On this page10 sections
Published October 27, 2026.
A purchased or scraped list rarely fails at import. It fails on day three, when the sequence has already burned the domain and the sender reputation. Revalidation is the step between "we have 4,000 rows" and "we are willing to put our domain behind these contacts." It should be a short, repeatable gate with a written pass rule, not a spot check by whoever built the list.
Start with the entry criteria
Write down what makes a row eligible before you look at the data. A typical rule includes a business email on a domain that matches the company, a title in the buying group, a country the team can legally contact, and no existing open opportunity. If the team cannot state the rule, the list is not ready to validate.
Separate the list's purpose from its source. A list built for a webinar invitation has different eligibility than a list built for a direct outbound sequence. Reusing one list for both is how unsubscribes climb.
Check the domain before the person
Group rows by domain and inspect the domain first. A domain with no MX record will bounce every contact. A domain with a catch-all server may accept mail but never route it to a person. A domain that redirects to a parked page is a dead company.
Flag free-mail domains separately. They are not automatically invalid, but they change the deliverability risk and the way the sequence should be written. Do not mix them into the same sending pool without a reason.
Verify the person and the role
Run the list through the team's approved verification tool, then spot-check the results manually. Verification scores are evidence, not a guarantee. A "valid" result on a role address such as info@ is still a shared inbox and should be treated as such.
Check the title against the buying group. A list that claims 800 "marketing leaders" but contains 300 coordinators and 200 sales reps is not a targeting problem to fix in the copy; it is a list problem. Recount the roles before the sequence is built.
Deduplicate by person and by account
Deduplicate on the normalized email first, then on the person's name plus company domain. Then inspect the account level. Five contacts at one company may be a legitimate buying committee or five versions of the same person from different sources.
Keep the source and the last-verified date on every surviving record. If two sources disagree about the title, keep both and mark the conflict rather than silently choosing the newer one. The sequence writer needs to know which claim is supported.
Apply suppression before enrichment
Load the global suppression list, the customer list, open opportunities, recent opt-outs, and any contractual exclusions. Do this before enrichment so the team does not pay to enrich a contact it cannot email. Re-apply suppression after enrichment as well, because enrichment can change the domain or the person.
Keep a written record of the suppression snapshot date. A contact who opted out after the snapshot must still be excluded; the sequence tool's own unsubscribe handling is not a substitute for the pre-send check.
Sample the list by hand
Pull 30 rows at random and inspect them. Check the domain, the title, the company size, the country, and the source. If more than a small share fail the entry criteria, stop and fix the source rather than cleaning row by row. Manual cleaning at scale is a sign that the list should not be used.
For a high-value account list, call or verify the top 50 accounts individually. The cost is small compared with the cost of a damaged sender reputation or an embarrassing first touch.
Gate the sequence entry
Do not let the sequence tool import directly from the raw file. Use a staging table with a status field: passed, needs review, excluded. Only passed rows enter the sequence. The person who built the list should not be the only approver; a second reviewer should check a sample and the exclusion counts.
Record the counts at each stage: raw rows, duplicates removed, suppressed, failed verification, failed role, passed. Those counts are the list's audit trail and the first place to look when reply rates fall.
Measure the first send, not just the import
After the first 100 sends, check bounce rate, unsubscribe rate, and reply rate against the team's baseline. A bounce rate above the agreed threshold should pause the sequence, not trigger a second batch. Inspect the bounced domains to see whether the verification step missed a pattern.
If replies are low but deliverability is healthy, the problem may be the offer or the role, not the list. Do not "fix" a messaging problem by buying more rows.
Keep the list from decaying
Set a revalidation interval, typically 90 days for outbound lists and 30 days for high-turnover roles. Re-check the suppression list and the open opportunities before every reactivation. A list that was clean in January is not clean in April.
Store the validation evidence with the campaign record so a later reviewer can see which tool, which date, and which rule produced the pass.
Implementation checklist
- Write the entry criteria and the list's purpose.
- Inspect domains for MX, catch-all, and parked pages.
- Verify emails, then manually sample the results.
- Recount titles against the buying group.
- Deduplicate by person and by account.
- Apply suppression before and after enrichment.
- Stage rows with a pass-review-exclude status.
- Check the first 100 sends before releasing the rest.
