What Does Cost per User Acquisition Really Look Like in Crypto in 2026?

• Fintech app installs ran between $2.50 and $6.00 in 2024 on Business of Apps’ user acquisition cost research, and the install line is the cheapest number in any crypto acquisition report.
• eToro spent $208.7 million on selling and marketing and grew funded accounts from 3.48 to 3.81 million in its FY2025 results, released February 17, 2026, roughly $632 per net new funded account, blended rather than a rate card.
• First Page Sage puts consumer cryptocurrency customer acquisition cost at $188 in a report updated December 23, 2025, drawn from its own client book with no stated treatment of token incentives.
• A sign-up bonus, fee rebate, or airdrop tied to a deposit is a financial promotion under the UK regime and a marketing communication under MiCA Article 66(2), and the review each triggers is an acquisition cost line.
• Twelve-month wallet retention ran from 26.2 percent on Ethereum to 7.9 percent on Solana in CoinGecko’s analysis updated June 24, 2026, so a 3:1 lifetime value to CAC ratio borrowed from software will not survive a market cycle.

• What Does Cost per User Acquisition Actually Mean in 2026?
• Why Does User Acquisition Cost More in Crypto Than in Fintech or SaaS?
• How Do You Calculate Cost per User Acquisition in Crypto?
• What Are the Crypto CAC Benchmarks by Channel in 2026?
• How Does CAC Change Across the Funnel, From Install to First Trade?
• What Is a Good CAC for an Exchange or Trading App?
• How Does Crypto CAC Compare With Fintech and SaaS?
• The Crypto CAC Playbook
• What Are the Challenges and Limitations of CAC Benchmarks?
• What Are the Regulatory Considerations for Incentive-Funded Acquisition?
• How Surgence Labs Does This
• Conclusion
• FAQs

Cost per user acquisition in crypto is the fully loaded cost, including token incentives and compliance review, of bringing one user to a funded account that is still active after 30 days. Three terms in that sentence carry the argument.

Fully loaded means the numerator holds media alongside incentives at grant-day value, referral payouts, compliance review hours, and fraud write-offs. Funded means a deposit above whatever minimum triggers your bonus, not a registration and not an identity check, which the industry calls KYC. Still active after 30 days means trading at day 30, the first point at which you can tell a trader from a bonus farmer. Call that retained.

Software calls this cost per customer acquisition, retail calls it consumer acquisition cost, and in both a customer is whoever paid you. Crypto has no single moment of payment, so it has no single customer. Declare the denominator, or the number means nothing.

Because the funnel has two extra paid gates, an identity check and a first deposit, and because the incentive layer is priced in a volatile asset. Four structural drivers set the gap, and none is a media buying problem.

• Ad platform restrictions push spend toward people. Google’s cryptocurrency advertising policy, read September 7, 2026, requires certification plus local authorization, meaning CASP status under MiCA in the European Union and, in the United States, FinCEN registration plus a state money transmitter license or a bank charter. Every market you cannot certify in pushes budget toward key opinion leaders quoting higher prices for the same reach.
• Identity checks multiply the cost of every gate below them. A registration is cheap, but a KYC pass is not, because you paid full price for every registration that never cleared it.
• Incentive competition raises the price of a first deposit. Points programs, airdrops, and fee rebates are how venues compete for the same depositor, and each is real spend booked somewhere other than the marketing line.
• Compliance review is a per-campaign cost rather than overhead. Sign-off on a rebate offer consumes hours that belong to that campaign.

The cost of user acquisition in crypto is set at the deposit gate, not at the install. A $4.00 media cost per install that becomes a $145 fully loaded cost per funded account is not a failing campaign, it is the normal shape of the funnel. Budgeting off the $4.00 is what kills a quarter.

Take the total cost of user acquisition for the period and divide it by the users who cleared one named gate, usually the funded account.

The crypto-specific numerator lines are media spend, KOL fees, referral payouts, sign-up bonuses, fee rebates, token or points incentives at grant-day value, compliance review hours, team time on acquisition, fraud write-offs, and amortized listing fees. Leave one out, and you have measured media efficiency, nothing more.

CPA and CAC are not the same measurement, and most of the pages ranking for this term treat them as one. CPA is cost per action inside a single campaign, usually an install or a KYC pass. CAC is the fully loaded cost to acquire customer accounts across every channel, divided by those still trading at day 30.

So the first decision in how to calculate user acquisition cost is which of the two you are reporting. Pick one denominator per table, cost per first trade or cost per funded account, and print it in the caption.

How to calculate user acquisition cost is easiest to see against one campaign. The figures below are a worked model built from a $4.00 install cost inside the 2024 fintech range Business of Apps published, not campaign data.

One campaign, three defensible numbers. $9.30 per install, $145.31 per funded account, and $322.92 per trader still active at day 30. The board has seen only the $4.00 media cost per install, one-eightieth of the last.

Yes, valued at grant day, not at vesting and not at the campaign’s start price.

Most software CAC frameworks exclude discounts, and some crypto finance teams book incentives below the marketing line as a treasury item. Userpilot’s benchmark guide is representative, listing ad spend, salaries, content production, and software tools, and naming no incentive at all.

That framing works when a discount is a price concession on a subscription. It fails in crypto because the incentive is the price of the deposit itself.

Book it below the line, and it never reaches CAC. Value it at spot rather than grant day, and CAC falls on paper every time the token does. Nothing about acquisition improved. The accounting moved.

No public source prices an average user acquisition cost by channel for crypto, so four of the six rows below have no number to give you. Know that before you benchmark against anything.

How to read this table. The denominator is one funded account, a deposit above the venue’s bonus-trigger minimum, across the United States and EMEA. The two rows carrying a published figure measure it differently, and each says which.

What this table can tell you. Which channels carry a citable cost anchor, and the design decision that breaks each one.

What it cannot tell you. Your own customer acquisition costs, which depend on your fee tier, bonus design, and KYC pass rate.

KOL cost collapses once the deal is repriced per funded account instead of per post, and the per-post rate cards by follower tier sit in the Surgence influencer list. Quests recruit wallets rather than traders when completion needs no deposit, the failure that makes a badly designed trading competition look cheap.

The referral row is the quiet killer. In most internal channel reports, it is the cheapest line and the most likely to be wrong because a referral program with no hold period, paying the moment an account is funded, is a bonus-farming product with a budget attached. Cheap on the dashboard, dear in the write-off.

Cost per user climbs at every gate, because the same spend divides among fewer people.

How to read this table. One campaign, six stages, fully loaded cost carried forward. The conversion rates are illustrative and show how the cost climbs, not what your rates should be.

What this table can tell you. What each gate costs, and whether your problem sits at the deposit or the first trade.

What it cannot tell you. Why users drop at either. The Surgence guide to install to first trade covers that gate by gate.

Measure at the funded gate, and channel comparisons hold. Measure at the install, and the ranking you get back is wrong.

Send us your channel spend and funded-account counts for last quarter, and we will return a fully loaded CAC by channel. A free acquisition cost audit. Free 30-minute call. No pitch deck. No obligation.

A good CAC for an exchange is one that pays back inside 90 days of trading-fee revenue at the fee tier the acquired cohort actually reaches, not one that beats a software benchmark. Divide fully loaded cost per funded account by monthly trading-fee revenue at that tier, and read the payback period in months.

Run the $145 funded account from the worked example through it. Paying back inside 90 days needs about $48 a month in fees. At $80 a year, it takes 22 months, and at the 26.2 percent twelve-month retention CoinGecko measured on Ethereum, most of that cohort is gone first. At the 7.9 percent Solana figure, almost all of it is.

The 3:1 lifetime value to CAC ratio, LTV:CAC in most decks, assumes revenue that renews on a contract. Crypto LTV is fee revenue multiplied by how long an account keeps trading, and that decays with the crypto market cycle. Model the cohort at your top fee tier, and the CAC looks affordable. Model it at the tier an incentive-led cohort actually reaches, and the same CAC does not pay back. Bonus-hunters arrive on the smallest ticket and stay there.

Every figure below measures something different from a crypto funded account, which is why quoting one to an investor goes badly.

How to read this table. These are customer acquisition costs by industry as their publishers report them, not as we would, and each denominator sits beside its figure.

What this table can tell you. The order of magnitude between an install, a customer, and a retained funded trader.

What it cannot tell you. Whether your own CAC is good. None of these denominators is yours.

Quote the $188 crypto consumer figure to an investor, and the next question is what a funded account that traded twice cost you. The deck does not have it.

Five moves, each with a mechanism and the cost of skipping it.

• Reprice KOL deals per funded account. Put a funded-account floor and a 30-day hold in the contract instead of paying per post. Skip it, and you buy impressions priced on follower count.
• Put a deposit floor and a hold period on every bonus. Pay it only after the deposit sits above the minimum for 30 days. Skip it, and the referral row becomes a bonus-farming budget.
• Stage incentives to the first trade rather than the sign-up. Move the grant to the trade gate so it buys a trader, not a registration. Skip it, and the cheapest cost per sign-up produces the worst day-30 cohort.
• Attribute at the wallet or account level, not the click. Join the ad identifier to the funded account rather than the last touch. Skip it, and paid takes credit every time a crypto growth hacking channel did the work.
• Report one CAC per denominator. Publish the install, funded, and day-30 figures side by side, and take the funded one to the board. Skip it, and finance finds the gap at quarter close, not you.

Ninety days is enough, and order matters more than speed.

Channel mix is the decision this feeds, and it is easier with channel-by-channel costs than a blended average. A free acquisition cost audit. Free 30-minute call. No pitch deck. No obligation.

Published ranges are campaign-level rather than company-level. They move with the market cycle and miss the organic and brand spillover a listing produces.

Incentive costs are the least stable input because a token grant valued on day one is worth something different by day 30.

Attribution across a KOL post, a Telegram group, and an app store install is rarely clean, and last-click flatters paid social in almost every account we audit. Sybil activity corrupts it from the other side because bonus-farmed accounts inflate the funded denominator and deflate CAC until the churn arrives.

• Reporting one blended CAC on mixed denominators. It hides which channel is failing.
• Booking incentives below the marketing line. It removes the biggest cost from the numerator entirely.
• Pricing KOL deals per post. It transfers the funnel risk to you.
• Counting a funded account before the bonus hold period ends. It counts sybil wallets as traders.
• Comparing against software benchmarks. It answers a question your CFO did not ask.

A sign-up bonus, fee rebate, or airdrop tied to a deposit is a financial promotion in the United Kingdom and a marketing communication under MiCA in the EU, and the review each triggers is an acquisition cost line.

The Financial Conduct Authority is explicit on incentives. COBS 4.12A.7R(1) states that a firm “must not communicate or approve a financial promotion which relates to a restricted mass market investment and which offers to a retail client any monetary or non-monetary incentive,” and qualifying cryptoassets sit inside that category. The FCA’s guidance on marketing cryptoassets to UK consumers, last updated February 6, 2026, states that a promotion made outside the permitted routes breaches section 21 of the Financial Services and Markets Act 2000, “which is a criminal offence punishable by up to 2 years imprisonment, an unlimited fine, or both.”

MiCA Article 66(2) requires crypto-asset service providers to give clients “information that is fair, clear and not misleading, including in marketing communications, which shall be identified as such.”

United States exposure works differently. State money transmission and FinCEN registration do not change the arithmetic of cost per user acquisition, but they decide which channels a US-facing venue may buy. None of this is legal advice. Name your jurisdiction, and take it to counsel.

A campaign that clears compliance review late runs at half the planned volume and twice the planned CAC.

We build the measurement stack before buying anything. Every cost line is mapped to a channel first, incentives at grant-day value and compliance hours included, so the numerator is agreed before the spend starts.

Attribution runs at the wallet or account level rather than the click, which makes a KOL campaign and a paid install comparable in one table. Where a client arrives with quarters of mixed-denominator history, we rebuild it on the funded gate so the trend survives.

Our work sits behind Binance, Animoca Brands, Polygon, Falcon Finance, Incentiv, Rails, Biptap, SendIt, Black Mirror, and Boost. What those teams wanted was not a smaller number on the dashboard but one that finance and growth would both sign.

A free acquisition cost audit. Free 30-minute call. No pitch deck. No obligation.

The denominator is the thing to carry out of this. Pick one stage, print it beside every number, and hold every channel to it because a blended CAC built on mixed denominators is not a benchmark, it is an average of things that were never comparable.

Run that discipline for a quarter, and the channel table stops being an argument between growth and finance. The KOL deal gets repriced, the referral program gets a hold period, and the incentive line appears in the numerator where it belongs. Cost per user acquisition stops being a figure you report and starts being a cost you control, and that shift shows up as a better crypto marketing strategy before it shows up as a smaller number.

It is the fully loaded cost of bringing one user to a retained funded account, counting media, influencer fees, referral payouts, bonuses, token grants, and compliance review. The crypto figure runs above software and fintech because the funnel charges you twice more than theirs does, once at the identity check and once at the deposit.

Add every acquisition cost line for the period, then divide by the accounts that reached one named stage. Every line belongs in the cost to acquire customer accounts, the incentives and the compliance hours included, and the worked example in this article runs the full list against one campaign. Use funded accounts as the denominator.

No. Cost per action measures one action inside one campaign, usually an install or a KYC pass, and it is a media efficiency number. Customer acquisition cost measures the fully loaded spend per retained funded user across every channel in a period. Mixing them is one of the commonest reporting errors in crypto.

A good figure pays back inside 90 days from trading-fee revenue at the fee tier the acquired cohort actually reaches. Divide fully loaded cost per funded account by monthly fee revenue, and read the answer in months. The software 3:1 ratio misleads here because crypto retention decays with the market cycle.

There is no defensible average user acquisition cost for crypto in 2026, so publish a range and say what each figure counts. eToro’s FY2025 filing implies roughly $632 per net new funded account on a blended basis, while First Page Sage puts consumer cryptocurrency acquisition at $188 per customer. Neither is an industry average.

Yes, valued on the day the grant is made rather than at vesting. Most software frameworks exclude discounts, and some crypto teams book incentives as a treasury item, which removes the cost entirely. Valuing a grant at spot rather than grant day is another error, since it makes acquisition look cheaper as the token falls.

Yes, for the share of team time and review hours attributable to the campaign. Allocate by the proportion of hours worked on acquisition in the period, and book campaign-triggered compliance review to that campaign. Leaving review hours in overhead understates the cost of running any incentive-funded offer, which is where those hours are actually spent.

It is total acquisition spend for a period divided by the customers gained in that period, expressed per customer. The definition is stable across industries, and the disagreement is always about what a customer is. For an exchange or a trading app, a funded account still trading at day 30 is the defensible answer.

They vary by an order of magnitude and by denominator. First Page Sage reports business-to-consumer financial services at $146 organic and $173 paid, and it puts consumer software at $135 organic and $197 paid. Business of Apps puts fintech installs between $2.50 and $6.00 for 2024, though an install is not a customer.

We map fully loaded cost lines to a channel and attribute at the wallet or account level rather than the click, so a KOL campaign and a paid install sit in the same table. Every report states one denominator at the top, and the funded-account version is the one taken to the board.

• Crypto Exchange Marketing
• App User Acquisition for Trading Apps
• Top Crypto Influencers in 2026