Lead enrichment waterfalls: how to stack vendors without burning credits

We benchmarked six lead-enrichment vendors across 12,000 LinkedIn-sourced contacts in March 2026. No single vendor cleared 70% verified-email coverage on its own. Stacked correctly, the waterfall cleared 85% at 40% lower per-contact cost than running any single vendor exhaustively.

Here's how to build the stack.

What "enrichment" actually means

Enrichment is the process of taking a LinkedIn URL (or name + company) and returning verified work contact info — primary work email, direct phone, sometimes title and seniority. Each vendor scrapes, infers, or buys this data from different sources, so coverage and quality vary widely by ICP.

Per-vendor performance, March 2026

Our benchmark, 12,000 LinkedIn URLs across SaaS, services, and consumer brands:

Caveats: hit rates skew higher for US-based SaaS contacts, lower for international or non-tech roles. ZoomInfo wins on raw coverage but loses on cost. Limadata wins on cost-quality balance.

The waterfall logic

A waterfall is: try the cheapest, highest-quality vendor first. If it misses, fall through to the next. Stop the moment you get a verified contact.

The 2026 stack that maximizes coverage while minimizing spend:

Email waterfall

1. Limadata — try first. Cheapest at 1¢/credit and high hit rate.
2. Findymail — fall through if Limadata misses. Strong B2B coverage.
3. Apollo — third try. Best for SMB/mid-market.
4. Wiza — fourth try. Catches LinkedIn-native contacts the others miss.
5. ZoomInfo — last resort, only for high-ACV deals.

Phone waterfall

1. Limadata (1¢/credit for engager surfacing, 5-10¢ for phones)
2. Datagma — strongest phone vendor at low cost
3. Apollo — fallback for verified mobile
4. ZoomInfo — high-ACV only

Email + phone deduped per contact

Run them as parallel waterfalls, not sequential. Hit email and phone vendors at the same time. Stop each waterfall the moment that field is filled.

Critical rules

1. Always verify before billing the customer. A vendor "match" isn't the same as a verified deliverable address. Run every email through Twilio Lookup or NeverBounce before charging an enrichment credit. ~15–18% of "matched" emails will fail verification — don't burn customer trust on them.

2. Cache aggressively. LinkedIn URLs don't change. Email-to-LinkedIn-URL mappings don't change often. Cache by canonical URL. We see 30–40% cache hit rate inside a month for active workspaces, which translates to direct cost savings.

3. Detect out-of-credits errors. Every vendor returns subtly different errors when your account hits a credit cap. Detect these specifically (Datagma's 402, Apollo's specific JSON shape) and short-circuit the waterfall to the next vendor instead of treating them like real misses.

4. Use the vanity URL. Many vendors return obfuscated LinkedIn URLs (ACoAA...). Use the vanity URL (e.g., linkedin.com/in/yardenhofer) for cache keys. We've seen workspaces over-paying because the same person was being enriched repeatedly under different URL shapes.

5. Set a per-contact budget cap. A runaway waterfall can spend $1.50 on a single contact across 6 vendors. Cap at 3 vendors per field per contact. Better to leave a phone unfilled than to burn credits chasing a contact that doesn't exist.

When the waterfall fails

For ~15% of contacts, no vendor returns a verified work email. The honest move: surface this clearly in the UI ("no public contact found") and don't bill a credit. Reps will appreciate the transparency. Trust is the long-game asset.

How Saava handles this

Saava ships with a built-in waterfall using Limadata, Apollo, Datagma, Findymail, and Wiza in that order. We only charge enrichment credits on verified contacts. The whole motion happens in milliseconds, transparent to the rep.

For high-ACV workspaces, we can wire in ZoomInfo as a final fallback — but most teams don't need it once the rest of the waterfall is tuned.