Agency email verification is not the same job as in-house verification. You are not cleaning one list — you are cleaning a different client's list every week, often uploaded by someone who does not know where the addresses came from, and you are the one who gets blamed when the campaign bounces.
That changes the buying criteria. A tool that is fine for a single SaaS sender can be unusable in an agency: no per-client separation, credit pools that one client exhausts, or an API rate limit that breaks when three campaigns launch on the same morning.
This is the checklist we would use as an agency, plus a test you can run on any vendor before signing.
The seven agency-specific requirements
1. Predictable cost per client list
Agencies buy verification as a cost of delivery, not as a subscription someone forgets. You need to know the cost per 10,000 checks before you quote a client, and you need it not to change because a list turned out to be 40% catch-all.
Ask: "What does it cost to verify 50,000 addresses where 30% are catch-all?" If catch-all results consume full credits, your quote is wrong on exactly the lists that need the most work.
2. Per-client separation
Mixing ten clients in one dashboard creates two problems: an accidental export that exposes another client's list, and no clean audit trail when a client asks what you deleted. Look for separate projects, folders or sub-accounts, and check whether results are retained or purged after download.
3. Status codes a junior can act on
Agency work gets delegated. If the export says "invalid" for both a dead mailbox and a timed-out server, someone will eventually delete a recoverable contact. You want distinct statuses for invalid, catch-all, risky/unknown, role, disposable, spam trap and over-quota — and a documented rule for what to do with each.
4. Volume without surprises
Check three numbers: rows per upload, rows per day, and what happens to API limits while a bulk job runs. Then check the same three numbers on the plan one tier below, because that is the plan your client will ask you to use.
5. An API you can put in a workflow
Verification should run at three moments: before a list enters the CRM, before every campaign send, and at signup for lead-gen clients. That requires batch endpoints, sane rate limits and error responses that name the field.
6. GDPR and data-processing terms
You are a processor for your client's data, which makes the vendor a sub-processor. You need a DPA, a retention period you can state in a contract, and a deletion path. "We keep uploaded files indefinitely" is a sentence that should end the evaluation.
7. A repeatable schedule
Lists decay between campaigns. An agency that re-verifies before each send has a materially different deliverability story than one that verifies once a year. Pick a vendor where re-running a stored list is cheap and fast, and write it into your process.
The agency workflow that keeps bounce rates under 2%
- Intake. Client sends a CSV. Record where it came from and when. Lists older than 12 months go straight to a re-verification step, not to a campaign.
- First pass. Verify the whole file. Never send to an unverified list "just to test".
- Segment by verdict. Valid → main campaign. Catch-all → separate, smaller batches from a warmed inbox. Risky/unknown → retry after 24h before deciding. Invalid, disposable, spam trap → suppress permanently.
- Suppression. Merge unsubscribes and hard bounces into a client-level suppression list that survives list changes.
- Pre-send check. Re-verify anything older than 30 days, plus any address added since the last run.
- Post-send review. Compare bounce rate by segment. A bounce spike almost always traces back to one source list, one catch-all domain, or one import.
If you run this loop, the vendor matters less than the process. If you skip steps 3 and 5, the best vendor on the market will still produce a 5% bounce rate.
How to test any vendor in 15 minutes
Build one seed file of 200 addresses where you already know the correct answer:
| Group | Count | What you are testing |
|---|---|---|
| Deleted mailboxes you control | 25 | False negatives (missed invalids) |
| Catch-all domains from client lists | 20 | Classification, and whether they bill you |
Role addresses (info@, sales@) |
15 | Whether role addresses are flagged |
| Disposable domains | 8 | Signup-fraud coverage |
| Spam traps | 2 | Whether traps are flagged before you send |
| Known-good contacts | 30 | False positives (killing real pipeline) |
Then score three numbers: recall on invalid addresses, catch-all handling and billing, and false-positive rate on known-good contacts. The last one is the one vendors optimise against and buyers forget to measure.
Run the same file through your current provider. If your current tool scores well, the honest answer is that you do not need to switch — you need the workflow above.
Comparison: what separates vendors for agency use
| Criterion | Weak answer | Strong answer for agencies |
|---|---|---|
| Catch-all billing | Full credit per catch-all check | Catch-all and risky results not billed as extra credits |
| Upload size | 10k–25k rows per job | 100k rows in one upload |
| Status detail | valid / invalid | 6+ statuses with a documented action per status |
| API | single-check only | batch + single, documented rate limits |
| Retention | indefinite | stated period, deletion on request |
| Free tier for testing | none | enough checks to run the seed-file test |
Where email-validaton.com fits
We are biased, so treat this as a claim to test rather than a conclusion:
- Catch-all and risky results are not billed as extra credits, at any volume, up to 100,000 addresses per upload — which is the part that usually breaks an agency quote.
- Free plan: 100 checks per month, no card. That is enough to run the 100-address seed test above on real data before you commit.
- REST API plus WordPress, Zapier and Google Sheets integrations, and PHP, Python and CLI clients for the intake step.
- Statuses are granular enough to drive the segmentation in step 3 above.
Limits and current pricing are on the pricing page — verify them there, and verify them again at renewal, including on us.
FAQ
How many checks does a typical agency client need per month?
Most SMB lists run 5,000–50,000 addresses per campaign, verified 1–4 times a month. Model the top of that range before choosing a plan; overage is where budgets break.
Should we verify before every campaign or once a quarter?
Before every campaign if the list is older than 30 days. Quarterly for stored databases that are actively used. Lists decay continuously, not in steps.
Is a free plan enough to evaluate a vendor?
Yes, if it allows at least ~100 checks so you can upload a seed file with known answers. Anything less is a demo, not a test.
What bounce rate should we promise a client?
Under 2% on a verified list is a defensible target; under 1% usually requires list-level work (segmentation, suppression, source quality), not just verification. See what a normal bounce rate looks like.
Do catch-all addresses have to be deleted?
No. Segment them into smaller batches from a warmed inbox and watch the bounce rate. Deleting them removes a large share of legitimate B2B contacts — here is how catch-all actually behaves.
