Every B2B data provider leads with a number: 200 million contacts, 300 million, 500 million. None of them lead with how many are still true, which is the only property that affects whether your outreach lands. Judging this market means ignoring the headline and running a test.
The field
| Provider | Strongest in | Model | Watch out for |
|---|---|---|---|
 Stalar |
Any market, verified at request time | Pay per contact found | Early access, request-based onboarding |
 ZoomInfo |
US enterprise, org charts, intent | Custom, five figures | Thin outside North America, sticky renewals |
 Cognism |
Europe, verified mobiles, GDPR | Custom contract | Contract-priced, no self-serve |
 Apollo |
Self-serve breadth, US-weighted | Seats plus credits | Phone lookups cost 8x email credits |
 Clay |
Running many providers at once | Credits | Needs an operator to be worth it |
 Lusha |
Fast individual lookups | Credits, free tier | Coverage varies by region |
 Kaspr |
LinkedIn-side European lookups | Credits | Rep-level tool, not a data platform |
 Hunter |
Email discovery by domain | Cheap tiers | Email only, no phone data |
ZoomInfo
The deepest US dataset, priced accordingly. Buyers on review platforms and in procurement communities report $10,000 to $40,000 a year for small and mid-sized teams, and well into six figures once intent and API access are in the contract. Org charts and intent are what that money buys. Outside North America the same contract buys materially less, which is the single most common reason a European team lets it lapse. The pricing detail and the alternatives are both worth reading before a renewal, because the renewal is where whatever leverage you have lives.
Cognism
The European answer. Its mobile numbers are checked by people, which is slower and dearer than scraping them and is why the hit rate holds up in markets where US-first databases thin out. Contract-priced, no self-serve tier. The compliance posture is built to survive a procurement review, which in Europe is not a nice-to-have but the thing that decides whether you can buy at all.
Apollo
Apollo won its adoption on packaging: $49, $79 and $119 per user per month, a free tier under that, and a database, a sequencer and a dialer behind one login.
Model the credits before you sign. A phone lookup costs eight times an email lookup, allowances expire at the end of each month, and a team doing phone-heavy outbound can spend more on credits than on seats.
Clay
Clay is not a provider. It waterfalls through the other providers until one returns a result, which is real leverage for anyone who will sit and maintain the tables, and an expensive subscription for anyone who will not: roughly $149 a month at Starter, $349 at Explorer, $800 at Pro. Budget the operator, not the seat.
Lusha, Kaspr and Hunter
The light end. Lusha and Kaspr both work per lookup from a browser extension, Kaspr with better European data since Cognism bought it. Hunter finds emails by domain from about $34 a month and does not do phone numbers at all.
All three are rep tools, and none of them build a list. Knowing whom to look up is still your job.
Stalar
Stalar is not a database. Describe the market you sell into and agents map it, pick the right person at each company, and verify the email and the mobile at that moment, in the local language. You pay per contact found, so a number that turns out to be dead is not billed to you, and coverage gets built for your market instead of inherited from wherever the vendor’s data happens to be deep.
What arrives is not a file, either. The contacts land in a workspace whose sales brain holds every prospect, conversation and touchpoint, kept current from what was actually said on the call, and the agents there brief your meetings and draft the outreach for you to approve.
The test that settles it
- Pick 100 real target accounts in your actual market, not a demo list they choose.
- Ask each provider for one specific role at each: not “a contact”, the role you sell to.
- Count three numbers: people returned, emails that pass verification, and mobiles that connect when dialled.
- Divide the annual price by the third number. That is your real cost per usable contact, and it reorders most shortlists.
What to weigh beyond coverage
Three questions, in the order they will bite you.
How were the records sourced, and what happens when someone asks to be told? In Europe this is not a preference, it is the question that stops a purchase in legal review.
What does the contract do when you stop paying attention? Auto-renewal dates and expiring credit allowances are where the real cost lands, not the list price on the plan page.
When was this record last checked? That is the number that predicts your bounce rate, and almost nobody publishes it. Our own production data shows how fast the decay sets in once records sit in a CRM with nobody maintaining them.
