Currently Empty: $0.00
Crucial Email Database Quality Checks Before Buying Any Marketing Leads
Last quarter, a friend who runs a digital agency called me in a mild panic. She’d just paid $800 for a “premium, verified” email list of 20,000 B2B contacts — and after her first campaign, her sending domain was flagged by Google Postmaster as a spam source. Three weeks of careful domain warming, gone in 48 hours.
The database seller? Completely unresponsive. The contacts? A mix of dead addresses, role-based emails like info@ and support@, and a surprising number of addresses that looked like they’d been harvested from decade-old forum registrations. She didn’t do any quality checks before buying. Most people don’t — until it’s too late.
After spending the last couple of years working closely with email lists for my own projects and helping clients audit theirs, I’ve put together a repeatable checklist I run through before spending a cent on any database. Here’s everything, in the exact order I check it.
Why database quality checks matter more than ever in 2026
Email service providers have gotten significantly smarter. Google, Microsoft, and Yahoo tightened their bulk sender requirements in 2024, and the ripple effects are still being felt. A high bounce rate or a flood of spam complaints no longer just costs you a few undelivered emails — it can permanently damage your domain reputation and affect deliverability to your own opt-in list too.
Here’s a reality check on the numbers:
- 23% — average bounce rate on purchased lists
- 0.4% — spam complaint threshold before ISPs start flagging your domain
- 95%+ — typical deliverability rate on opt-in lists
- 30% — average data decay per year in B2B contact lists
That 30% annual decay figure is the one that catches most buyers off guard. A list that was “fresh” 18 months ago has almost certainly rotted by a third. People change jobs, abandon old email addresses, and companies get acquired or shut down. A database is not a static asset — it starts expiring the moment it’s compiled.
The 9-point quality checklist I run before every purchase
Go through these in order. If a seller can’t or won’t help you verify any of these points, that alone is your answer about whether to buy.
1. Ask for the data collection date — and verify it
Any reputable seller should be able to tell you exactly when the data was collected. If they hedge with vague answers like “recently updated” or “continuously refreshed,” push harder. Ask for the specific last-collection date of each batch. A good rule: anything older than 6 months for B2B, or 12 months for B2C, needs to be treated with significant skepticism and must be re-verified before use.
2. Request a sample — and actually test it
Any legitimate list broker will offer a free sample of 100–500 records before you commit to a full purchase. If they refuse, walk away. Once you have the sample, run it through NeverBounce or ZeroBounce immediately. A healthy sample should have less than 5% invalid or risky addresses. If your 200-record sample comes back with 40 bounces, that’s a 20% failure rate — and you haven’t even bought the full list yet.
3. Check the ratio of role-based addresses
Role-based emails — think info@, support@, contact@, sales@, admin@ — are terrible for marketing campaigns. They’re usually monitored by multiple people (or no one), rarely associated with real decision-makers, and frequently configured to auto-report unsolicited email as spam. In a quality B2B database, role-based addresses should make up no more than 5% of the list. Anything above 10% is a red flag worth calling out directly with the seller.
4. Look for duplicate entries
This sounds basic, but you’d be surprised. I’ve seen purchased databases with duplicate rates above 15% — the same contact appearing three or four times with slightly different formatting. Open the sample in Excel or Google Sheets and run a quick deduplication. More than 3–4% duplication on a sample is a sign of poor data hygiene across the whole list.
5. Verify the data source and collection method
This is the one most buyers skip because it feels awkward to ask. Don’t skip it. Ask directly: “How was this data collected?” Legitimate answers include: scraped from public professional profiles with consent frameworks in place, compiled from opt-in B2B directories, or aggregated from industry events with explicit permission. Vague answers like “proprietary sources” or outright refusal to explain are serious warning signs — and a GDPR liability waiting to happen.
6. Check the deliverability score on the sample
Beyond simple bounce checking, tools like Mail-Tester or GlockApps can analyze whether your emails are likely to hit inboxes or spam folders. Before committing, send a test campaign to your sample contacts using your intended subject line, from name, and body copy. Check where it lands. If even your warm-up test is landing in spam on a sample of 100 addresses, the full list will be worse.
7. Confirm compliance with applicable privacy laws
Ask the seller for a written statement confirming the list complies with GDPR (if any EU contacts are included), CAN-SPAM (for US contacts), CASL (Canada), and CCPA (California). Reputable providers like Apollo or ZoomInfo have compliance documentation ready. If a seller can’t provide anything in writing, you’re accepting full legal liability the moment you use their data — and that risk is not worth any price point.
8. Spot-check 10–15 contacts manually
Pick random names from the sample and look them up on LinkedIn. Do the job titles, companies, and locations match what the database says? Are these people still at those companies? Even a quick manual check of 10–15 contacts gives you a ground-truth accuracy read that no automated tool can fully replace. If three out of ten contacts have obviously moved on from the listed company, extrapolate that to the full list size.
9. Ask about their refund or replacement policy
Serious data vendors stand behind their accuracy claims. Most will offer credit or replacement contacts if verified bounce rates exceed a certain threshold — typically 5–10%. If a seller offers no refund or replacement policy at all, they already know the data quality doesn’t hold up. This one question tells you more about a vendor’s confidence in their product than any sales pitch.
Red flags that should make you walk away immediately
Price seems impossibly low — $29 for 100,000 contacts. $49 for “1 million verified emails.” These exist only as scams or recycled breach data. Legitimate, verified data has real acquisition costs — and the pricing reflects that.
No sample offered — Every credible list vendor offers a sample. Period. Refusal to provide one — regardless of excuse — means they already know what you’d find if you tested it.
Vague or evasive on data source — “Proprietary network.” “Exclusive sources.” “Compiled from the web.” These non-answers suggest the data was scraped without consent frameworks, potentially from breached databases.
No written compliance statement — If they won’t put their GDPR or CAN-SPAM compliance claim in writing, they’re either uninformed or deliberately avoiding accountability. Either way, the risk lands on you.
Seller pushes urgency hard — “This deal expires tonight.” “Only 3 copies left.” High-pressure tactics are a hallmark of data brokers who know that anyone who looks closely won’t buy. Don’t rush a decision that affects your domain reputation.
No refund or replacement policy — A vendor with no accuracy guarantee has zero confidence in their own product. This should be a hard stop, not a negotiating point.
What “acceptable” quality actually looks like
Here’s how I score a database sample using the metrics that actually matter:
| Metric | Pass | Warning | Fail |
|---|---|---|---|
| Invalid / bounce rate | Under 5% | 5–10% | Over 10% |
| Role-based address ratio | Under 5% | 5–10% | Over 10% |
| Duplicate entry rate | Under 3% | 3–7% | Over 7% |
| Manual LinkedIn verification (10 contacts) | 8–10 still accurate | 6–7 accurate | 5 or fewer accurate |
| Data age | Under 6 months | 6–12 months | Over 12 months |
| Compliance documentation provided | Written statement | Verbal only | None offered |
If a list scores “warning” on two or more of these, I’d either negotiate a significant price reduction to account for the extra cleaning work, or walk away entirely. If anything scores “fail,” I walk away — no exceptions.
What to do after you’ve passed the quality check
Even a database that clears every check above still needs careful handling before you use it. Here’s the post-purchase process I follow without fail.
- Run the full list through a verifier. Even if you tested the sample, run the complete purchased list through NeverBounce or ZeroBounce before your first send. List quality degrades during transit and formatting issues can introduce new problems. Expect to remove 5–15% of even a “clean” list.
- Remove all role-based addresses before sending. NeverBounce and ZeroBounce flag these. Remove them. Even if a few role-based addresses turn out to be active, the risk they pose to your spam complaint rate is not worth the reach.
- Use a secondary sending domain. Never cold-email from your primary company domain. Set up a closely matched alternate (e.g., getcompanyname.com or companynamehq.com), warm it for 2–3 weeks, and send from there. Your main domain is your brand reputation — protect it.
- Start with a small batch — monitor, then scale. Begin with 100–200 sends. Check deliverability, open rate, and spam complaint rate before scaling. Use Google Postmaster Tools to monitor your domain’s sender reputation in real time. Only expand volume once you’ve confirmed the metrics are healthy.
- Set up proper unsubscribe infrastructure. CAN-SPAM requires an unsubscribe mechanism. GDPR requires it even more strictly. Make sure your ESP supports one-click unsubscribes, and that bounces and unsubscribes are automatically synced back to your master list in real time — not weekly, not manually.
The uncomfortable truth about “quality” databases
Here’s the thing I’ve come to accept after years of working with purchased data: even a list that passes every single check above is still fundamentally inferior to an opt-in list you built yourself. The best purchased database I’ve ever used still converted at about a third of the rate of my organic subscribers — because those organic subscribers actually chose to hear from me.
Bought data is a rental. An opt-in list is an asset. The quality check process exists to make sure you’re at least renting something real — not paying for a parking spot that doesn’t exist.
That said, when you’re launching something new, testing a market, or trying to reach an audience you haven’t built yet, purchased databases can absolutely be a useful tool — as long as you’re honest with yourself about what they are, and you do the work to ensure you’re not torching your domain reputation in the process.
Use the checklist. Ask the hard questions. And if a seller makes you feel like you’re being unreasonable for wanting basic quality assurance — that feeling is your answer.
Quick reference: the non-negotiables
Always ask for a sample and test it. Always get a written compliance statement. Always verify with NeverBounce or ZeroBounce. Always use a secondary sending domain. Never buy from a seller with no refund policy. Never skip the manual LinkedIn spot-check. Never send at volume before testing a small batch first.
If even one of those feels like too much work — the database probably isn’t worth buying in the first place.