Contact Waterfall Workflow: Find Verified Emails for 80%+ of Any List

A waterfall is a sequential chain of data providers where each provider is tried in order, and the chain stops as soon as a result is found. For email finding, the waterfall might be: try Apollo first (strong B2B database, good for US tech companies), then try Hunter (strong for European companies and smaller businesses), then try Clay's native email finder (good fallback with its own unique dataset), and finally try a LinkedIn email extraction as a last resort. Each step only runs if the previous step returned no result or returned a low-confidence result. This matters because Clay charges credits for each enrichment step that runs. A well-configured waterfall minimizes credit burn by only escalating to expensive providers when cheaper ones fail. The goal is maximum coverage at minimum cost per contact found.

Create your Clay table with the input contacts. Each row needs at minimum a first name, last name, and company domain. LinkedIn URL significantly improves match rates, so include it if available. Add your first email finding column using Apollo's Clay integration. Set the confidence threshold to accept results at 80%+ confidence only. Next, add a second column using Hunter's API, configured to run only when the Apollo column returns empty. Then add Clay's native email finder as the third column, running only when both Apollo and Hunter return empty. Finally, add Apollo as a fourth option configured for LinkedIn-based email finding, which uses a different lookup path than domain-based search. For each column, check 'Run only if previous column is empty' to enforce the cascade logic. This prevents redundant lookups and keeps credit usage efficient.

Finding an email is not the same as finding a valid email. After the waterfall columns run, add a validation column that checks each found email against a verification service. Clay integrates directly with NeverBounce and ZeroBounce for this purpose. Configure the validation column to run after the waterfall and categorize each result: valid (safe to send), catch-all (server accepts all emails, deliverability uncertain), risky (likely invalid), and invalid (do not send). For your outreach, only accept 'valid' results. Flag 'catch-all' emails in a separate column for manual review: some catch-all domains are fine, others are not, and you need to make that call per domain. Remove 'risky' and 'invalid' emails entirely before exporting to your sequencer. This single validation step is what keeps your bounce rate below 2%, which protects your domain reputation.

After your first batch of 500+ contacts runs through the waterfall, pull the analytics to measure performance. Calculate fill rate per provider: how many emails did Apollo find versus Hunter versus Clay native? Calculate validation pass rate: of the emails found by each provider, what percentage passed as 'valid'? A provider with high fill rate but low validation pass rate is filling your list with bad emails, which is worse than not filling it at all. Based on this data, reorder your waterfall to prioritize providers with the best combination of fill rate and validation pass rate. Re-run the batch analysis every month as your list profile changes. Enterprise contacts on one domain type may have different provider coverage than SMB contacts on another. The waterfall order that works best for your ICP today may need adjustment as you move into new market segments.