Short answer: web3 marketing is the work of getting people to find, join and use a crypto product, and it runs across the same channels everyone lists — Discord, Telegram, X, creators, quests, airdrops, partnerships, events, paid ads. What almost nobody attaches to those channels is a number. We went through Zealy's 118 published case studies looking for a cost and an outcome sitting side by side, and found both for two channels: performance-based creator posts, and an exchange trading competition. This post publishes those figures, says plainly which channels we cannot price, and explains why that gap exists everywhere, including here.
Disclosure: Zealy publishes this page and Zealy sells one of the things discussed on it — a quest and engagement platform that projects use for several of the channels below. Every number in this post comes from Zealy's own marketing case studies. That means it is first-party data: a claim by Zealy about a Zealy customer, or a customer's self-report published by Zealy. None of it has been independently audited, and you should read it with the scepticism you would apply to any vendor quoting its own results.
Last verified: 17 August 2026.
What is web3 marketing?
Web3 marketing is the work of getting people to find, join and use a crypto product: a token, a chain, a wallet, a game, an exchange listing. The channel list is familiar: Discord, X, Telegram, creators, quests, airdrops, paid ads. What is missing almost everywhere, including here, is what any of it costs.
The difference from ordinary consumer marketing is not the channels. It is that in crypto, the product often is the incentive. You can pay for attention with a token, an allowlist spot, XP, a raffle entry or cash, and the people you attract will respond differently to each. That is a genuine structural difference, and it is why a web3 team ends up running a Discord, an X account, a quest board and a rewards budget at the same time, in a way a normal SaaS company does not.
It is also why the honest version of this topic is uncomfortable. When your reach is bought with tokens, "what did it cost" and "what did it produce" both become measurable in principle — and stay unmeasured in practice, because publishing either number is a competitive disclosure nobody has to make.
We ran our own search for web3 marketing on 17 August 2026 to see what a reader currently gets. The results were agency landing pages and vendor blogs: singular.net, formo.so, coinbound.io, channelsale.com, eakdigital.com, luvkaizen.com, cryptotrafficmarket.com, plus a Shopify blog post and a Substack essay. Every one of them names the channels. Of the pages we opened, not one attached a cost to any channel it recommended. That is an observation about the pages we read on one day, not a claim about the whole internet — but it held for every page we checked.
So this post does the thing the category does not do, as far as our own evidence allows. Which is less far than we would like.
Which channels did we check, and what does our own evidence show?
We went looking for cost and outcome across the channels Zealy's own customers run: Discord, Telegram, X, creator posts, quests, airdrops, ambassador programs, referrals, partnerships, events, paid ads, email, content, exchange campaigns. This is our list, not a canonical one. Two of them have both recorded.
Here is the state of the evidence, honestly grouped. The point of the table is the right-hand column.
| Channel | Outcome recorded in Zealy's case studies | Spend recorded |
|---|---|---|
| Performance-based creator posts | Yes | Yes |
| Exchange trading competition (MEXC) | Yes | Yes |
| Quests and campaigns | Yes | No |
| Partnerships between communities | Yes | No |
| Discord | Yes | No |
| X organic | Yes | No |
| Events and in-person activations | Yes | No |
| Paid ads | No | No |
| Email and newsletter | No | No |
| Content and SEO | No | No |
| Referrals | No | No |
| Airdrops | No | No |
| Ambassador programs | No | No |
| Telegram | No | No |
Two rows carry a price. Five carry a result with no price attached. Seven carry nothing at all. That is the actual shape of first-party evidence at a company whose entire product is campaign participation, and it is worth sitting with before you trust anyone else's channel breakdown.
One more thing the table hides, and it matters for every figure that follows: no Zealy case study records what the Zealy subscription cost, and none records staff time. Every "cost" in this post is reward budget only. Platform fees and labour are invisible in this data, so any cost-per-outcome you derive from it understates the real number. We have written separately about both of those hidden costs — what community tooling actually costs across the category and how the review and moderation labour behaves — because they are usually the larger half of the bill.
What a performance-based creator campaign cost
Two Zealy campaigns published a price per post and a reach figure. Zealy's own campaign paid $15 USDT per approved post, spending $525 across 35 posts in five days. Raid Arena paid $100 per post, spending $500 across five posts in under 24 hours. Both are Zealy's first-party case-study data.
This is the clearest priced channel we have, because both campaigns used the same mechanism: a creator posts, submits proof, gets reviewed, and is paid a fixed amount per approved post. The unit price is the campaign design, not an estimate after the fact.
| Zealy's own KOL campaign | Raid Arena | |
|---|---|---|
| Price per approved post | $15 USDT | $100 USDT |
| Total spend | $525 (35 posts × $15) | $500 (5 posts × $100) |
| Posts | 35 | 5 |
| Reach claimed | 100,000+ views | 100,000+ views |
| Window | 5 days | under 24 hours |
| Creator qualification | Minimum 1,000 followers + X Premium | 6.5k+ followers + 1,500+ average impressions on your last 5 posts + X Premium |
Source: Zealy's own case studies, the $15-per-post campaign, published October 2025, and the Raid Arena campaign, published November 2025. First-party data, not independently audited.
The Raid Arena case study prints one cost-per-outcome figure directly: $5 per 1,000 views. That is the one cost-per-outcome figure we quote from a campaign narrative, this post quotes, and it is quotable because the campaign published its spend, its reach and its rate together, and those three agree with each other.
The interesting part of this comparison is not which campaign was cheaper. It is the qualification gate. Raid Arena paid $100 a post and required a creator to have 6.5k+ followers and a recent average of 1,500+ impressions; the $15 campaign asked for 1,000 followers and X Premium. The higher price bought a much narrower creator pool and a much faster result — 100,000+ claimed views in under a day rather than five. If you take one operating lesson from this section, take that one: in performance-based creator work, the price you set is a filter on who applies, and you are choosing your applicant pool more than you are choosing your cost.
Two limits on both numbers, stated because they are the parts a case study will not volunteer. "Views" is a creator-reported and platform-reported metric with no audit behind it, and "100,000+" is a floor, not a measurement. And neither campaign records what the reach converted into — no signups, no wallets, no purchases. These are reach numbers, and reach is not a result.
If X is the channel you are actually trying to make work, the mechanics of getting attention there are a different problem from paying for it, and we covered them in how to get X followers for a crypto project.
What a 28-day exchange trading competition cost
$WALLET allocated a $9,800 USDT reward pool for a 28-day trading competition on MEXC, hosted by Zealy, and the case study reports $9,500,000 in total trading volume and around 200 daily active traders. Trading volume is gross turnover, not revenue, return or profit.
Read those three figures carefully, because each carries a caveat that changes what it means.
- The case study uses three different words for the $9,800. It is an "allocated" reward pool in the body, "Total Rewards Distributed" in the results table, and "spent" in the page description. Allocated, distributed and spent are three different facts, and a page that uses all three interchangeably has not told you which one it measured. We quote the figure and flag the ambiguity rather than picking the flattering reading.
- $9,500,000 is gross trading volume on MEXC. It is the sum of trades on a pair, so the same dollar traded repeatedly counts repeatedly. It is not money the project received, kept or earned. Anyone who converts a volume figure into a "return" is comparing two unlike things.
- "~200 daily active traders" has no stated denominator. The case study does not say whether that is an average across the 28 days, a peak day, or unique traders over the whole period. Those are three very different populations and we cannot tell you which one it is.
Source: Zealy's $WALLET MEXC case study, published November 2025. First-party data, not independently audited.
What the figure is genuinely useful for is sizing. If you are listing a token and considering a trading competition, this tells you that a four-figure reward pool over four weeks is a real campaign shape that a real project ran, and roughly what order of participation came back. It does not tell you the campaign caused the volume, because there is no control and no counterfactual, and the case study does not offer one.
This is also the channel where the reward-budget-only caveat bites hardest. Running a competition means someone builds the quest, watches for wash trading, handles disputes and pays out. None of that labour is in the $9,800. If you are designing the payout side of something like this, how to distribute tokens to active community members covers who actually sends what, and to whom.
The channels where we have an outcome but no cost
Five of the channels we checked appear in Zealy's case studies with a result attached and no spend recorded: Discord, X organic, quests, events, and partnerships. One partnership quest between Kyoto Blockchain and Qorpo World reports 10,000 members in under a week, rewarded with 150 XP and no cash.
The partnership example is the most useful of the five, because it is the closest thing we have to a zero-cash channel with a recorded outcome. Both communities agreed on the same reward — 150 XP, no money — for joining the other's community, and Zealy's partnership quest checks membership automatically, so there was no review labour behind it either.
Read the 10,000 precisely. The case study describes those members as transferred between the two communities, which is a different claim from 10,000 new people. The same person can appear on both sides, so "grew by 10,000" overstates it. Treat it as evidence that a cross-community swap moves a large number of memberships quickly, not as an acquisition figure — and note the date, because this one is older than the rest.
Source: Zealy's partnership quest case study, published April 2024. First-party data, not independently audited.
For the other four, we have outcomes and no budgets. Quests, Discord activity, organic X and events all appear repeatedly across the corpus with member counts, claim counts and engagement figures — and no spend line anywhere near them. Some of that is because the spend genuinely was zero cash and all staff time, which is the least visible cost of the lot. Some of it is because nobody wrote it down.
Where those channels are concerned, we would rather point you at mechanics than at numbers we do not have. Discord engagement and growth tools covers what a server actually needs before you spend anything on it, and if quests are the channel you are evaluating, what Zealy is explains the mechanism, Zealy vs Galxe and Zealy vs Layer3 compare it against the two obvious alternatives, and our survey of gamification platforms for communities looks at whether the mechanics work at all.
The channels where we have nothing
Seven of the channels we checked have no first-party cost or outcome evidence in Zealy's 118 case studies: paid ads, email and newsletter, content and SEO, referrals, airdrops, ambassador programs, and Telegram. That is a statement about our own corpus, not about whether those channels work.
We also searched the case-study, academy and insights folders for the vocabulary of unit economics — cost per acquisition, cost per signup, lifetime value, CAC, LTV — and none of those terms appears anywhere with a figure attached. What the corpus does contain is cost per unit of output, in three campaigns: $5 per 1,000 views, $10 per approved creator, and about $1.75 per daily active trader per day. Only the first of those appears elsewhere in this post; we are printing the other two here because leaving them out would have made the absence look wider than it is. What is missing is the thing a marketer actually budgets against: a cost per member who arrived and stayed, and any lifetime value at all.
That absence is worth naming for two of these channels in particular, because they are the ones people most often assume a platform like ours has data on:
Airdrops. We have published guidance on the mechanics — how to run an airdrop that doesn't get farmed is the practical version. What we do not have is a case study saying what an airdrop cost and what it retained. There is a structural reason as well as an editorial one: Zealy computes no retention metric at all. We searched its source for retention, churn, D7, D30 and the daily and monthly active-user measures. Its analytics interface exposes fifteen methods: two running totals that take no dates at all, five ranked lists, six series bucketed by interval, and a member count as of one date. Not one of them joins a member's earlier activity to their later activity, which is the operation retention is made of. So even where a project ran an airdrop campaign through Zealy, the platform was never measuring the thing the airdrop argument is about.
Ambassador programs. Same shape. How to run a crypto ambassador program describes the design, the incentives and the failure modes. No case study attaches a budget to one, and none reports what an ambassador cohort produced.
For the remaining five — paid ads, email, content and SEO, referrals, Telegram — we simply are not the right source, and it would be worse to invent a benchmark than to tell you that. The 2023 community management modules cover some of this ground from the operations side, including content and social and building an ambassador squad; read them for the method, not for figures, since they are three years old and their numbers are not current.
Why web3 marketing costs so rarely get published, Zealy included
Cost data makes a vendor comparable, dates badly, and exposes failures. Zealy's own corpus shows the pattern: 118 case studies, five that print a spend at all, all five inside one six-week window between 2 October and 11 November 2025, and not one that reports a campaign performing below expectation. Survivorship is total.
Take those three reasons in order, because they are not equally excusable.
Comparability. A published cost per outcome is a stick every competitor and every prospect gets to measure you with, forever, out of context. A vendor that publishes $5 per 1,000 views has handed everyone a benchmark it must keep beating. A vendor that publishes nothing keeps the conversation on stories. The incentive points one way and everybody follows it.
Staleness. Reward markets move. What a creator with 6.5k followers charged in November 2025 is not what they charge now, and a figure that was accurate on publication becomes a misleading benchmark within a year. This is a real problem, not just an excuse — which is why this post carries a verification date and the campaign dates sit next to every number.
Survivorship. This is the one with no defence. Five case studies with a spend, and every single one is a success story. No campaign that missed. No retention decay. No "we spent this and it did not work". We looked for a failure across all 118 and did not find one. A body of evidence with no failures in it is not a body of evidence about what works; it is a body of evidence about what got written up. Read every vendor's case study library that way, ours very much included.
The fix is not for us to stop publishing case studies. It is for the numbers to come with their denominators, their timeframes and their conditions attached, and for at least some of them to be about campaigns that did not go well. We are not there yet. This post is what our archive supports today, and the gap between that and a real cost model is the honest subject of this article.
How to budget when nobody will show you numbers
Price the unit you can verify, not the channel. Every figure in this post that survives scrutiny is a unit price with a completion condition attached: $15 per approved post, $100 per approved post, and a reward pool allocated up front to a fixed 28-day competition. Budget the same way and your worst case is bounded.
That is the practical takeaway, and it generalises past the two channels we can price. A budget built as "$8,000 for KOLs this quarter" has no floor on waste. A budget built as "$100 per post from creators above 6.5k followers with proof of engagement, up to 50 posts" cannot cost more than $5,000 and cannot pay for a post that was never made. The second is not a cleverer forecast; it is a different contract.
Five things to do with that, in order:
- Define the unit before the budget. An approved post, a completed quest, a verified trade, a member who joined and claimed something. If you cannot name the unit, you are buying exposure, and exposure is exactly the thing that generates the numbers nobody can check.
- Attach the payment to a condition you can verify without trusting the counterparty. The two priced campaigns above both worked because payment followed proof, and proof was reviewed before money moved. That single design choice is why they have numbers at all.
- Set the qualification gate deliberately, because it prices itself. A higher per-unit price with a tighter creator gate and a lower one with a loose gate are two different campaigns, not two budgets for the same campaign.
- Add the two costs no case study shows you. Platform subscription and human hours. In our experience the second is larger than the first for anything involving review, and neither appears in a single figure in this post.
- Write down what you expected before you start. Not for a report. So that when it comes in at half, you have a record of a campaign that underperformed — which is the exact data the entire category, us included, is failing to publish.
The last one is the whole argument compressed. The reason this post can price two of the channels we checked and none of the others is not that the rest are unmeasurable. It is that nobody, including us, was writing the numbers down at the time. You can fix that for your own team this quarter, and you will end up with better cost data on your own channels than any agency page or vendor blog — this one included — will ever hand you.