Last updated: 2026-05-06
How a DeFi Lending Protocol Attributed Their First $1M in TVL
Six channels, $25K of marketing spend, 90 days, and one unambiguous question: which dollars actually moved on-chain TVL? Here is what the attribution data showed — and what it changed about the next quarter.
Disclaimer: simulated scenario, anonymized protocol, illustrative figures. Channel ranges are consistent with published 2026 DeFi marketing benchmarks.
The setup
A new lending protocol launched on Base in early Q1 2026. Like most new DeFi protocols, the question was not whether to spend on marketing — it was whether the spend was actually buying TVL. The team allocated $25,000 across six channels for the first 90 days, and instrumented every dollar with on-chain attribution before the first campaign went live.
The instrumentation stack was deliberately simple: UTM-tagged links on every campaign asset, a wallet-connect handler that stored the UTM alongside the connected wallet address, a 30-day on-chain attribution window, and wallet quality scoring to filter bot traffic before it polluted the channel ROAS numbers. No proprietary tooling, no machine-learning model — a UTM table, a wallet-event listener, and a 30-day SQL query.
The decision rule was set in advance: any channel under 1.5x ROAS at day 90 gets reduced or cut. Any channel over 4x gets its budget doubled in Q2. Anything in between gets discussed.
The question
Six channels were running by week two: three Tier 1 KOLs (more than 100K followers), five Tier 2 KOLs (10–50K followers), Twitter / X Ads, four sponsored Discord communities, two newsletter sponsorships, and a small content/SEO investment on three long-form guides. By the end of Q1 the protocol had crossed $1,000,000 in TVL attributed to these six channels. The non-trivial question was not the total — it was how that million was distributed.
Channel breakdown — Q1 2026
| Channel | Spend | Clicks | Wallets | Real | Deposits | TVL | ROAS |
|---|---|---|---|---|---|---|---|
| Tier 1 KOLs (3 × >100K followers) | $9,000 | 14,200 | 1,180 | 540 | 168 | $252,000 | 2.8x |
| Tier 2 KOLs (5 × 10–50K) | $5,000 | 6,800 | 1,020 | 690 | 172 | $220,000 | 4.4x |
| Twitter / X Ads | $4,000 | 11,500 | 690 | 320 | 48 | $52,000 | 1.3x |
| Discord Ops (4 communities) | $3,000 | 2,200 | 880 | 760 | 195 | $240,000 | 8.0x |
| Newsletter Sponsorship (2 sends) | $2,000 | 3,400 | 480 | 410 | 87 | $70,000 | 3.5x |
| Content / SEO (in-quarter) | $2,000 | 1,200 | 95 | 88 | 12 | $11,000 | 0.6x (Q1) / 5–8x (projected Q2–Q4) |
| Total | $25,000 | 39,300 | 4,345 | 2,808 | 682 | $845,000+ | 3.4x blended |
Direct/organic and brand traffic added the remaining ~$155K in TVL to clear the $1M mark. The six measured channels accounted for $845K of attributed deposits at a 3.4x blended ROAS on $25K of spend.
Five surprises in the data
Discord ops beat every paid channel on capital-efficiency
A $3,000 spend on four community managers, AMA hosting, and pinned-thread coordination produced more attributable TVL than the entire $9,000 Tier 1 KOL line. The win came from connect-to-deposit conversion: 22% of Discord connects deposited, vs 14% for KOLs and 7% for Twitter Ads. People who show up via Discord arrive with intent.
Tier 2 micro-KOLs beat Tier 1 macros on ROAS
Tier 1 KOLs delivered higher absolute deposits ($252K vs $220K) — but on $9K of spend vs $5K. ROAS: 2.8x vs 4.4x. Tier 1 paid for follower count we did not need; Tier 2 paid for engaged niche audiences. Reallocating $4K from Tier 1 to Tier 2 in Q2 was the single most profitable reallocation.
Newsletter sponsorship was the sleeper
Two newsletter sponsorships ($1K each) drove 3.5x ROAS — middle of the pack — but the connect-to-deposit conversion was 18%, second only to Discord. Newsletter audiences self-select for long-form attention. We added one more sponsorship in Q2 and held the same ROAS.
Twitter Ads moved volume, not deposits
11,500 clicks at the lowest CPC on the chart. Looked great on top-of-funnel reporting. ROAS: 1.3x. The connect-to-deposit was 7%, lowest of any channel. Without on-chain attribution we would have doubled this budget; with it, we cut it 50%.
Content/SEO returned 0.6x in Q1 — and that’s the right number
Twelve deposits attributed in-quarter against $2K spend on three guides and a benchmarks page. SEO does not pay back inside 90 days. By Q2 the same three pages were generating 35–55 deposits/month with no additional spend. Honest attribution requires honest time horizons; killing the SEO spend at the Q1 mark would have been the wrong call.
What changed in Q2
The decision rule produced four concrete moves for Q2, all driven by the channel data above and not by what the team wanted to be true:
Doubled Discord ops
$3K → $6K. Hired one part-time community manager to cover two more communities. Expected ROAS held at 7–8x in Q2.
Reallocated $4K from Tier 1 to Tier 2 KOLs
Same $14K total KOL line. Tier 2 went from 5 creators to 9. Tier 1 dropped to one campaign. Forecast ROAS: 4.0–4.5x.
Cut Twitter Ads 50%
$4K → $2K. Narrowed targeting to retargeting only — high-quality traffic from people already on the site. Expected ROAS: 2.0–2.5x.
Held Content/SEO budget despite Q1 ROAS
The 0.6x first-quarter ROAS was the expected value, not a failure signal. Held at $2K and added one more long-form guide. Q2 ROAS came in at 4.8x once the Q1 pages ranked.
The lesson — and the part that does not generalize
The lesson is not that Discord beats KOLs, or that Tier 2 beats Tier 1. The lesson is that you cannot run that decision rule without on-chain attribution. Every channel looked broadly comparable on top-of-funnel metrics. The 6x ROAS spread between Discord and Twitter Ads only existed once we attached deposits to UTMs. Without that, the next quarter would have been: bigger Tier 1 KOLs, more Twitter Ads, less Discord. The opposite of what worked.
The part that does not generalize: channel ROAS is protocol-specific. A perpetuals DEX, a yield aggregator, and a stablecoin issuer all have different audiences, different deposit-velocity, and different attribution windows. The ranking we found here — Discord > Tier 2 KOL > Newsletter > Tier 1 KOL > Twitter Ads > SEO (in-quarter) — is a starting hypothesis for a similar protocol, not a universal law. The right move is to instrument first, run for 90 days, and let your own data set the rule.
The full instrumentation pattern is documented in our DeFi marketing attribution guide and the channel benchmark ranges in our 2026 benchmarks page.
FAQs
How do you attribute TVL to specific marketing channels in DeFi?
Tag every campaign link with a unique UTM. When a wallet connects on the landing page, store the UTM alongside the wallet address. Watch that wallet on-chain for deposits within an attribution window (typically 14–30 days). The deposit is then attributable to the original UTM, and the channel ROAS is the deposit value divided by the channel spend.
What is a realistic ROAS for DeFi marketing channels?
ROAS is highly channel-dependent. In our simulated scenario, Discord operations delivered the highest ROAS (8x), followed by Tier 2 micro-KOLs (4.4x), newsletter sponsorship (3.5x), Tier 1 macro-KOLs (2.8x), and Twitter Ads (1.3x). Content/SEO returned 0x in the first 90 days but compounds in months 4–12, where it typically passes 5x.
How long should an attribution window be for a DeFi lending protocol?
The median deposit happens 5–7 days after the click; the 90th percentile is around 30 days. We used a 30-day window, which captured 92% of attributable deposits. A 7-day window would have under-credited slow channels (newsletter, content) by roughly 35%.
Do Tier 1 macro-KOLs (>100K followers) outperform Tier 2 (10–50K)?
Not in our scenario. Tier 1 KOLs delivered higher absolute deposits but at 1.6x the ROAS of Tier 2. Tier 2 micro-KOLs had lower bot rates, more engaged audiences, and lower per-creator fees. The same pattern appears in published industry data: niche, smaller KOLs typically beat broad influencers on capital-efficiency.
What’s the biggest mistake teams make when attributing DeFi marketing?
Stopping at wallet connections. A wallet connect is not a deposit. In our scenario, only 22% of connected wallets actually made an on-chain deposit. Channels with high connect-to-deposit drop-off (Twitter Ads, broad KOLs) look attractive on top-of-funnel metrics and look terrible on TVL-per-dollar. Always attribute to the on-chain action, not the wallet connect.
Attribute your own first million
Set up UTM-tagged campaign links, wallet quality scoring, and a 30-day attribution window in a single afternoon. Your data will tell a different story — and the right story.