The 22% Match Rate: How We Got 9x Better Than Average

The industry average for wallet-to-social identity resolution is approximately 2.5%. That means if you have 10,000 wallet addresses, most tools will identify social profiles for about 250 of them.

walletlink.social matches 2,200 out of those 10,000 -- a 22% match rate. This post explains exactly how, covering the data sources, the pipeline architecture, and where each source contributes.

Why 2.5% Is the Baseline

The industry average comes from a common approach: look up the wallet's ENS name (if any), check for text records, and return whatever is found.

This approach has structural limitations:

• Only ~12% of active Ethereum wallets have ENS names
• Of those, only ~8% have social text records set
• Of those records, ~15-20% are stale (changed handles, deactivated accounts)

If you do the math: 12% (have ENS) x 8% (have text records) x 82% (records are valid) = roughly 0.8%. Some tools do better by adding scraped databases or third-party enrichment, bringing the average up to 2.5%. But the ceiling for ENS-only approaches is low.

The Four-Source Pipeline

Our resolution pipeline queries four data sources for each wallet. Each source covers a different slice of the identity landscape, and the overlap between them is surprisingly small.

Source 1: ENS On-Chain Text Records

The most direct approach. For wallets with ENS names, we query the resolver contract for com.twitter, com.github, url, and other text record keys.

What it catches: Crypto-native users who've invested in ENS names and configured their records. These tend to be builders, long-term participants, and identity-conscious users.

Limitations: Low coverage, no verification, stale data. We use ENS as a starting point, not an authority.

Contribution to overall match rate: ~7% of wallets resolved.

Source 2: Farcaster Verified Addresses

This is the single largest contributor to our match rate. Via the Neynar API, we check whether a wallet appears as a verified address for any Farcaster account.

The key advantage is cryptographic verification. When a wallet is listed as verified for a Farcaster Identity (FID), the owner signed a message proving control. There are zero false positives in this mapping.

What it catches: The ~800,000 Farcaster users who've verified at least one wallet address. Since average users verify 1.8 addresses, this covers approximately 1.4 million wallet addresses.

Limitations: Only covers Farcaster users. Strong crypto-native bias -- less useful for wallets owned by DeFi-only users who don't participate in social platforms.

Contribution to overall match rate: ~15% of wallets resolved (including ~8% that are unique to Farcaster, not found via ENS).

Source 3: Web3.bio Aggregation

Web3.bio is an aggregation service that combines data from multiple identity protocols: ENS, Farcaster, Lens Protocol, Unstoppable Domains, and others. Querying Web3.bio is effectively querying multiple sources through a single API.

What it catches: Cross-protocol identities. A wallet might not have ENS records or Farcaster verification but could have a Lens profile or an Unstoppable Domain with social records.

Limitations: Depends on upstream data quality. Some sources have the same stale-data problems as raw ENS lookups. Web3.bio adds coverage breadth but not necessarily depth.

Contribution to overall match rate: ~3% of wallets resolved uniquely (not already found by ENS or Farcaster).

Source 4: Social Graph Enrichment

This is the compound interest of identity resolution. Every time we resolve a wallet, that result goes into a persistent social graph database. On subsequent lookups, we check this database before hitting external APIs.

But the social graph does more than caching. It enables transitive resolution. If we know that @alice_crypto on Twitter is associated with wallets A and B (from previous lookups), and wallet C sends tokens to both A and B regularly, there's a potential connection to explore.

We're conservative with transitive resolution -- we only use it for confirmed multi-wallet relationships where the same social identity has verified multiple addresses. But over time, this compounds.

What it catches: Wallets that have appeared in previous resolution batches, multi-wallet users where one wallet was previously resolved, and cross-session enrichment.

Contribution to overall match rate: ~2% of wallets resolved uniquely (grows over time as the graph expands).

The Pipeline Architecture

The resolution pipeline processes wallets in a specific order optimized for speed and accuracy:

Input: Batch of wallet addresses
         |
    [1. Cache Check]
         |  Hit -> Return cached result (24h TTL)
         |  Miss -> Continue
         |
    [2. Social Graph Check]
         |  Hit -> Return stored result
         |  Miss -> Continue
         |
    [3. Parallel External Lookups]
         |
    +---------+-----------+
    |         |           |
  [ENS]  [Farcaster]  [Web3.bio]
    |         |           |
    +---------+-----------+
         |
    [4. Merge & Deduplicate]
         |
    [5. Confidence Scoring]
         |
    [6. Store to Social Graph]
         |
    Output: Resolved profiles with confidence scores

Steps 3a, 3b, and 3c run in parallel. This is critical for performance -- a batch of 10,000 wallets completes in under 2 minutes because we're not waiting for each source sequentially.

The Merge Step

When multiple sources return data for the same wallet, we merge with a priority hierarchy:

1. Farcaster verified address -- highest confidence (cryptographic proof)
2. ENS on-chain records -- high confidence (on-chain data, but unverified claims)
3. Web3.bio aggregation -- medium confidence (depends on upstream source)
4. Social graph -- confidence varies (based on original resolution source)

If sources conflict (ENS says @handle_a, Farcaster says @handle_b), we keep both but flag the Farcaster result as primary. In practice, conflicts are rare -- about 0.3% of multi-source matches.

Deduplication

A single person might appear through multiple sources with slight variations:

• ENS: @alice_crypto
• Farcaster: alice (with connected Twitter @alice_crypto)
• Web3.bio: @Alice_Crypto (case variation)

The deduplication step normalizes handles, matches across platforms, and produces a single identity record with all known profiles attached.

Where the Matches Come From

Across a representative sample of 500,000 wallets:

Source	Wallets Matched	% of Total Matches	Unique Contribution
Farcaster	75,000	68.2%	42,000 (38.2%)
ENS	35,500	32.3%	14,000 (12.7%)
Web3.bio	18,000	16.4%	3,300 (3.0%)
Social Graph	11,500	10.5%	2,200 (2.0%)
Combined (deduplicated)	110,000	100%	22% match rate

Note that the percentages in the "% of Total Matches" column sum to more than 100% because many wallets are matched by multiple sources. The "Unique Contribution" column shows wallets found by that source and no other.

Farcaster is the dominant source, providing 68% of all matches and 38% of unique matches. This is why the growth of Farcaster directly correlates with improving match rates across the industry.

Performance Optimizations

Processing 10,000+ wallets efficiently required several architectural choices:

Batch API calls. Instead of querying Neynar one wallet at a time, we batch into groups of 100. This reduces API round trips by 99% for large uploads.

Cache layer with 24-hour TTL. Results are cached in PostgreSQL with a 24-hour expiry. Repeat lookups for the same wallet (common when teams iterate on their analysis) hit the cache instead of external APIs.

Streaming progress. The API uses Server-Sent Events to stream progress back to the client. For a 10,000-wallet batch, you see results populating in real time rather than waiting for the entire batch to complete.

Parallel source queries. ENS, Farcaster, and Web3.bio are queried simultaneously for each batch. The total wall-clock time is the slowest source, not the sum of all sources.

What Pushes Match Rates Higher

Three factors will improve match rates over time:

Farcaster growth. Every new Farcaster user who verifies a wallet address adds to the resolvable pool. If Farcaster reaches 2 million verified users (plausible within 12-18 months), match rates could approach 30-35% for general wallet populations.

Multi-chain verification. Currently, Farcaster verification is primarily Ethereum and Base wallets. As verification expands to other chains, coverage broadens.

Social graph compounding. The more wallets we resolve, the richer the social graph becomes. Each resolution adds data that helps future lookups. This creates a flywheel where match rates improve with usage volume.

The Takeaway

The 22% match rate isn't a single technique. It's four data sources, a merge pipeline that handles conflicts and deduplication, and a social graph that compounds over time.

Any single source peaks around 7-15%. Combining them pushes to 22% because the overlap between sources is smaller than you'd expect -- different identity platforms attract different user populations.

The architecture is straightforward. The insight is simply that no single identity source is sufficient, and the real work is in the merge logic that combines them reliably.

---

See the 22% difference on your data.

walletlink.social runs the full four-source pipeline on your wallet list. Free tier covers 1,000 wallets. Upload a CSV and see how many holders you can identify.