Most DeFi teams talk about “the funnel,” but very few measure it in a way that reflects how users actually behave onchain.

In Web2, the funnel is linear: view → signup → activate → retain.
In DeFi, users move across chains, connect multiple wallets, interact with several protocols, and often disappear and reappear months later.

Growth only becomes meaningful when teams understand:

• How awareness leads to acquisition

• How acquisition lead to activation

• How those actions lead to revenue and retention

This article explains how the DeFi growth funnel works in practice, how to instrument each stage, and how funnel insights change growth priorities.

Why DeFi funnels are non-linear

DeFi funnels are non-linear because user journeys span multiple wallets, chains, and protocols, which breaks any single straight path from discovery to retention.

Users often bridge assets, switch interfaces, and re-enter the ecosystem after long periods of inactivity. Some users skip the frontend UI altogether and interact onchain directly. This leads to fragmented paths that do not follow a simple top-to-bottom flow. As a result, linear funnels undercount real engagement and misrepresent where drop off actually happens.

What makes DeFi funnels non-linear

• Users bridge between chains before first transaction

• The same user may use multiple wallets

• Activity can pause for weeks or months before reactivation

• Value can flow across protocols rather than stay in one app

Non-linear paths make it harder to interpret drop off, which is why funnel design must account for re-entry and cross protocol behavior.

The identity problem in DeFi growth funnels

The identity problem is worse than it looks. A user sees your protocol on Twitter, clicks through to your site, reads the docs on their laptop, connects wallet A briefly to check gas fees, then closes the browser. Two days later they return on their phone, connect wallet B, and deposit $40,000.

A standard funnel sees this as: one visitor who bounced (laptop session), plus one new user from direct traffic who activated on first visit (phone session). The Twitter acquisition gets no attribution credit. The considered path that took 48 hours shows as instant activation. The same user appears as two separate users with two separate wallets and zero attribution overlap.

This is why wallet-based funnels require cross-device session stitching (typically using wallet address as the canonical identity once a user connects) and why first-touch attribution models undercount consideration-heavy channels like documentation, Twitter threads, and influencer content.

Mapping funnel stages to onchain outcomes

DeFi funnel stages must be mapped to verifiable onchain events because there is no centralized account system to rely on.

Each stage in the funnel should correspond to a measurable event that signals real progress. This leads to better attribution and clearer growth diagnosis.

Mapping funnel stages to chain level events reduces ambiguity because each step is verifiable and auditable.

A rough benchmark for a healthy DeFi acquisition funnel: 10,000 visitors produce roughly 300-500 wallet connections (3-5% connect rate), of which 30-50% will complete a first transaction within 7 days. Of those first transactors, 25-40% will return for a second transaction within 30 days. Of repeat transactors, roughly 20-30% become retained depositors with meaningful TVL.

These numbers vary significantly by protocol type. DEXes tend to see higher activation rates (the first transaction is the whole point, friction is low) and lower retention (users go where liquidity and fees are best). Lending protocols see lower activation rates (deposits require more consideration) but higher retention among users who do activate (positions are sticky). Vaults and yield aggregators have the longest consideration phase and the highest LTV once users commit.

The diagnostic value is in comparing stage-to-stage transition rates against your own historical baseline and against your channel mix. A 40% wallet-to-transaction rate is not inherently good or bad: it depends on whether your channels bring in active DeFi users or crypto-curious newcomers.

How to measure funnels and what to track at each stage

Funnel instrumentation works when each stage is tied to a measurable signal because this creates causal links between marketing actions and onchain outcomes, as detailed in our guide to tracking crypto acquisition funnels.

Tracking only traffic or wallet connections leads to shallow insights. Tracking contract interactions and capital retention reveals where growth actually compounds.

Instrumentation should follow the value flow, not just user counts. This leads to better prioritization of product work over top of funnel spending.

How to segment funnels by channel

One of the highest-value analyses you can run once your funnel is instrumented is breaking transition rates by acquisition source. The same funnel looks completely different depending on where users came from.

A DeFi team might find that Discord users convert wallet-to-transaction at 55%, while Twitter users convert at 18%, and DeFiLlama organic referrals convert at 62%. The traffic volume might be reversed (Twitter sends more raw visitors), but the quality picture inverts when you look at who actually transacts.

Segment your funnel by utm_source across at least 90 days of data before drawing conclusions. Small channels with high conversion quality are often underfunded relative to their true acquisition value.

Common funnel blind spots DeFi teams miss

DeFi teams miss blind spots because they focus on volume metrics instead of transition quality between stages.

High wallet connections do not guarantee first transactions. High first transactions do not guarantee revenue or TVL. Each transition has its own friction that requires a separate diagnosis.

Frequent blind spots

• Overvaluing wallet connections without transaction follow-through

• Measuring volume, revenue, and TVL without separating short-term incentive-driven deposits

• Ignoring time to first transaction

• Treating all users as equal value segments

• Ignoring churned users

This leads to growth strategies that inflate surface-level metrics while long term usage decays.

How funnel insights change growth priorities

Funnel insights change priorities because they reveal where marginal effort creates the highest long term value.

If most users connect wallets but never transact, onboarding friction should be prioritized. If users transact once but never return, retention mechanics matter more than acquisition. This is why funnel analysis leads to reallocation of budget from campaigns to product improvements.

How priorities shift

• If wallet-to-transaction rate is below 25%: The most common causes are: (1) the first action requires too large a commitment (asking for a deposit before users understand the yield), (2) gas fees are visible but value is not, (3) the time between connect and the first clear call to action is too long. Tactical fixes: surface the action directly on the connect confirmation screen, show expected return before any deposit, add a low-minimum "try it" path.

• If first-to-repeat transaction rate is below 30%: Users completed one action but found no reason to return. Common causes: no in-app activity feed showing what other users are doing, no email/notification hook after first transaction, no visual progress toward a goal. Tactical fixes: post-transaction confirmation should show next steps, not just a success state. Email sequence triggered by first transaction within 24 hours.

• If retained TVL declines despite repeat transactions: Users are transacting but withdrawing net capital. This is the mercenary liquidity pattern and is distinct from churn. It means users find utility in the product (they keep using it) but store value elsewhere. The fix is not retention mechanics but yield or risk architecture.

Funnel insights shift growth from channel optimization to system optimization.

Learn more about Web3 Funnel Analysis: Improve User Activation & Retention.

Final takeaway

The DeFi funnel is not one problem with one solution. It is five distinct transitions, each with its own conversion rate, each with its own failure mode, and each requiring a different fix. Teams that do this well go from chasing activity to building sustainable revenue.

The most useful thing you can do with this framework is calculate four numbers: site-to-connect rate, connect-to-first-transaction rate, first-to-repeat transaction rate, repeat-to-retained TVL rate. Those four numbers show you where the sharpest drop occurs, which determines where intervention creates the most leverage. Teams that skip this and go straight to tactics end up spending growth budget optimizing stages that are not the constraint.

Two misreadings to avoid: high TVL does not mean you have solved the retention stage. If TVL tracks reward campaigns and withdraws when incentives end, you have rented capital, not retained capital. The diagnostic is whether your baseline, non-incentivized TVL is growing. Second, low conversion at any stage does not automatically require product work. It might mean your acquisition channels are bringing in users who were never going to convert at this protocol. Run the funnel segmented by source before spending on product fixes.

The protocols that compound growth are not the ones with the biggest launches or the most TVL at peak. They are the ones that steadily improve each stage transition, measure the real signal at each stage rather than the vanity metric, and build products that give users a reason to stay when the incentives run out.

Check out our complete DeFi marketing framework to learn more.

FAQs

How do I calculate funnel conversion rates between stages?

Each transition rate is a simple division, but the time window matters.

Site-to-wallet-connect rate: unique wallet connections in a period divided by unique visitors in that period. Use a 30-day window rather than same-session comparisons, since many users research across multiple sessions before connecting.

Connect-to-first-transaction rate: unique wallets with at least one confirmed transaction divided by unique wallets connected. Measure this at 7, 30, and 90 days to understand your activation curve shape.

First-to-repeat transaction rate: wallets with two or more transactions divided by wallets with at least one. Typically measured at 30 days.

Retained TVL rate: wallets with active deposits at day 60+ divided by wallets that made their first deposit. A net-deposit version (deposits minus withdrawals as a percentage of peak period TVL) tells you how much of your TVL base is actually sticky.

The most important thing: calculate these per cohort, meaning wallets that first connected in the same week, rather than pooling all time periods. Cohort-level rates reveal whether the product is improving or degrading over time. Aggregate rates hide this.

What is a reasonable wallet-to-first-transaction conversion rate?

It depends on protocol type and traffic quality, but some rough orientation helps.

For a DEX, where the first transaction is a swap and the value is immediate, a 30-day connect-to-first-transaction rate of 40-60% is achievable for intent-driven traffic (DeFiLlama referrals, protocol-specific search). For a lending protocol, where the first action requires understanding collateral and liquidation risk, 20-35% is more typical. For vaults, where trust in the strategy is a prerequisite to depositing, 15-25% is common for organic traffic.

These numbers degrade for low-intent traffic. Broad influencer or Twitter campaigns typically produce site-to-connect rates around 1-2% and connect-to-transaction rates in the 10-20% range. This is not necessarily a product problem: the audience was not actively looking for your protocol.

The most reliable benchmark is your own historical data by cohort. If 30-day activation rate is falling cohort-over-cohort without a change in traffic mix, that is a product signal. If it is flat while traffic sources are changing, diagnose the source composition first.

How do I know which funnel stage to fix first?

Find the stage where the gap between your current rate and the realistic ceiling for your traffic quality is largest, then start there.

As a general rule: fix activation before acquisition. Every dollar spent bringing in new wallets returns less if 80% of those wallets never transact. Activation improvement compounds across every existing and future acquisition campaign.

One useful heuristic: if comparable protocols with similar traffic sources see meaningfully better rates at a given stage, that gap is a real opportunity. If your rate is already at or above the realistic ceiling given your traffic mix and protocol type, adding more top-of-funnel volume is the right move.

When you cannot determine which stage to prioritize from rates alone, look at absolute numbers. If 9,500 visitors do not connect wallets but 200 who did connect never transact, the activation gap is smaller in absolute terms. The stage with the largest absolute loss of users is usually where fixing the leak has the most immediate impact on growth.

Why do users connect a wallet but never make a transaction?

This drop happens for a few distinct reasons that require different fixes.

The first action asks for too much commitment too soon. Connecting a wallet costs nothing. Depositing real capital into an unfamiliar contract is a different decision. If there is no low-stakes entry path, users stall at the commitment threshold. Surfacing a minimal viable first action (a small swap, a test deposit with low minimum) reduces this.

The expected value is unclear at the decision point. Users can see the cost of the first action (gas, slippage) before they can see the return. If yield, output, or savings are not displayed at or before the transaction confirmation screen, the perceived risk-reward calculation does not close.

Trust signals are absent where they are needed most. Protocol audits, TVL figures, and user counts on the marketing page do not help users who have already connected and are deciding whether to transact. These signals need to appear near the action, not just at the top of the funnel.

To diagnose which of these applies: measure median time from wallet connect to first transaction attempt in your event data. If the median is over 20 minutes, users are deliberating, and the barrier is trust or value clarity. If it is under 2 minutes and they still do not transact, the barrier is friction in the action itself.

We get TVL during campaigns but it disappears after rewards end. Why?

Reward-driven TVL leaves because it was rented, not retained. The capital was seeking the highest short-term yield, your protocol was offering it, and when the rate fell, that capital migrated to the next opportunity. This is mercenary liquidity and it is structurally different from organic retained TVL.

The diagnostic test: measure baseline TVL, the non-incentivized TVL excluding new deposits made during the campaign period, in the 30 days before and after a campaign ends. If baseline is growing, you have a real retention foundation and campaigns are additive. If baseline is flat or declining and all TVL movement tracks incentive periods, the protocol's organic value proposition has not yet been validated.

Retained TVL comes from users who have a reason to stay unrelated to rewards: yield better than alternatives in the same risk tier, utility they cannot get elsewhere, or switching costs that make moving expensive. Building one of these is a product problem. Campaigns are a distribution problem. Conflating them is why TVL grows only when the next campaign is running.

Which metric matters more, TVL or active users?

Neither alone is sufficient, and the most useful single metric is usually Revenue Per Wallet: protocol fees generated per active wallet per month.

TVL is misleading because it includes rented and speculative capital. Active users are misleading because a $50-a-week transactor counts the same as a $2M liquidity provider. Revenue Per Wallet combines both dimensions. It tells you the revenue density of your user base, which determines whether growth is compounding or diluting.

A protocol with rising TVL and falling RPW is likely onboarding lower-quality wallets or seeing high-value users reduce positions. A protocol with flat TVL and rising RPW is seeing its best users deepen engagement even as acquisition slows. The second situation is a more durable growth foundation.

If RPW is not yet trackable, use active transacting wallets as the user proxy and net deposits (new deposits minus withdrawals in the period) as the TVL quality proxy. Avoid raw TVL as a headline internal metric.

Why does our funnel look broken compared to SaaS funnels?

The comparison benchmarks are wrong. A SaaS product with a 3% signup-to-activation rate has a serious problem. In DeFi, a 3% site-to-wallet-connect rate for cold traffic is typical. A 40% first-transaction-to-repeat rate would be exceptional SaaS retention. In DeFi, this is achievable for protocols with real utility.

The structural reason: SaaS users create accounts, which are free and reversible. DeFi users transact, which costs gas and involves real capital. This raises the intent threshold significantly. Users who do activate have already passed a higher bar than equivalent SaaS signups, which is why post-activation retention in DeFi often looks better than raw comparison to SaaS suggests.

The other source of confusion is identity fragmentation. One person using multiple wallets across multiple devices appears as 3-5 separate users in a standard analytics tool. This inflates new user counts, deflates returning user counts, and makes retention curves look worse than they are. Wallet-based identity resolution, using wallet address as the canonical identity once a user connects, is what makes DeFi funnel analysis accurate.

Is it normal that growth slows after launch even when traffic is still coming in?

Yes, and it follows a predictable pattern. Launch creates a spike driven by novelty (coverage, community announcements, early adopters) and usually by launch incentives. Both are one-time drivers. After they dissipate, what remains is the protocol's organic pull.

The number to watch in the weeks after launch is not total users or TVL but the week-over-week retention rate of activated wallets. If users who transacted in week 1 are still active in week 4 at a rate above 30%, you have real product retention and the slowdown is a distribution problem: you need more sustained awareness channels. If week-1 users have largely churned by week 4, the slowdown is a product problem and more acquisition spend will not fix it.

Most post-launch slowdowns are a combination: the top of funnel is harder to refill without another launch moment, and bottom-of-funnel retention is weaker than it appeared during the incentive period. Diagnosing which is dominant determines whether the next 90 days should focus on growth marketing or product depth.