World of Warcraft: Forever launches November 4, 2026 — rates below are launch-day rates, nothing is on sale yet

WoW Forever Gold → Guides → Which profession pays

Guide · the money side of tradeskills

Which Profession Actually Pays

Every profession guide lists what a trade can make. Almost none of them ask the only question that decides your income: does anybody buy it twice?

One test, and it sorts everything

Split the trades by what happens after a sale. A consumable is used up and bought again — ammunition, bandages, potions, food, enchanting materials. A durable is bought once and worn for 10 levels — armour, weapons, bags, goggles.

Consumables give you a customer who returns on a schedule. Durables give you a customer who returns when they outgrow the item, which on the way to 60 is often never, because the next upgrade drops in a dungeon instead. That difference decides more about your income than the recipe list does.

There is a third case the two columns miss, and Engineering is where it shows up. A bound-on-pickup consumable is used up exactly like any other consumable — the crafted ammunition a hunter burns through in a night is gone by the end of it — but it cannot be sold, because the only character allowed to use it is the one who made it. The demand is real and it repeats, and none of it reaches the auction house. For the crafter it is a saving on their own supply bill; for the profession it is not income at all.

That matters when you judge a trade by its recipe list, because a BoP line reads like the strongest entry on the page and contributes nothing to the ledger. The useful question is not whether a recipe is consumed but whether somebody other than you can be the one consuming it. Engineering earns through the parts of its list that trade freely and through the materials it feeds on, not through the items locked to the maker, which is the same filter worth applying to any trade before you count it as a source of gold earned at each level.

The same bind shows up at the top of the durable column, and it changes what you are actually selling. The better blue recipes in most crafting trades produce items that bind when they are made, so the finished piece never reaches the auction house and never appears in a price history. What trades instead is the service: the buyer brings the reagents, the crafter supplies the skill point and the recipe, and the payment is whatever the two agree on in the trade window. The item is not inventory, it is work done on someone else's materials.

That has three consequences worth planning around. The first is that income from a bound recipe is unpredictable in a way a listing is not, because a tip is negotiated each time rather than set by the market. The second is that you are competing on availability rather than on price, since a buyer who has already gathered the reagents will take whichever crafter is online and has the pattern. The third is that the materials cost is not yours, so the only thing you spend is the time to meet the buyer, which makes the line worth keeping even when the broader auction house price of the reagents moves against crafters. Count the bound blue recipes as a reason to hold a trade, not as a product you can list and walk away from.

Enchanting is the clearest version of that case, because most of it never produces an item at all. The enchantment goes straight onto the buyer's weapon or armour, so there is nothing to post, nothing to hold in a bank slot and no price history to compare against. With no vellums or scrolls to carry the work, the profession trades the way bound recipes do: the buyer brings the gear and usually the reagents, the enchanter supplies the formula, and the payment is a tip agreed in the trade window.

That puts the income in one place, the capital city bank and auction steps where buyers gather and ask for the formula they want, and it makes availability the thing you compete on rather than price. The part of the trade that does list is what comes off disenchanting: dust, essences and shards are consumed by every other enchanter working through skill points, so they repeat on a schedule and carry an auction house price for enchanting materials you can plan around. If you want listings as well as tips, that is the side of the profession to count.

If the bind comes off the top-tier ammunition, that line moves out of the self-supply column and into the consumable one, and the arithmetic changes with it. A consumable that every engineer can make and every hunter burns through is the easiest thing on the table to undercut, because the seller who wants the slot cleared will take whatever covers the trip to the mailbox. The failure case is not slow sales, it is a stack price that settles under what the ore and powder cost to assemble, at which point the recipe is still recurring demand and still a loss on each batch. Price the inputs first and treat the auction house price for crafted ammunition as the number that decides whether the line is worth running, not the fact that it gets used up.

Tailoring sits on the other side of the same test for a reason that has nothing to do with the recipe list. Beta reports describe quest reward sets that are thin for cloth wearers, so a caster levelling through the middle brackets goes longer between upgrades than a plate or mail character does. That stretches the window in which a crafted robe or set of bracers is still an improvement, and a durable only sells while that window is open. It is still a one-off sale per buyer, so the trade does not become recurring demand; it becomes a durable with a wider buyer pool and a slower replacement cycle, which is the best a durable gets. We have no counts behind the reward gap, only what players reported during testing, so check the reward tables for the bracket you intend to sell into before committing the skill points.

The beta forums put the same problem in blunter terms. The recurring complaint is that a crafting trade is worth having up to roughly level 30, and then stops paying its way until 60, because the gear it makes in the middle brackets is replaced by dungeon drops faster than the skill points arrive to make the next tier. That is a player judgement rather than a measurement, and we have no numbers behind it, but it describes the durable column exactly: a sale that depends on the buyer being between upgrades, in a stretch of levelling where nobody stays between upgrades for long.

Testers described the same squeeze from the seller's side. Crafted durables in the middle brackets were reported moving at less than the materials cost to assemble them, because the maker needed the skill point more than the margin and listed the output at whatever cleared the slot. When that happens the recipe is still consuming ore, cloth or leather at full price and returning part of it, so the loss is paid on the way up and the sale only decides how much of it comes back. Anyone planning a crafting slot around resale should price the inputs against the auction house listings for crafted armour in the same bracket before assuming the two meet.

One beta report made the comparison concrete. A quest handed out the Enchanted Gold Bloodrobe as a reward, and testers rated it behind the Crimson Silk Vest, which was going for 20 silver at auction at the time. A tailored item beating a quest reward sounds like demand, but it is the cheap end of the tailoring list doing it, and the buyer needs it once. We have no figures for how long that held or what the vest cost later, so treat it as a single observation from testing rather than a price to plan around.

The stat rework changes the durable column in one specific place. If spell damage and healing are folded into a single Spell Power value, that value lands on items the original tables never carried it on, and crafted plate and mail are where the effect is visible. A caster in a plate or mail class has very few drops competing for those slots, so a blacksmith selling a piece with the new stat is not undercutting a dungeon reward — there is no equivalent reward to undercut. That is the one case where a durable sells on the strength of having no substitute rather than on the buyer being briefly between upgrades. We have no item list for which crafted pieces carry Spell Power or at what level they appear, so confirm the recipe before planning a slot around it.

This is also what it costs to pick Engineering instead. Engineering's durable output is concentrated in pieces that bind when they are made, so the crafter wears them and nobody else buys them, and its recurring line is the ammunition the maker burns through on their own character. A blacksmith with a gear slot that nothing else fills is selling a one-off item, but it is a one-off item with a buyer pool and a listing, which Engineering's best durables do not have. Judge the two by what reaches the auction house rather than by the recipe count, and price the inputs against the auction house listings for crafted plate before assuming the shortage carries the margin.

The hunter case shows what that costs in practice. A hunter carries Engineering and Leatherworking not because either one sells, but because one supplies ammunition and the other supplies a quiver, and both are needed by the character making them. Two profession slots go to self-supply, and neither produces a single auction listing. The saving is real and the income is zero, which is the clearest argument on this table for putting at least one gathering trade on a character who intends to earn, since ore, leather and herbs have no dead bracket and keep their auction house demand for gathered materials at every level in between.

There is one more column that this table used to ignore, and it does not get paid in gold at all. Forever ties account progression to crafting skill: passing a skill threshold in a crafting trade is reported to award Legacy points, and raising a gathering skill to the same level awards none. Mining, Herbalism and Skinning keep every advantage listed above — no dead bracket, no bind, a buyer at every level — and they contribute nothing to that track. So the two gathering slots that look like the obvious earning choice are also the two slots that leave the account progression line flat.

That turns the profession decision into a trade between two different currencies, and they do not convert into each other. A crafting trade with a thin middle bracket may still be the better slot on a character you intend to keep, because the skill-up thresholds pay out whether or not anybody buys the output, including on the bound recipes that never reach the auction house. A gathering trade pays in coin and stops there. The practical split is to keep the earning trade on whichever character is funding the rest — the levelling arithmetic for that sits with gold earned at each level — and accept that the crafting slot is being carried for the progression rather than the ledger. We have no figures for how the thresholds are spaced, so treat this as a reason to check before committing skill points, not as a number to plan around.

The threshold reward points back at gathering from a direction the two columns miss. Skill points in a crafting trade are bought with materials, and the thresholds that pay out are reached by consuming ore, leather, cloth and herbs in volume. Whoever is supplying that volume is either gathering it or paying the auction house price for crafting materials for every point on the way up. A gathering trade feeds the progression track directly while still selling the surplus, so the slot that earns nothing on the Legacy line is also the slot that makes the Legacy line affordable.

There is a cost in that arithmetic that never shows up on a materials list. Every craft has a cast, and a skill bracket is cleared one cast at a time, with the stack split into single items that each have to finish before the next one starts. The reagent bill is what the bracket costs in coin; the casting is what it costs in hours at the forge or the anvil, and those hours are not spent gathering, running dungeons or listing anything. A trade that needs a long run of cheap crafts to move through a bracket can look inexpensive on reagents and still be the slower slot to raise.

That is worth pricing before the skill points are committed, because it changes how the two columns compare. Gathering skill goes up while the character is already moving through a zone and collecting things worth selling, so the time spent raising it produces inventory at the same moment. Crafting skill goes up during casts that produce items which are frequently vendor fodder at that tier. Neither position is wrong, but the comparison between them is only honest once the casting time sits next to the reagents in the same total, alongside the gold earned at each level that the same hours could have covered instead. We have no figures for cast lengths or bracket sizes in Forever, so this is a column to measure on your own characters rather than a number to plan around.

The second half of the argument depends on something we cannot confirm. The reward is described as account progression, and if a threshold that has already been passed once does not pay again on a second character, then the crafting slot only needs to exist somewhere on the account, not on every character. Every alt after that is free to carry 2 gathering trades with nothing lost on the progression side. We have not seen this tested and the award rules may work differently in practice, so check whether a repeated skill-up pays before planning a roster around it — but if it does not, the case for gathering on everything except the one crafter gets considerably simpler.

The same caveat applies to the recipes handed out by Waylaid supply crates. Some of what comes out of the crate reward line produces items that bind when crafted, which puts them outside both columns on this page: they are neither a durable you sell once nor a consumable somebody rebuys, because no buyer ever sees them. The recipe still counts as progress on the skill track and the item can be genuinely good for the character holding it, but the ledger does not move.

That also removes the usual fallback for a crafter with more than one character. A bind-on-pickup craft cannot be posted, cannot be mailed to an alt, and cannot be made on the main for the alt who actually needs it, so the value stays with whichever character ran the turn-ins. Worth checking which side of that line a crate recipe falls on before treating it as a reason to pick the trade, and worth treating the freely traded part of the list as the only part that shows up in auction house income from crafting.

Blacksmithing is the clearest example of both problems arriving at once. Beta reports describe a recipe track that sits behind the Azeroth Commerce Authority: the plans are bought with a faction currency, and the currency comes from daily reputation quests rather than from anything you can farm on a schedule that suits you. The gate is time rather than materials, so a second smith cannot catch up by spending more, and the smith who does the dailies ends up with plans nobody else has. That sounds like an advantage until you look at what the plans produce.

What they produce is mostly bound on pickup. The output is armour and weapons for the character who ran the turn-ins, which means the trade combines the weaker half of both columns on this table: a durable, so the same buyer never needs a second one, and bound, so there is no buyer in the first place. The handful of freely traded entries is where the whole income of the profession has to come from, and the list is short. We have no figures for how short, and the beta is the only sample, so treat the shape of it rather than the size as the useful part.

Income is only half the comparison, and the other half is what a skill point costs to buy. Crafting pays for points in materials: a yellow recipe either advances the skill or it does not, and the attempt consumes the inputs either way. The Working overtime talent is meant to soften that, and the stated figure players quote for it is a 91% chance on yellow recipes, but the recurring beta complaint is runs of consecutive failures that do not look like the stated number. We have no measurements behind either side of that, only what testers reported, so treat the complaint as a reason to budget for failures rather than as a correction to the figure.

The practical effect is that the cost per point is not the material cost printed next to the recipe, it is that cost plus whatever the failed attempts burned before one landed. On a bracket where the yellow range is long, that difference decides whether a trade is worth levelling at all, and it is paid in stock you could otherwise have sold, so price it against the auction house price of the materials rather than against the time spent gathering them. Gathering has no equivalent line: a point there costs travel and nothing else, which is why the two sides of this comparison are not settled by income alone.

Enchanting hits the same wall from the other side, and Wizard Oil is the clearest case of it. The oil is a consumable, which puts it in the column that is supposed to pay, but one of its reagents is a vendor purchase rather than a drop. That sets the floor under the recipe in the shop instead of on the market: with Maple Seed at 2 silver each, the reagent bill alone sits above what the early oil goes for once a second enchanter is posting, so the batch is made at a loss while the demand for it is still genuine. Recurring demand does not rescue a line whose input price cannot be undercut, because every other crafter pays the same vendor and none of them can go lower.

The route crafters normally use around that is closed here as well: buying oils off the vendor and flipping them is not a working source of margin in Forever, so the recipe has to stand on its own numbers or not at all. The practical reading is to treat the early enchanting brackets as a skill-up cost rather than a source of income, and to compare the auction house price for Wizard Oil against the vendor reagent before committing to a run. If the gap is negative, the skill points still arrive and the gold does not, which is a different outcome from a slow seller and worth telling apart before the materials are bought.

None of that makes Blacksmithing a bad trade to carry. A smith who wants the gear gets the gear, and the daily quests are work they were going to do anyway. It does make it a poor answer to the question this page asks, because the recurring demand it generates is demand from the crafter for their own output, and that never reaches the auction house. If the goal is gold made while levelling rather than equipment, the currency gate is a cost with no sale at the end of it.

The reputation route has a second cost that is easy to miss while the standing is still climbing. Daily turn-ins are paid for in time, and the plans they unlock are permanent, so the spending feels like it builds something. It does, but what it builds is equipment rather than stock. A recipe that produces a bound item converts every hour of reputation work into gear worn by one character, and gear worn by one character cannot be listed, cannot be mailed to an alt who needs it more, and cannot be turned back into coin when the next tier makes it redundant.

Enchanting fails the same test from the other direction. The demand genuinely repeats: an enchant is destroyed the moment the item under it is replaced, and on the way to 60 items are replaced constantly, so the same customer comes back for the same enchant on a new weapon within a few levels. What breaks is not the demand but the way it has to be sold. Without vellums there is no object to post, so the enchant is applied directly to an item the buyer is holding, which means both characters have to be standing in the same spot at the same time. The trade is arranged by voice in a city channel and completed by hand, not listed and forgotten.

That changes what the profession is paying you for. An auction listing works while you are offline; a trade chat advertisement works only while you are in a capital repeating it and waiting for someone to answer. Income tracks hours spent standing in the city rather than the number of players who want the enchant, and the enchanter who logs in to farm instead of to advertise earns nothing from the recipe list at all. The part of Enchanting that behaves like a normal trade is the materials: dust, essences and shards are freely traded, consumed by every other enchanter, and sell the way any other auction house demand for gathered materials does. Treat the spell list as a service you have to be present to provide and the material supply as the part that earns on its own.

That is the difference between an investment and an expense. Working capital is anything you can sell to fund the next batch: ore in the bank, a stack of cloth, a finished item somebody else can equip. A bound craft is none of those, so the reputation track never compounds — each day of turn-ins buys the right to make another item for yourself, and the balance at the end of the week looks the same as the balance at the start. A gathering trade behaves the opposite way, because everything it produces is tradeable the moment it leaves the node and keeps its value as resellable materials whatever the gatherer decides to do next. Before committing to a reputation line, it is worth checking which of its recipes make something a buyer can own, and treating the rest as a purchase rather than as income.

Cooking is the plainest illustration of the first half of that sentence. Beta testers describe a dish called Spicy Wonderfowl Treat that behaves the way a consumable is supposed to behave for the person selling it: the buff runs for 1 hour, it is lost on death, and the ingredients are cheap meat rather than anything contested. Every one of those three properties is a reason for the same customer to come back. The timer alone guarantees repeat purchases on a long evening; the death clause adds the ones nobody planned for, and the cheap inputs mean a wipe-heavy group can keep rebuying without thinking about it.

Compare that with the durable side of the table. A crafted chest piece is one sale to one character and then silence until they outlevel it. A stack of food is one sale to one character this week and another next week, and the buyer does not need to have improved at anything in between. That is the whole argument of this page in a single recipe, and it is why Cooking sits in the constant column despite making nothing anybody would call an upgrade. We cannot say what the dish costs to make or what it lists for, since the beta is the only sample and it wipes on 4 November, but the shape of the demand is readable now, and it matches the rest of making gold at each level more closely than most crafting trades do.

There is a second reason the food line holds, and it has nothing to do with the stat on the dish. In Forever, a well-fed buff carries a 5% experience bonus alongside whatever the recipe gives in stats, and there is no plainer version of the recipe that leaves the bonus out — a cooked dish is a cooked dish. That turns food from a raid consumable into a levelling consumable. A player who would never buy a stat buff for a questing session will still buy the thing that shortens the session, and they will buy it again tomorrow, because the buff goes away when they log off or die.

For a seller the practical consequence is the size of the customer list rather than the price of the item. Potions sell to characters doing something difficult; food sells to everyone moving from 1 to 60, which is most of the realm for the first weeks. The same recipe serves a dungeon group restocking between pulls and a solo player grinding a zone, and neither of them finishes the stack. That is the clearest case on this table of a product whose demand does not depend on the buyer improving at anything, which is the only property that makes gold earned at each level repeatable rather than a one-off.

Spellpower oils and elixirs follow the same logic from the caster side, and in the beta threads they turn up as a direct answer to a new player asking what to keep in the bags rather than as a theory about consumable demand. A weapon oil runs out on a timer, an elixir is gone on death or on logging off, and neither of them survives into tomorrow's session. The buyer is back for the same reason the food buyer is back: the item ended, not because anything about the character changed.

What makes the line steady is the gap underneath it. Drops in the levelling brackets hand out spellpower in small amounts and not in every slot, and the crafted gear available at the same level does not reliably close the difference either, so the only part of the kit a caster fully controls is the consumable. Until the drops and the level-appropriate crafts cover the need, that gap is filled by something bought again each time. We cannot say which recipes or prices Forever will settle on, since the beta is the only sample, but the demand is of the repeating kind rather than the one-off kind, which is what decides whether a trade is worth the grind and what to watch for in auction house consumable prices once the realm is live.

One column on this table is less stable than it looks, and that is the part of a trade that sells to an NPC rather than to a player. A craft that is only worth making because the vendor buys it back above the cost of its materials depends on a value that lives in the game files, not in player behaviour, and that value can be edited. On 10 November beta testers reported exactly that: a vendor return that had been worth crafting for stopped being worth it, with nothing in the notes to mark the change. Nobody announced it and nobody had to.

The practical difference is who you have to watch. Player demand moves slowly and visibly, so a consumable line gives you warning before it dries up. A vendor line gives you none, because the only signal is the number itself changing between one session and the next. That does not make vendor sales useless, but it does mean they belong in a different column than the recurring player demand this page is built around, and it is worth checking the current return by hand before committing a long crafting run to it rather than assuming the figure you measured last week still holds. The same caution applies to anything you plan around in auction house income from crafting: the part priced by players is the part you can reason about.

There is a cost on the consumable side that the demand argument hides, and it is measured in minutes rather than in materials. A player comparing the two clients reports that crafting casts in Forever run longer than their Classic equivalents: the tiers they timed came out at 5 seconds and roughly 8 seconds per craft, against 3 seconds for the same kind of action in Classic. One batch of 158 grenades, counting the time spent getting the materials ready, took about 50 minutes from start to finish. We cannot say whether those durations apply to every recipe or whether they stay as they are once the realm is live, since a single player's timings are the only sample.

What that changes is which half of the comparison on this table you are actually paid for. A gatherer is paid per node and a crafter is paid per cast, so a long cast turns a thin margin per item into a thin margin per hour, and the trades that sell in stacks are the ones that feel it first — ammunition, food and oils are only worth selling in volume. The margin on one unit is still the right number for deciding whether a recipe sells, but it stops being the right number for deciding what to do with an evening, and it is worth timing a full batch yourself before treating a repeat-demand line as a reliable source of gold earned at each level.

The second consequence is that Cooking starts paying earlier in its own skill track than most trades do. A durable craft only becomes sellable once the skill is high enough to reach a recipe somebody wants, which means the whole early half of the track is spent making things that exist to raise a number. Food does not work that way: a dish that applies the well-fed buff applies the experience bonus with it, and a levelling buyer is not comparing the stat line against a better dish they cannot afford yet. The low-skill recipes are selling the same thing the high-skill ones sell.

The forum threads from beta make the same point from the losing side. Testers levelling a durable trade report the familiar early-track problem: the cloth, leather and bars needed for a low-level recipe cost more bought off the auction house than the finished item fetches when it is listed, so each craft is a loss taken on purpose to move the skill number up. That is not a pricing mistake by the sellers. It is what happens when the inputs are contested by several trades at once and the output is wanted by nobody who is not already past that level bracket.

Gathering trades have their own version of that trap, and Skinning is where it catches people out. The skill number controls what you are permitted to skin, not what the corpse contains. A character sitting at 100 Skinning who spends the evening on low-level beasts in a starting zone will finish it with scraps and very little else, because the drop the mob carries is set by the mob, not by the gatherer. Raising the skill past a threshold opens access to tougher targets; it does not improve the yield of the targets already being farmed.

So the honest way to judge a gathering trade is by the route rather than by the number next to the profession. Before deciding Skinning is underpaying, check what is actually being skinned: which zone, which mob family, and whether anything there drops the hide grade that sells. A player who moves to the right beasts sees the income change without touching the skill track at all, and one who grinds the skill up while staying on the same mobs sees nothing change. The same reading applies when you compare the result against leather and hide prices on the auction house, since a thin evening is usually a routing problem and not evidence that the trade itself is weak.

Enchanting sits at the far edge of the same table, because it produces nothing that can be put in a bag and listed. The dust and essences that come out of disenchanting are consumed by the skill track itself, and the finished enchant is applied straight onto somebody else's weapon or chest, so the thing a customer pays for never exists as an item. An enchanter has no stock. What is being sold is a service, and a service only sells while the seller is online and visible in a trade channel or standing where buyers gather.

That changes what the profession is worth rather than whether it is worth anything. The demand repeats on the same schedule as gear replacement, which during levelling is frequent, and it repeats again every time a buyer rerolls or regears at the cap. But none of it accumulates while you are logged out, and none of it benefits from listing overnight the way a stack of food does. The part of Enchanting that does trade freely is the materials, which other enchanters buy to push their own skill up, so the auction side of the trade is a supplier business and the enchant side is a time business. Both are real income and they behave nothing alike, which is worth separating before you judge the profession against anything else in auction house consumable prices.

The vendor floor finishes the job. Starter gear sold by NPCs covers the same slots as the first tier of crafted armour and weapons, and a buyer comparing the two is not weighing stats so much as weighing a walk to a vendor against a trip to the auction house. A crafted item has to beat that convenience before it can ask for a price at all, which most low-level recipes cannot do. We cannot say how wide the gap runs in Forever, since the beta is the only sample, but the direction matches what auction house material prices do to any craft whose output has a vendor substitute: the materials keep their value and the finished item does not inherit it.

We cannot say which recipes sit at which skill values in Forever, or how early the first dish with that buff becomes available, because the beta is the only sample anyone has worked from. What is readable is the shape: the gap between starting the trade and having something worth listing is shorter here than on the durable side, and the early output is a consumable rather than a stack of vendor fodder. If you are picking a trade for income rather than for the recipe list, that is worth weighing alongside auction house consumable listings when you decide how much skill to grind before expecting the first sale.

There is one more skew worth checking before you commit to a durable trade, and it sits inside the recipe list rather than in the demand for it. Crafted leather and mail lean heavily toward Agility and Strength, so the trainer-taught part of the tree points most crafters at the same armour buyers. The Intellect and spell power versions of the same slots are thinner on the ground, and the ones that exist tend to sit behind drops and faction vendors rather than behind the trainer. The practical result is a market where the physical sets are listed by everyone who levelled the trade and the caster pieces are listed by whoever went looking for the pattern.

That changes what a durable craft is worth to you. A recipe anybody can train is priced by the most impatient seller holding it, while a recipe that took a dungeon run or a reputation grind to obtain has a short list of competitors and a buyer group with few alternatives. We cannot say which patterns drop where in Forever, or how the faction vendors are stocked, since the beta is the only sample. The useful habit is to read the supply side before the stat line: check how many sellers are already posting the Agility sets against auction house caster gear listings, and treat the thin column as the one that justifies the effort of chasing a recipe.

One change in the announced ruleset pushes the constant column further in the same direction. Downranking — casting a cheaper, lower rank of a spell to stretch a mana bar — has been nerfed, so the habit casters and healers rely on to finish a fight on one bar works less well than it did. The practical result is more mana spent per pull and more time spent refilling it: more water carried, more drink breaks between packs, more mana potions bought before a run rather than after one goes badly. We cannot put a figure on the difference, because nobody has played the live version of it yet, but the direction is not ambiguous.

For a seller that is the repeat-demand test answering itself. Water, food and potions are consumed by the same characters every evening, and a rule that makes casting more expensive adds customers to that line without adding anything to the durable side — a chest piece does not sell twice because healers drink more. If you are weighing trades on recurring income, this is a reason to read the auction house for consumables as the steadier half of the market rather than the boring one.

Enchanting sits outside both columns, and the reason is mechanical rather than economic: the build has no enchanting vellums. An enchant is applied straight to an item, so there is no object to list — nothing goes on the auction house, and no buyer finds you by searching for what you sell. The demand itself repeats as reliably as any consumable, since every new weapon and every replaced chest piece is another application, but it reaches you through trade chat, a guild channel, or whoever is standing by the enchanting table when you are.

That changes the arithmetic in two places. Skill goes up through your own gear and through work done for other players who bring the materials and pay for the application, rather than through a batch of goods you craft once and sell while logged out. And income tracks the hours you are visible: a seller with consumables on the auction house keeps earning offline, an enchanter earns only while answering in the channel. The tradeable half of the trade is the materials — dust, essences and shards list and move like any other input — so if you are judging the profession as a source of gold earned at each level, count the materials as the market line and the enchants themselves as a service business with no shopfront.

That recurring line also has a date attached to it. Raid content in Forever is announced in 10, 20 and 40 player sizes, and the consumable demand that comes with organised groups does not build evenly from the first evening of the realm — it forms by 9 December, which is the point by which rosters are set and clears run on a schedule rather than whenever people happen to be online. Up to then, food, water and potions sell to levelling characters one or two stacks at a time. After it, the same recipes sell to one person buying for 9, 19 or 39 others.

The other thing that moves these columns is balance work done before the realm opens. A pre-launch hotfix that strengthens a class does not only change how that class plays — it changes how many people roll it in the first week, and every one of those characters arrives with the same shopping list. A buff to hunters pulls ammunition, pet food and leather armour together, because the class consumes all three and nothing else consumes them in the same volume. A buff to a caster pulls water and reagents instead. The recipe you can make did not change; the number of characters who need it did.

The caveat is that the same lever works in reverse, and that is the direction sellers get caught by. A class tuned down in the last pass before the realm opens loses characters to rerolls, and a reroll does not carry the old shopping list with it. The player who was buying ammunition and leather starts buying water and reagents, and the stock built for the first list sits while the second one empties. Nothing about the recipes changed; the population behind them moved. The same applies to the durable half of the table, since armour type follows the reroll as well — mail and leather demand drop together when the people wearing them switch to cloth.

We cannot say which classes the final tuning will touch, because that is decided after the beta sample ends and the notes are not out yet. What is worth planning around is the shape of the risk rather than the specific class: a deep stock in one class-specific line is a bet on the roster of the first week, while the same gold spread across water, food and bandages is not, because those sell to everyone regardless of what they rolled. If you are reading the auction house for consumables in the opening days, the lines that serve every class are the ones that survive a reroll wave intact.

Forum threads where players post their planned builds show the same effect before anyone has logged in. The pairings people name are the familiar ones and they cluster by class: hunters and rogues writing down Skinning with Leatherworking, warriors and paladins pairing Mining with Blacksmithing, casters taking Herbalism with Alchemy, and a visible group of clothies planning Tailoring with Enchanting so the greens they make feed the disenchant pile instead of the vendor. What is worth noting is not that the pairings are sensible but that almost nobody plans to gather one thing and craft another, which means most of these characters will be supplying themselves and buying only what their own gathering slot does not cover.

Read as a demand forecast, that skews the first weeks towards a short list of materials rather than spreading across the table. Leather and light ore get worked by the people who collect them, while cloth, which no gathering profession produces, falls out of mobs into the bags of players who have no use for it and every reason to list it. The same goes for cheap meat, which the Cooking crowd buys rather than farms. If the planned pairings hold, the gaps are in the materials nobody chose a profession to collect, and that is where the auction house for consumables is likely to be thin first. We cannot weight any of this, since a forum sample is self-selected and plans change in the first evening, but the direction of the gap is readable from what people are writing down.

We cannot say which classes will be adjusted, or how hard, since that is decided by the developers and announced when it is announced. What is worth doing is reading the change notes with the supply question in mind rather than the gameplay one: ask which consumables the buffed class burns through and which durable slots it fills from crafted gear. A trade that sits on the receiving end of a popular class has a wider customer list for as long as the fashion lasts, and a trade that sells to nobody in particular keeps whatever demand it had. That is the same recurring-income filter applied one layer up, and it is the fastest way to tell which half of the auction house for consumables is about to get busier.

The practical difference for a crafter is order size rather than product. Nothing on the recipe list changes on that date; what changes is that a single buyer clears a shelf instead of a slot, and comes back the following week because the stock was used. A seller who wants that trade has to have inventory ready before 9 December rather than start levelling the profession when the first group asks, since the materials behind a 40 player restock take longer to gather than the restock takes to sell. We are not going to guess which consumables each raid size leans on hardest, because nobody has run them on a live realm yet, but the volume shift itself is a scheduling problem more than a pricing one, and it is a reason to treat the auction house during levelling weeks as preparation time for the raid weeks that follow.

Organised rosters usually write the same arrangement into their joining rules, and it is worth reading as a demand statement rather than as a chore list. The main carries a crafting trade, the alts carry gathering, and every member is expected to arrive with full consumables for the night — food, water, potions and whatever the fight asks for. The first two lines are one-off: a character picks up the profession once and the requirement is satisfied for good. The third line resets every single raid night, which is why it is the one that moves gold across the realm week after week.

Engineering on a hunter is the same test read from the weapon side. Crafted ammunition is spent as it is fired, so a hunter who raids on a schedule works through stacks the way a healer works through water, and the person making the bullets sells to that character again the next week without either of them having improved at anything. A leatherworker selling the same hunter a crafted chest or a set of bracers makes one sale instead, and the next one waits until the buyer outgrows the piece or stops wearing crafted gear altogether. Both trades make something a hunter wants; only one of them has a customer who comes back on a timer, which is the whole of the argument on this table.

Forum posts about Forever hunters usually attach a specific damage figure to the better ammunition, stated as a flat percentage gain over the vendor option. We are not repeating that number, because it comes from player testing rather than from anything announced, and nobody has run it on a live realm long enough to confirm it. The part that matters to a seller holds either way: as long as the crafted ammunition is believed to be worth carrying, it is bought in stacks and burned in an evening, which puts it in the same column as food and potions on the auction house for ammunition rather than in the durable column with the armour.

There is a second way to earn from a trade that this table does not measure at all, because it never reaches the auction house. In a commission craft the customer brings the materials and pays only for the work: they hand over the bars, the leather or the cloth, you run the recipe, and the fee is for having the pattern and the skill rather than for the goods. The crafter puts up nothing. No stock is bought in advance, no stack sits unsold, and a recipe that nobody lists profitably can still pay, because the question stops being what the item sells for and becomes whether anyone else on the realm can make it.

That changes which professions are worth keeping. Durables lose the repeat-demand argument when you sell them as a service, since the one-off chest piece costs you nothing to produce and the fee arrives either way; the risk of the materials sits with the buyer, who wanted the item anyway. It also explains why crafting guilds organise around rare patterns rather than around volume — a pattern few people hold is a steady stream of commissions, while a common one is just competition on price. We cannot say what a fee looks like in Forever, because that is set by players rather than by the ruleset, but the model works alongside selling output rather than instead of it, and it fits the same habit as the rest of gold made while levelling: charge for the part nobody else can supply.

The durable half of the table is not uniformly weak either, and the reason is who shops for which stat. Crafted leather and cloth carrying spell power and Intellect look like caster-only stock on the recipe list, but under the current Forever mechanics retribution paladins scale with spell power as well, so they shop in the same bracket as mages, priests, druids and shamans rather than in the strength bracket their armour class suggests. A tailor or leatherworker listing an Intellect piece is therefore not selling to one archetype, and the buyer list is wider than the item tooltip reads.

For a crafter the practical effect is on how long a piece sits rather than on what it sells for. Stat-specific durables normally die on the shelf because the slice of the realm that wants them is small; a stat that two different playstyles are buying keeps the listing moving, and it keeps moving while players are still levelling and respeccing rather than only during gearing weeks. That does not turn armour into a consumable, since each buyer still takes one piece and leaves, but it is the one case where a durable line holds up on repeat relistings, and it is worth checking against the auction house for crafted armour before writing the gear recipes off entirely.

On the durable side of the table there is a change worth noting before you pick a trade for the armour market. Leatherworking in Forever has recipes carrying Intellect and spell power, which is the stat line Tailoring used to hold on its own. The buyer that matters here is the caster who was never locked into cloth in the first place: balance and restoration druids, and shamans, all of whom could already wear leather or mail but bought cloth anyway, because that was where the casting stats were. If the same stats now appear on leather, that part of the customer list stops being Tailoring's by default.

The effect is not that one trade stops selling. It is that a single sale, which was already a one-off, now has two suppliers competing for it, while the buyer still only needs the slot filled once. That is the opposite of what the repeat-demand test rewards: durables already depend on a customer who may never come back, and splitting them between two trades thins each seller's share of the same list. We cannot say how many recipes overlap or at which levels they land, since that depends on the final recipe lists rather than on anything measurable now, but if you are choosing a crafting trade for income, treat caster armour as a contested line and check auction house listings for armour before committing materials to it.

That split is the argument of this table stated by somebody else. A guild that demands a crafting trade on the main is not creating a customer, because the member supplies himself; a guild that demands a full consumable bag before every pull creates a customer who buys again on a schedule set by the raid calendar rather than by his own progress. We cannot say which Forever guilds will phrase the requirement this way, since none of them have run a clear yet, but the shape is the same one that makes gold earned at each level repeat instead of ending after a single sale, and it points at the consumable column rather than the crafted gear column.

The durable side has a second problem that the table above cannot show, and it comes from how the new blue crafted pieces are bound. Several of them are bind on pickup, which means the crafter cannot list them at all — the item exists for the character who made it and for nobody else. A recipe that would have been one sale to one stranger becomes zero sales, and the only value that reaches the market is the materials that went into it. We are not going to claim a count of how many recipes fall under that rule, because the beta list is still moving, but the direction it pushes players is already visible.

That rule is also where the choice stops being about recipes and starts being about your own character. Leatherworking leans toward sets built for a specific class, which is useful if you are the one wearing them and worth nothing to the market if the pieces cannot be listed. Tailoring paired with Enchanting sits on the other side of the same decision: the cloth gear covers a healer or caster through the early levels, the disenchanting turns the leftovers into materials, and bags come out of the same trade as something any character on the realm will take. One combination equips you, the other equips you and leaves stock behind.

Neither is wrong, but they pay differently and it helps to know which one you picked. Gear made for your own class is a saving, not an income — it removes a purchase instead of adding a sale, and the moment you outlevel the piece the saving ends. Bags are the opposite case on the durable side of the table: one sale per character rather than per week, but every new character needs them and nobody replaces a bag because the next tier of content opened. If you are choosing a trade mainly to fund the rest of the account, the side that produces something a stranger can buy is the one that connects to the auction house for bags rather than to your own equipment list.

Tailoring shows the same restriction from the seller's side. Part of what a tailor can make is bound on pickup, so those pieces never reach a listing — they are gear for the character who holds the profession, not stock. That does not make the trade worthless, but it does mean the recipe list and the sellable list are two different lists, and only the second one earns anything. When you weigh a durable trade against a consumable one, count the sellable half only, because a recipe you cannot put on the auction house for crafted gear contributes nothing to income no matter how good the item is.

There is also a cost that does not come out of the ground. Some tailoring work uses dyes bought from a vendor for gold, which means part of the input is a fixed purchase rather than something you gather. Gathered materials cost time and can be replaced by farming; vendor reagents leave your pocket every time you craft, whether or not the result sells, and on a bind on pickup result they leave it with no sale on the other side at all. We are not listing which recipes take which reagents, since the beta list is still changing, but the principle holds when you compare trades: a consumable line pays the material cost back on every sale, while a durable line with vendor inputs can spend gold on items that were never going to be listed.

Camp crafting has a ceiling of its own, and it is a rule rather than a price. Only one camp item can be placed at a time: putting down a tent replaces the fishing, skinning and bandaging buffs that the other camp items give, and each new placement pushes out the last one. A player who owns the full set still benefits from exactly one of them in any given moment, which means the set does not multiply demand the way a stack of food does. The second, third and fourth camp item are convenience purchases for a character who already has the first, not consumables that run out and get rebought.

For a seller that puts camp goods closer to the durable column than their price tag suggests. The buyer comes once, picks the camp item that matches what they do most, and has no mechanical reason to come back unless the item itself is consumed on use — and nothing in the exclusivity rule creates that reason. We cannot say how many of these recipes will end up worth levelling for, since the beta is the only sample, but the limit is the part of the demand picture that does not change with tuning: one placement per character at a time is a hard cap on how often the same customer returns, and it is worth weighing against the auction house for consumables before committing a profession to it.

There is one more argument in circulation that this table cannot settle, and it belongs here as opinion rather than as mechanics. In forum threads about the Legacy system, players report that the account-wide perks they have seen so far attach to crafting professions and not to gathering ones, and several of them say they intend to drop gathering entirely once the lists are confirmed. We cannot verify that reading, and we are not going to repeat a perk list we have not seen on a live realm, but the belief itself is worth noting because people act on it before it is proven.

If that belief holds, it adds a term this comparison does not otherwise contain. The recurring-versus-one-off test weighs what a trade earns at the auction house; a Legacy perk is a benefit sitting outside the auction house entirely, and a player who values it will take a crafting trade with weaker repeat demand over a gathering trade with steadier income. For a seller the practical consequence is the supply side rather than the demand side: if enough people abandon gathering for perk reasons, the characters feeding ore, herbs and leather into the auction house for raw materials get thinner, and the crafters who stayed pay more for inputs. That is a scenario, not a forecast, and it is the part of choosing a trade while levelling that no table of recipes will answer for you.

There is a cost side to this choice that the income table never shows. You have 2 profession slots, and the pair that pays while you level is often not the pair you want at 60. Gathering trades fund the trip up because they sell into a market of levelling crafters and need no investment beyond travel time; a crafting pair earns at the top because it serves raid rosters that buy on a schedule. Plenty of players run the first combination to 60 and then want the second one, which means dropping a trade that is already maxed and starting the replacement from the bottom.

That restart is paid in materials rather than in a fee. Levelling a craft a second time means buying or farming the whole low tier again — the cheap recipes nobody raids with, made in bulk and vendored or disenchanted, plus the recipes you have to go back and learn from trainers and drops. The skill you dropped does not come back for free either; if you change your mind, it is the same climb in the other direction. We are not going to put a figure on what either climb costs, since prices on a fresh realm are set by what other people are farming that week, but it is spending that happens once per switch and does not earn anything while it is happening, which is worth weighing before you treat a profession as pure income. If you expect to switch, the usual way to cover it is to bank the proceeds from gold earned at each level on the way up rather than to spend them on gear that the next tier replaces.

There is one durable line that behaves like a constant anyway, and it exists because of a gap in the drop tables rather than because of a recipe. A paladin levelling from 1 to 30 who wants spell power on their gear does not find it on quest rewards and does not find it on mail that drops in the dungeons of that range — the armour handed out at those levels is built around melee stats. The only way that character closes the gap is to buy a crafted piece, which means the demand lands on a crafter and then on the auction house for crafted gear rather than on a boss.

What makes this worth naming on a page about recurring income is that it is not one buyer. Every paladin who levels through that range meets the same missing slot at roughly the same point, and the ones who care about spell power all arrive at the same short list of craftable options. The sale is still a durable — it is worn until it is outgrown, and it will not sell twice to the same character — but the stream of buyers refills on its own as long as people are rolling the class. We cannot say how many of them will pay rather than level in whatever they looted, since nobody has run the live realm yet, and we are not going to guess at a price for the pieces involved.

Gathering has a version of the same problem, and it shows up most clearly in ore. A miner is not selling to the realm at large; a miner is selling to the trades that smelt bars, and in Forever that list is short. Blacksmithing and Engineering draw from the same stacks, so the two of them are competing for one supply rather than widening it, and there is no Jewelcrafting in this version to absorb the leftovers the way it does in later expansions. Herbs feed Alchemy and the first aid line, leather feeds Leatherworking, but ore feeds two trades that often want the same metal at the same stage of levelling.

The practical consequence is that mining income depends on how many people are levelling those two trades at that moment rather than on how much ore you can pull out of a zone. A gatherer who can only mine is tied to a narrow customer list; a gatherer who also smelts, or who pairs mining with the crafting trade that consumes the bars, keeps the value instead of passing it on at the first sale. That is the same recurring-demand test applied to raw materials, and it is worth checking against the auction house for ore stacks before committing a second character to a gathering pair.

There is one more input that sits outside the recipe list entirely: which classes people actually roll. Demand for a crafted good is demand from the characters that need it, so a class that is projected to perform poorly ends up thin on the realm, and every service only that class provides goes thin with it. Rogues are the clearest case. If the class reads as weak in Forever and fewer players level one, the realm has fewer characters able to open a lockbox on request, and the lockboxes keep dropping at the same rate regardless.

That shortage does not remove the demand, it moves it to the trades that solve the same problem. Blacksmithing covers it with skeleton keys and Engineering with explosive charges, and both are consumed on use, which puts them on the repeat side of the test rather than the one-off side. The same logic runs through the rest of the class list — fewer of a class means fewer buyers for whatever that class alone consumes, and more buyers for the crafted substitute. We cannot put numbers on any of this, since no class population figures exist for a realm nobody has played yet, but it is worth checking which classes the community expects to be scarce before committing a trade to them, and worth watching auction house demand for lockbox keys once the realm is live rather than assuming it follows the beta.

Enchanting sits outside the consumable and durable split entirely, because the thing it produces is not an item. An enchant is an effect applied to equipment the buyer already owns, so there is nothing to put on the auction house and nothing that sells while you are offline. Every sale is arranged in trade chat, finished face to face in a city, and priced by agreement rather than by a listing — the buyer brings the materials and pays a tip, or pays for both and waits while you gather. The profession earns, but it earns per conversation.

That changes what the trade is worth to you rather than whether it is worth anything. A crafter who lists stacks of food can log off and come back to sold mail; an enchanter who logs off earns nothing until the next time they stand in a capital and advertise. The one part of the trade that behaves like the rest of the market is the materials — dust, essences and shards from disenchanting are ordinary items, they list like anything else, and for a lot of enchanters that line brings in more over a week than the enchants do. If you are judging trades by income that arrives without you, weigh the materials side on the auction house for crafting reagents separately from the service side, because only one of them works on its own.

What it pushes them towards is duplication. If the piece cannot be bought, the way to wear it is to carry the profession yourself, so the trade gets levelled again on an alt that wants the same armour rather than bought from someone who already has it. For a seller that cuts both ways: your customer list for that item disappears, while demand for the raw materials behind it goes up, since every player repeating the climb buys the same inputs. It is one more reason the gathering half of the auction house for crafting materials tends to hold its customers better than the finished-goods half, and one more reason to judge a crafting trade by what you can actually list rather than by what the recipe tooltip promises.

There is one window where the durable side still finds a buyer, and it sits lower than most crafters look. Between levels 2 and 19 the supply of blue cloth gear with intellect on it comes mostly from the crafting trades rather than from bosses: a caster at that stage has noticeably more craftable options than drops available to them, because the early dungeons simply do not hand out many of those pieces. A tailor or a leatherworker who stocks that range is selling into a gap, not competing with a loot table.

The limit is the same one that applies to every durable: the customer outgrows it. What makes the low-level window work anyway is volume of buyers rather than repeat purchases from one of them — new characters pass through levels 2 to 19 constantly, and each one passes through once. We cannot give a figure for how many such pieces exist or what they list for, since the beta list is still moving, but a trade with a buyer before level 20 starts paying for itself far earlier than one that waits for the endgame, which is the part of gold making at low levels most crafters skip.

Binding applies to more than crafted armour. Part of what a trade hands you comes as a drop or a reward rather than as something you make to order, and some of those are bind on pickup as well. That means the profession is producing an item for the character carrying it and nothing for the market behind it. The fishing ring discussed further down is the plainest version of that shape: the demand is obvious, since every player who hears about the ring wants one, and the supply on the auction house is nil, because every copy stays with whoever pulled it up.

The same test can be run on stats instead of slots, and for cloth casters in the 2-19 bracket the sources are badly lopsided. Intellect is close to absent from what bosses drop in that range, while it appears on 27 crafted cloth pieces. A levelling mage, warlock or priest who wants the stat therefore has two realistic routes to it, and both of them are purchases: a tailor who will make the piece, or a green with the right suffix bought off the market. Neither route is a drop they can farm on the way past.

For a seller that is the rare case where a durable produces something close to recurring demand, because the customer list refills itself. The gap is not closed by one sale — it reopens every time a character enters the bracket, which includes every alt the realm rolls, and the pieces are outgrown fast enough that nobody holds onto them. Tailoring picks up the made-to-order half of it, and the auction house for caster greens picks up the rest, since a suffix roll cannot be crafted on demand and has to be found. We are not going to put a price on either half while the beta is the only sample, but the shape of the demand is set by the item list rather than by how the realm settles.

For payback arithmetic those rewards belong in a different column from anything you can list. A sellable craft returns gold you can spend on whatever you like; a bound reward returns gear, and it is only worth the price of the item you would otherwise have bought to fill that slot. If you were never going to buy that item, the reward is worth the time it saved you and nothing more. Levelling a trade for a bound drop is a reasonable choice, but it is a choice about your character rather than about income, and it should not be counted twice when you are weighing which profession pays.

There is one more structural point worth naming before the table. Jewelcrafting does not exist in Forever, so there is no second trade selling a small permanent improvement to gear a character is already wearing. That leaves Enchanting as the only trade in the list whose product attaches to an item the buyer already owns, and it is why Enchanting sits in the per-upgrade column rather than the occasional one: a weapon, a cloak and a chest piece that get kept for a while each pass through an enchanter once, and the dust behind those enchants comes from the same trade disenchanting what nobody kept. In later versions that demand is split; here it is not.

That same test applies to the gathering side, and this is where the clearest new example sits. Beta testers report that several of the added enchanting recipes call for herbs alongside the usual dust and essence, which would mean the same stack of herbs is bought by two different trades rather than one. Alchemy has always been the single buyer for most herbs; a second buyer with a different reason to want them changes how often a gatherer finds a bid waiting. We are treating this as unconfirmed, because it comes from players on a beta that wipes on 4 November and the recipe list is still moving, but it is worth watching before you commit a character to a gathering trade.

The reason it matters is the one this whole page runs on. A material with one customer profession is only as steady as that profession's demand on a given evening; a material two professions consume keeps selling when either of them is active, and the two are not busy at the same times — alchemists restock before a raid night, enchanters work through whatever came out of it afterwards. For a gatherer that smooths the gaps rather than raising any single sale, which is the same property that makes consumables beat durables. If the overlap holds on live realms, herbs on the auction house become a two-sided listing rather than a one-sided one, and that is a better argument for picking Herbalism than any figure quoted from a beta.

Engineering is the trade where the two columns sit inside one profession, and the beta discussion made that visible. After an early 1v1 tournament, a large part of the player commentary settled on the same conclusion: engineering is close to mandatory for anyone who takes PvP seriously. That is a player opinion rather than a tested fact, and we are not going to put a figure on how many duellists actually followed it. But opinions about what is mandatory are what move a market, and the items behind this one are not the goggles. They are the throwables — bombs, sappers, nets, the small devices that leave the bag the moment they are used.

For a seller that puts part of Engineering in the constant column while the rest of it stays durable. A pair of crafted goggles is one sale and then nothing for 10 levels; a stack of grenades is consumed in an afternoon of duels whether the buyer wins or loses, which is the same shape as food and potions rather than armour. It also changes the second slot. Engineering is hungry, so the usual pairing is mining to feed it, and the character who takes both has no room left for enchanting, alchemy or tailoring. Those players stop supplying themselves and start buying, which is worth keeping in mind when you read the auction house for consumables as a single market rather than as separate trades.

That column has a delivery problem the others do not. An enchant is applied to an item the buyer is holding, so it cannot be posted and forgotten the way a stack of food can be. The sale happens in trade chat: you advertise, somebody answers, you meet, and one of you hands over goods before the other has finished their half. Everything that is slow about selling enchants follows from that. You are paid for the time you spend sitting in a capital spamming a channel, not for the time your goods spend listed on the auction house, and an evening with no replies earns nothing at all.

Two further costs come with the channel. The first is the handover itself: either the buyer sends materials and trusts you to return the enchanted item, or you apply first and trust them to pay, and there is no system holding the middle. Sellers usually solve this by stocking the materials themselves and charging a flat figure for the finished enchant, which removes the trust problem but ties up capital in reagents. The second is noise. Trade chat is where advertising bots post, so a genuine offer scrolls away between them, and buyers learn to ignore the channel rather than read it. None of this makes Enchanting a bad trade, but it does mean the income depends on hours spent present and visible, which is a different kind of work from restocking a listing once a day.

The level cap pushes the durable column the same way, from a direction that is easy to miss. Forever stops at 60 and there is no announced content patch that raises it or prints a tier of gear that makes the current one obsolete, so a crafted blue a character buys at 60 is not devalued on a schedule. For the person wearing it that is the good outcome. For the person who made it it means the sale is final: the buyer has no patch-driven reason to return, and the item stays in circulation on that character instead of being replaced and resold.

That stretches the payback on a one-off craft rather than shortening it. A recipe you levelled for pays back only as fast as new buyers arrive, and once the characters who want that piece have one, the line goes quiet with nothing scheduled to restart it. A consumable has no such ceiling, because the same customer empties the stack and comes back regardless of what the gear situation looks like. If you are deciding which trade to carry, treat no patch cycle as a point against durables and read the auction house for finished gear as a market that fills up rather than one that refreshes.

Vendor income is the other half of the picture, and it no longer behaves the same way across trades. Most professions have had their vendor payouts trimmed, so the fallback of crafting something and selling it to an NPC when nobody bids returns less than it used to. Engineering moved the other way on at least one item: the Satchel of Dark Iron Bombs vendors for 2g 66s 2c instead of the 9c it was worth before. That is a bag, which on the auction house is the dullest kind of durable — one sale per character, no repeat — but against an NPC it now pays per craft without needing a buyer at all.

The practical reading is that vendor value has stopped being a flat floor under every trade and has to be checked recipe by recipe. For most professions the NPC price is now a worse safety net than it was, which pushes their crafted goods back onto real demand and makes the consumable test above matter more. For Engineering, a single bag recipe turns materials into a fixed return that does not care how crowded auction house pricing for crafted goods gets on a given evening. We are not treating that as the profession's main case, since one item is one item and the beta list is still moving, but it is the clearest sign that vendor income should be read per recipe rather than assumed.

The table has a third column that most profession lists leave out: trades that consume money rather than produce it. Engineering is the clearest case. Its output is mostly consumable, but the consumer is usually the engineer — bombs, grenades, nets and trinket charges get thrown in battlegrounds and world fights, and every one of them is a stack of ore and powder that left your bags and did not come back as a sale. A player who enjoys the PvP side of the trade can hold a maxed profession, a full recipe book and an empty purse at the same time, because the production line ends at their own action bar.

That is worth naming plainly because it is not a mistake, it is a choice about what the profession is for. Gathering turns time into stock someone else buys; crafting for sale turns materials into a listing; Engineering in a PvP routine turns materials into an advantage that lasts one fight. If you are picking a trade to fund a character, the question is not whether the gadgets are good — several of them are — but whether you intend to sell anything at all. The honest version of the answer is to treat the bomb budget as a spending category and fund it from somewhere else, either a gathering trade on the same character or the steadier consumable lines on the auction house for reagents, rather than expecting the trade to pay for its own habit.

Enchanting also sits apart from every other trade on this table for a reason that has nothing to do with how often the demand repeats: there is no lot to post. The enchant is applied to an item the buyer is holding, so the sale happens as a service between 2 players rather than as a listing somebody browses. A tailor can put a bag up and walk away; an enchanter has to be online, in the same place as the customer, with the materials already in the bags. That is a different kind of work from stocking the auction house for finished goods, and it should be weighed before picking the trade, not after.

What this changes in practice is where the demand is found. An enchanter sells through trade chat and through whoever asks in a dungeon group, which means income depends on being present at the hours when people are upgrading rather than on a stock sitting in the mailbox overnight. The materials behind the enchants are listable and behave like any other consumable input, so the dust and essence side of the trade can be sold the ordinary way while the enchant itself cannot. If you do not want to spend evenings answering shouts, that is a real argument against the trade, regardless of how steady the underlying demand looks.

Taken together, the binding rules are the part of the build that changes the conclusion of this table rather than decorating it. In the current beta build the crafted gear a profession produces is bound to the character who made it, and forum threads from testers describe the same thing from the other side: people asking where to buy a piece and being told the only way to wear it is to level the trade themselves. A trade in that state has no finished product to put on the auction house at all. Whatever it is worth to the person carrying it, as a business it sells nothing.

So the income from a crafting trade, for now, sits in two places only: the consumables it can still list, and the raw materials that go into the gear nobody can sell. Both of those are repeat purchases, and the second one grows every time a player decides to level the profession on an alt instead of buying the armour. That is why the gathering half of the auction house for gathered goods reads as the safer side of the split here, and why a recipe list is a poor way to judge a trade in this build — it describes items that never reach a buyer. If the binding rules change before launch the arithmetic changes with them, but nothing in the current build suggests planning around that.

It is worth being blunt about what a bind on pickup blue actually returns. The payoff is a slot filled on your own character, not gold in your own bags. That is a real return, and for a player who would otherwise have bought something similar it may be the better one, but it is not income and it does not compound: you cannot make the same piece twice and sell the second. A trade whose best recipes are built that way is a self-supply operation with a profession tab attached, and it should be judged next to what you would have spent, not next to what another trade earns on the auction house.

The part that never changes its mind is the gathering half. Ore, herbs, leather and cloth carry no bind of any kind, so everything a gatherer pulls out of a zone can be listed exactly as it came, including the stacks the crafter in your own guild was going to need anyway. That is why the materials line keeps its customers while the finished-goods line loses them to the binding rule: the buyer of a bar does not have to be a stranger who wants your armour, only somebody levelling the same trade you levelled. If you are picking a pairing for income rather than for convenience, weigh the gathering side against gold earned at each level and treat the crafted blue as the part that pays you in gear.

The same thinning applies on the consumable side. Alchemy and Engineering are the main suppliers of the things that get used up in a fight — potions and elixirs on one side, explosives and ammunition on the other — without the later trades that eventually sold alternatives to the same buyers. We are not going to guess how many players pick each of them up on a fresh realm, since nobody has that number yet, but the list of trades competing for a given consumable is shorter than it is in anything released afterwards. For judging recurring income that matters more than the recipe count: fewer suppliers chasing a demand that renews every evening, and the auction house for enchanting materials feeding both halves of it at once.

TradeWhat it mostly sellsRepeat demand
Gathering — Herbalism, Mining, SkinningRaw materials into every other tradeConstant
AlchemyPotions, elixirs, flasksConstant
Cooking, First AidFood and bandagesConstant
EnchantingEnchants, and the dust to make themPer upgrade
Leatherworking, TailoringArmour, bags, armour kitsOccasional
BlacksmithingWeapons, plate, sharpening stonesOccasional, stones aside
EngineeringGadgets, explosives, ammunitionSplit — see below

Why Engineering is the argument

Engineering came up on the official forum this week and the verdict there was unsentimental. Most of the climb consumes components rather than producing anything a stranger wants: you buy bars, you turn them into parts, the parts become the next parts. The saleable end — explosives, gadgets, the famous jumper cables — arrives late.

The exception is ammunition, and it is a real one. A hunter who raids goes through it the way an alchemist goes through herbs. That is the only part of Engineering with a customer who comes back on a timetable.

That assessment is players', not ours and not Blizzard's, and it is about the beta. We are repeating the reasoning because the reasoning is checkable; the conclusion is theirs.

Something in the latest beta reports cuts the other way, and it belongs on the crafting side of the ledger: several trades now have a blue-quality item inside the first 20 levels rather than only at the top of the tree. By the test this page runs that is durable demand — the buyer wears the piece until something better drops and never orders a second one — but it is a single sale at a price no stack of food will match, landing in a window where most characters have no alternative in that slot. A trade with one early blue in its list is not a recurring business with better margins; it is an occasional business with a bigger ticket, and the two are worth judging separately instead of averaging them into one verdict. Whether a given recipe binds on pickup decides whether it counts at all, which is the filter set out further down.

Engineering is the complaint sitting directly opposite that, and it is the same testers making it. Working the trade through its early tiers, they describe components turning into components with nothing at the end a stranger would pay for: the skill-ups are funded out of pocket and there is no item to list while that is happening. An early blue is exactly what the trade does not have. Both observations come from a build that wipes on 4 November, so neither is a rule for the live realms, but the gap between them is the clearest version of the choice this page is about, and funding that stretch with a gathering skill at low levels remains the dull answer either way.

A newer argument on the same forum is about Merchant's Favor and who it actually favours. Testers comparing the two armour trades say the Blacksmithing list contains more pieces carrying defensive stats than the Leatherworking one does, and that the difference is not cosmetic: a tank shopping for a crafted slot filler reads the stat line first and the trade second. If that holds, the same mechanic produces two very different businesses — one with a queue of tanks asking for a specific piece, one selling to whoever happens to want leather.

We are repeating the comparison, not confirming it. The counts come from players reading a beta build that wipes on 4 November, nobody has posted the full list side by side, and we cannot say what Merchant's Favor ends up doing to prices on a live realm. What is worth taking from it is the method rather than the verdict: when two trades sell into the same durable market, the one whose recipes match a role people are short of gets the orders, and the other one waits. That is still occasional income on both sides, which is why it sits below consumables in the table above rather than beside them, and why the auction house for crafted armour rewards patience more than volume.

Enchanting needs a separate note, because the table row above describes its demand and not the way that demand reaches you. An enchant has no carrier item in Forever: there is nothing to craft, hold in a bag and post, so the enchant itself never appears in auction house listings. What does appear there is the material side — dust, essences, shards — which anyone can sell, enchanter or not.

The practical difference is the sales channel. A cook posts stacks and walks away; the listing works while the seller is offline. An enchanter advertises in trade chat from a capital, waits for a whisper, meets the buyer and applies the enchant in person, usually with the buyer supplying the materials and paying for the service on top. The demand is real and it repeats every time somebody replaces a weapon or a chest piece, but it is paid for in attention rather than in posting fees, and it stops the moment you log out. If you are picking a trade around the hours you actually play, that is a bigger factor than the recipe list, and it is the reason Enchanting and Cooking can share a column on repeat demand while behaving nothing alike day to day.

The same split shows up in the oldest piece of profession advice there is: take the armour trade that matches what your class wears, so you can gear yourself. As a way to keep a character equipped that is sound. As a way to earn, it fails the test this page runs twice over. A piece crafted for yourself is not a sale at all, and a piece crafted for somebody else is a single sale to a single character who will not order a second one in that slot. Alchemy sits at the other end: a potion is gone the moment it is drunk, and the buyer who wanted one before a hard pull wants another one before the next.

Beta testers have been arguing the opposite, and the argument is worth repeating rather than dismissing. Their position is that matched-armour crafting carries the early game because the pieces land in slots nothing else fills, which is the same big-ticket logic as the early blues above. The counter from the other side of that thread is the one this page would make: defence potions and health potions move steadily in the beta auction house, evening after evening, while an armour order arrives once and then goes quiet. Neither side has put numbers behind it, and the build they are testing wipes on 4 November, so treat both as descriptions of shape rather than of price. If you are choosing a trade for income instead of for your own gear, the steady half of that disagreement is the one that matches how the auction house handles consumables actually behaves.

There is a way of reading Engineering that the repeat-demand test does not cover, and the same forum thread puts it plainly: the trade is mostly a self-buff. Most of what it produces is used by the engineer who made it rather than sold to anyone — gadgets that only work in the hands of the person who trained the skill, explosives carried rather than listed, utility items that save a trip rather than earn a fee. By the accounting set out above that is gear budget, not profession income. It is the opposite case from Enchanting, where the output has no inventory but always has a second party; Engineering has an inventory that rarely leaves the bag.

The counter-argument in that thread goes further, and it is worth repeating as the opinion it is. Blacksmithing, Leatherworking and Tailoring sell durables, and durables stop mattering for the buyer the moment better pieces start dropping in raids — so a trade that enhances its owner permanently keeps paying when the armour trades have run out of customers. Whether that holds for Forever is not something anyone can state yet: no realm has reached that part of the game, and the beta wipes on 4 November. What the argument does show is that the ammunition line is not the only reason people pick the trade, and that the question it answers is what your own character can do rather than what the auction house will pay for it.

The levelling half of the durable problem has a concrete example in the current build. Testers handing in Caches at the Crossroads report crafting plans among the rewards, which means the recipes are reaching players without any auction house step at all. The pieces those plans produce are the issue: set against quest rewards and dungeon drops from the same level range, they are rarely the better item, and a character who is questing through the zone already has the alternative in hand. A plan that is free to obtain and makes something the buyer can out-roll on the next dungeon run is not the start of a business, it is a skill-up.

That is the durable case at its weakest, and it is worth separating from the early blue noted above. One good piece in a slot with no alternative still finds a buyer; a mid-range chest or pair of boots competing with a quest reward usually does not, because the competition costs the buyer nothing but time they were spending anyway. Neither observation is settled — the handins are beta behaviour and the rewards can change before launch — but if you are choosing a trade for income rather than for your own equipment, the levelling gear tier is the part least likely to pay for itself, and making gold at each level tends to come from what gets consumed instead.

There is a second risk on the durable side that the repeat-demand test does not price in, and the beta has just supplied an example of it. Wolfshead Helm is the Leatherworking piece that carried that trade in older versions of the game: one recipe, one slot, a buyer base that treated it as the obvious choice and kept ordering it long after the rest of the crafted list had been outgrown. Testers report that the helm has been changed in Forever and that it no longer does what it used to do. We are not quoting the new behaviour, because the values in a beta build are not values on a live realm and this one wipes on 4 November. The point is structural rather than numeric: a single patch note can take a craft that supported a whole profession's income and leave the crafter with a recipe nobody asks for. A trade whose case rests on one famous item is one line of a changelog away from having no case at all.

The same reports describe dungeon quest rewards being trimmed, and that cuts in the opposite direction for anyone selling durables. Every quest that hands out a usable blue for a slot is a sale that never happens, because the buyer already has the piece and did not pay a crafter for it. Weaken those rewards and the slot comes back onto the market: the character still needs something in it, and the remaining options are a drop they have not seen yet or a listing. Neither change is a rule for the live realms, and neither is worth planning a profession around on its own. Taken together they describe how this part of the economy actually moves — demand for a craft is not a property of the craft, it is whatever the rest of the game has stopped giving away for free, which is the half of the auction house in Forever that no recipe list will show you.

The sharper version of the complaint on the beta is comparative rather than absolute. Testers working several trades side by side say the crafting skills around Engineering were given something to sell — early blue recipes, additions to the recipe lists — while Engineering was left roughly as it was, so the gap between it and the trades it competes with for a slot widened without Engineering itself getting worse. That is a claim about one build, and that build wipes on 4 November, so it may not describe the live realms at all. It is worth recording because it is the reason the trade keeps coming up in these threads: not that the components chain changed, but that everything next to it moved.

The same testers single out the camp side of it. Where the camp station bonuses for other trades produce material outright — herbs grown, a vein spawned, reagents fermented or tanned — the Engineering bonus is reported as a vendor that stocks reagents, which is gold going out for parts rather than parts arriving for free. For the test this page runs that is a convenience, not a supply line: it shortens the trip to buy components, and the components still come out of your pocket. What the vendor stocks, and whether that is the whole of the bonus at higher skill, is not something we can state from a beta sample.

Alchemy is the same test run in the other direction, and it is worth spelling out because the table above puts it in the constant column. Potions are drunk and bought again, so the trade reads as recurring by default. In Forever, health potions sit with First Aid instead, which takes the single most repeated purchase in the game out of the alchemist's hands and hands it to a secondary skill every character can train for themselves.

What is left at the low end is thinner than it looks. By level 20 the notable thing an alchemist makes for other people is a helmet — worn once, replaced the moment something better drops, and never bought twice by the same customer. That is durable demand inside a trade with a consumable reputation, which is exactly the mismatch this page is about. It does not make Alchemy a dead end: elixirs and, later, flasks still empty and still get reordered. It moves the payday further up the level range, and it means the early skill-ups are funded by you rather than by a buyer. How the rest of the reagent chain behaves at higher skill is not something we can state yet, since the beta is the only sample and it wipes on 4 November. Before committing a slot, check the specifics against what Forever changed about professions.

The forum reading of the early range is sharper than that, and it cuts across the table above. Testers levelling more than one trade at the same time say Alchemy holds the opening stretch of the market more or less alone, with Cerulean Dye named as the item doing it: a cheap, repeatable output that other crafters need and that an alchemist can list from low skill, before anything else in the crafting column has produced something a stranger will pay for. The complaint is not that the dye is strong, it is that nothing sits next to it.

Enchanting and Tailoring are the two they name on the other side of that gap. An enchanter at low skill is disenchanting more than he is selling, and the service half needs a customer with an item worth improving, which most characters do not have that early; a tailor is buying cloth before he can sell anything made out of it, so the first tiers run at a loss by design. Both reach a list worth posting somewhere in the middle of the level range rather than at the start of it. This is testers' reporting on a build that wipes on 4 November, not a measurement, but it fits the accounting used here: the difference is in when the first sale arrives rather than in what the trade eventually earns, and covering that stretch with a gathering skill at low levels is the same dull answer as everywhere else on this page.

One Forever rule change pushes the other way, and it belongs in the same argument. Druids in Forever can use consumables while shapeshifted, so a class that previously had to leave form, drink and shift back becomes a routine buyer of elixirs and potions mid-fight. That does not change what an alchemist can craft. It changes how many characters treat a consumable as standard kit rather than as something they get round to between pulls, and feral and bear specs are the obvious case, since they spend most of a fight in form.

Read it as a widening of the customer list, not as a figure. We cannot say which consumables the rule covers in full, or how it behaves further up the level range, because the beta is the only sample and it wipes on 4 November. What is safe to plan around is the direction: one more class sitting in the recurring column, buying items that empty and get reordered, which is the side of the test this page is built on. The durable end of Alchemy is unchanged by it, and so is the advice to fund the early skill-ups with a gathering skill at low levels rather than out of pocket.

Blacksmithing has a gap of its own on the beta, and it is a strange one: players working through the trainer lists and the recipe drops they have seen so far report no shield recipe at all. Not a weak one, not a late one — none found. We cannot confirm that the category is absent from the game, only that nobody testing it has turned one up yet, and the beta is a partial sample by definition. If it holds, it removes one of the few genuinely repeating orders a smith gets, because shields are the slot tanks replace on a schedule and the one piece they will pay to fix early rather than wait out.

It matters for the comparison above rather than for Blacksmithing alone. A smith without shields is selling plate, weapons and sharpening stones, and only the stones sit in the consumable column. That pushes the trade further toward occasional demand than the table suggests, and it is worth watching after the wipe before you commit a slot to it. If shield recipes turn up later — from a blueprint rather than a trainer, which would fit how the new Forever recipe sources work — the answer changes, and so does the reason to pick the trade.

The stone line itself looks wider on the beta, and it pulls the other way. Testers report that sharpening stones and weightstones in Forever sit alongside poisons rather than replacing them on the weapon, so a rogue can run a poison and a stone at the same time. That puts a class with no previous reason to buy either into the stone market, next to the warriors and paladins who were already there.

By the test this page runs, that is the useful kind of change: it widens the buyer list for the one output a smith has that empties and gets reordered, rather than adding another item somebody wears until it is replaced. Stones are cheap, bought in stacks, and used up by anybody swinging a weapon for an evening, so the same customer returns without having improved at anything in between. It does not close the shield gap — plate and weapons are still single sales per character — but it means the consumable column of Blacksmithing is broader than the trade's reputation suggests. Whether every stone tier behaves this way, and which poisons it covers, is not something we can state: the beta is the only sample and it wipes on 4 November, so it is worth rechecking against what Forever changed about professions once realms are live.

There is a second filter underneath the consumable test, and it runs before it: can the finished item leave your hands at all. A recipe that produces a bound item pays you in skill points and in gear for your own character, and in nothing else — there is no second party to the transaction, so it cannot appear in any income plan no matter how good the item is. Every crafting trade has some of these, and the binding is a property of the recipe rather than of the profession, so two recipes off the same trainer can behave completely differently. Which Forever recipes bind and which do not is not something we can list; the beta has not been mapped that far, and it wipes before anyone finishes.

Beta testers have put a concrete case under that filter. A share of the blue-quality recipes they have found so far come from dungeon drops and from supply crates rather than from a trainer, and the items those recipes produce are reported as bound on pickup. The pattern they describe for the run up to level 20 is consistent: the recipe is a real upgrade, the crafted piece goes straight into your own slots, and there is nothing at the end of it to list. Skill points and a better character, no buyer. This is testers' reporting on a build that wipes, not a rule we can state for the live game, and we do not know how far up the level range it holds.

If it does hold, it sharpens the case for treating the early levels as gathering time. A crafting trade whose first tier is bound output funds itself out of your pocket twice over — once for the materials, once for the skill-ups — while a gathering skill produces something saleable from the first node. The crates and drops also mean the recipe list you end up with is partly a function of where you run rather than what you train, which is one more reason to weigh what a dungeon run is worth alongside the trainer's page before committing a profession slot.

It is worth stating the accounting plainly, because a good craft is easy to mistake for a good trade. A blue-quality piece that binds the moment it is made is gear budget, not profession income. It replaces something you would otherwise have looted or paid for, which is real value and worth the materials, but it never reaches a second party, so it cannot appear in the column this page is testing. The repeat-demand table above assumes the output can be listed at all; for a bound recipe the question of whether anybody buys it twice never comes up, because nobody buys it once.

The working version of that check is dull. Before spending materials on a recipe that came out of a drop or a crate rather than off a trainer, look at how the finished item binds, and decide which budget the run is coming from: the one that dresses your character or the one that is supposed to pay for your training. Both are legitimate, they just should not be added together. Mixing them is how a crafting trade ends up looking profitable on paper while the gold in the bag goes down every week. Which Forever recipes bind and which do not is not something we can list, since how Forever recipes are acquired varies by trade and the beta wipes on 4 November, so the binding is a property to read on the recipe itself rather than to assume from the profession.

Enchanting is the clean counter-example, and it is why the table puts it above the armour trades despite selling nothing that sits in a bag. The enchant is applied to the customer's own item, so binding never enters the exchange: they keep the gear, you keep the fee, and the work is repeated the next time they replace that slot. A tank who replaces a weapon at 40 and again at 50 is two sales, not one, and the disenchanting side keeps producing dust that other people need for their own orders. That is the same split behind making gold at each level — a service leaves you with a customer, a durable leaves you with an inventory problem.

The same property has a cost attached, and it is the one thing that separates Enchanting from every other trade in the table: the finished product cannot be listed. An enchant is applied to an item that belongs to someone else, so there is no object to post and no auction channel for the output at all. What an enchanter can actually put on the auction house is the reagent line — dust, essences, shards off the disenchanting side — and those sell to other crafters rather than to the person who wanted the enchant.

That splits the trade into two businesses with different working habits. The reagent half behaves like any other material: post it, walk away, let it sell while you are offline. The service half needs the customer and the enchanter in the same place at the same time, which means trade chat, a guild, or standing somewhere busy, and it means income that stops the moment you stop advertising. Neither half is worse, but only one of them earns while you are logged out, and that is worth knowing before a profession slot goes to a trade with no shopfront. It is also why an enchanter's income reads differently from the posted prices that shape why gold costs what it costs, since half the work never passes through the auction house to be measured.

The reason there is no shopfront is that an enchant has no object to put in one. Later versions of the game added vellums, a blank item the work can be applied to and then listed like any other craft; Forever has nothing of the kind, so the enchant exists only at the moment it is cast on gear that already belongs to somebody else. That takes the auction house out of the service half entirely. What is left is trade chat, a guild, a whisper from someone who remembers you from last week, and a customer standing next to you with the item equipped.

It is worth holding that in mind when comparing the trade's take against a craft that posts its output. A tailor or a smith sells into a market that works while he is logged out and that prices itself in public, so his income is visible and largely passive once the listings are up. An enchanter sells into a conversation he has to start, and stops earning the moment he stops starting them. Two players putting in the same hours can end up a long way apart on income with the same skill level, and the difference sits in the selling rather than in the recipe list. If a slot is going to Enchanting, count the advertising as part of the work, and treat the reagent line as the half that behaves like every other trade on what Forever changed about professions.

What Forever changed that matters here

  • Everyone gathers from level 3. The gathering skills stop being a choice and become a floor, which means more raw supply and thinner margins on it.
  • Camp stations make materials. A greenhouse grows herbs, a rock garden spawns a vein, a fermenter and a tanning rack produce reagents outright. Production no longer requires travel.
  • 600-plus new recipes. The obvious ones will be crowded from week 1; the scarce ones come from blueprints that have to be found.
  • Legacy points come from crafting, not gathering. Which means a pure gatherer never earns the points that buy the perk raising gathering yield.

So what do you pick

If you want income in the first month: a gathering skill you actually run, plus a consumable trade to convert it. If you want the best character at 60: whatever your class needs, and buy the consumables from someone who chose the first path.

One caveat on the first path, and it is about timing rather than about the trade itself. In the opening phase everyone is levelling at once, which means everyone is pulling the same low-level herbs, ore and hides out of the same zones on the same day. Supply of the cheapest materials arrives before the demand for them does: the buyers who need those stacks in volume are the players pushing a crafting skill through its early tiers, and they get there later than the gatherers do. Raw low-level material sold into that window is worth the least it will ever be worth, and no amount of route efficiency changes that.

It is not an argument against taking a gathering skill, only against expecting one to carry you in the first weeks. The usable responses are the dull ones: convert the material yourself instead of listing it raw, keep what stores well until the crafting crowd catches up, or treat the early gathering as skill-ups rather than as income. How long the overlap lasts is not something we can put a number on — no Forever realm has run through a launch yet, and the beta wipes on 4 November — but the direction is predictable, and it is the same pressure that shapes the early Forever economy more generally.

There is a ceiling on gathering income that has nothing to do with prices, and it is carrying space. What you earn per hour is bounded by how much fits in the bags before the run ends and the trip back begins, and the opening allowance is 4 slots, 6 once the quest reward is in. Skinning presses on that harder than the other gathering skills, because hides and scraps come in several types that stack separately, so a route fills slots faster than the stack sizes suggest and the return trips start eating the hour. A skinner who has to walk back twice as often is not running a worse route than the herbalist next to him; he is running the same route with less room.

The other cost this comparison leaves out is kit bought for the profession rather than for the character. Gathering and the trades hanging off it have equipment that does nothing in a fight and a good deal on a route: a cloak carrying a mining bonus, leg armour that helps at the forge, a glove enchant tied to the gathering skill itself. Each of those is a one-off cost that then applies to every node, every hide and every smelt for as long as the piece stays equipped, which puts it in a different column from the materials you burn on skill-ups and never get back.

The arithmetic is dull and it is the same each time: a fixed outlay against a small gain repeated per run, so the question is how many runs you actually expect. A character who works routes most evenings pays the kit back and then keeps the difference; one who gathers when it is convenient is better off spending the same gold on bag space, which raises the ceiling described above instead. Two details are worth checking before buying. A glove enchant occupies the slot a combat enchant would have taken, so the cost includes what you give up there, and it has to be reapplied if the gloves are replaced. Which of these items exist in Forever, what they grant and where they come from is not something we can state — the beta is the only sample and it wipes on 4 November — so weigh them against what Forever changed about professions once realms are live rather than against the old lists.

There is a second thing about Skinning that the table above does not capture, and it works in the opposite direction from the bag problem. A herb node or an ore vein is taken once: whoever reaches it first gets it, and the node is gone for everyone else. A corpse is not scarce in the same way. In a group or a raid, every skinner present can take a hide off the same body, so one kill can supply several gatherers at once, while the same kill run solo supplies one and then the corpse is gone. The supply of leather per monster killed therefore scales with how many skinners are standing around it, and nothing comparable happens to herbs or ore.

The practical effect is on price rather than on effort. Leather arrives on the auction house from group content in volumes that a single gatherer's route does not explain, which means a skinner is competing with supply he did not create and cannot see. That is an argument for converting hides yourself rather than listing them raw, and it is one more reason to weigh what a dungeon run is worth as a source of material and not only of loot. How the mechanic behaves across the Forever level range, and whether every hide type follows it, is not something we can state from the beta.

That is also where the bag trades sit in the test this page runs. A bag is a durable — bought once, kept for a long time — but the buyer list is every character rather than one class, and with gathering open to every character from level 3 the buyer list is the whole realm. On the auction house the crowded niches at launch are the obvious ones, the recipes every new crafter trains first; upgrades built out of leather are the part of the bag market nobody is competing over yet, with tailoring sitting beside it on the cloth side. What the Forever bag recipes are, what they need and where they come from is not something we can state — the beta is the only sample and it wipes on 4 November — but the demand side is readable in advance: a realm of gatherers working out of 4 to 6 slots.

Cloth deserves a line of its own here, because it behaves unlike every other material on the list. Herbs, ore and hides each have a gathering skill attached to them, so the person who collects them chose to and can train the yield up. Cloth has none. It drops from humanoids, which means a tailor has two supply routes and both cost something the other trades do not pay: buy it off the auction house at whatever the levelling crowd lists it for, or go and kill humanoids for it as a deliberate trip rather than as something picked up along a route that was already being run.

That is the gathering-versus-crafting split from the top of the page showing up inside a single trade. A skinner turns time into saleable material directly; a tailor turns gold into material first and only then into an item, so the margin depends on someone else's listing price rather than on his own route. It also means cloth supply is wide and uncoordinated, since most of it comes from players who were not collecting it and who sell it to clear bag slots. Which humanoids in Forever drop which cloth, and at what rate, is not something we can state — the beta is the only sample and it wipes on 4 November — but the structure holds regardless, and it is one more reason to read why gold costs what it costs before picking the trade over one that gathers its own inputs.

The mage case is the clearest illustration of recurring demand sitting behind a durable item. Beta testers report that a share of the caster pieces a mage would previously have waited on as drops now sits in Tailoring instead, and that part of that list is not made out of the bag at all: it is worked at a Spinning Wheel, with the recipe coming from a blueprint testers describe as rare rather than from a trainer. Each finished robe is still a one-off for the buyer, who wears it until something better turns up and never orders a second. The cloth behind it is not a one-off for anybody.

That is the split worth keeping separate when judging the trade. The tailor sells a durable and gets a single sale per customer; the material feeding it empties every time and has to be bought or farmed again, so the steady business is in the cloth rather than in the garment, and it belongs to whoever supplies it. For a tailor holding the blueprint, the recurring cost is the input, which is why the two cloth routes above decide the margin. Which mage pieces moved, what the Spinning Wheel needs and how the blueprint is found are not things we can state — the beta is the only sample and it wipes on 4 November — but the direction fits the rest of the Forever profession changes: output gated behind a found recipe, demand concentrated in the material.

The same skew shows up from the buyer's side, and testers describe it as a gap rather than a preference. Below level 30 the Leatherworking list is built around agility and stamina pieces, and the caster slots it covers are thin to absent, so the leather and mail wearers who cast — balance druids, shamans, a hunter after an intellect piece — end up on the auction house buying cloth armour made by somebody else's trade. Their own profession does not dress them in the stat they want, and the trade that does is the one with no gathering skill behind it.

Read that as demand rather than as a recipe complaint. Each of those characters is a cloth buyer who was not counted as one, which widens the customer list for the material rather than for the finished garment, and it is the recurring-input case from the paragraph above with more names attached to it. How far up the level range the gap runs, and which slots exactly it leaves open, is not something we can state: the beta is the only sample and it wipes on 4 November. Before a slot goes to leather, it is worth checking the early list against what Forever changed about professions and against the classes actually on your realm.

The dishonest answer is a ranked list with gold-per-hour figures. Nobody has those for Forever, the beta wipes on November 4, and the prices that will matter do not exist yet. What survives the wipe is the reasoning.

Questions

Which profession makes the most gold in WoW Forever?

Nobody can answer that yet with numbers, and anyone who gives you a figure this month is guessing. What can be answered is the shape of the question: professions that sell something people buy again beat professions that sell something people buy once.

On that test the gathering skills and the consumable trades start ahead. A hunter buys ammunition every week. A player buys a pair of goggles once.

Is Engineering worth levelling for money?

Players on the official forum are blunt about it: most of the way up, Engineering consumes components rather than producing goods anyone wants. It is a slog, and the slog is paid for in materials you bought.

That does not make it worthless — it makes it a late payoff rather than an early income. If you need gold in the first month, it is the wrong first profession.

Do I need a crafting profession at all?

For money, no. For Legacy points, yes. Points come from non-gathering professions at 150, 225 and 300, and the perk that raises gathering yield is bought with them.

Sources