What a verified B2B email actually costs in 2026.

Here is the surprise that reshaped this article while we fact-checked it: the villain we expected does not exist. We went looking for providers that charge you for failed lookups, and among the majors we could verify, we found none. Hunter charges a credit only when an email is found. Datagma says it plainly: no match, no credit. LeadMagic and Findymail, the same.

And yet the arithmetic still refuses to land anywhere near the credit price. We wired six finders into one production waterfall, and an address you can safely send to still costs two to three times what any pricing page suggests. The gap has three causes, and none of them is a provider cheating you.

The sticker prices, re-verified 5 September 2026

Two things to notice before the arithmetic. The per-found price swings almost sixfold between a starter plan and a volume plan of the same product, and annual billing quietly moves it another 30%, so compare plans on the terms you would actually buy. And the last column is now boringly uniform: pay-on-success has become table stakes. That makes the next section the entire story.

Three reasons your invoice still disagrees

The residue cascade. No provider finds everyone, so a waterfall sends the leftovers of each finder to the next one. But the leftovers are, by construction, the hard cases: every tier finds a smaller share of a harder pool, usually at a higher per-found price. Your first 550 addresses might cost $0.0085 each. Address number 650 did not.

Verification is a second bill, and catch-alls are a third. A found address is not a sendable address. Everything gets one verification pass, and a large share of B2B domains are catch-alls that accept mail for any name at all, real or invented: the server says yes to everyone, which is a shrug, not an answer. Those need a second, specialist pass, and it lands on exactly the addresses that were hardest to find.

Subscription minimums and the utilisation tax. A waterfall needs accounts at several providers, each with a monthly minimum, rollover caps, and credits that quietly expire. Pay $49 for 2,000 credits and use 900, and your real per-credit price just doubled without a single price changing. This is the cost nobody's calculator shows, because it lives in your utilisation, not their pricing.

A thousand-person list, walked through

Illustrative model, stated as one: hit rates swing hard by audience. Take 1,000 people. A volume-priced finder at $0.0085 per found address finds 55%: 550 addresses, $4.68. The 450 leftovers go to a second finder at roughly $0.02 per found; it finds 30% of what the first could not, 135 more, $2.70. Only finds were billed, and the blended finder cost is already $0.0108 per address, a quarter above the best sticker.

Now the part that is not optional: 685 addresses through verification at $0.008 is $5.48, and the roughly one third living on catch-all domains need the second layer, call it $2.26 more. Total: about $15.12 for 685 sendable addresses, which is $0.022 each. Two and a half times the best credit price, on a friendly list, before a single wasted subscription credit. Make the audience developers instead of marketers and watch the hit rates drop.

A useful triangulation, with the bias flagged: Findymail now markets on this exact distinction, charging only for verified results and pitching its cost per usable email at around two cents against higher figures for typical providers. Vendor maths, not an independent benchmark, but it is striking that a provider's own framing lands in the same band our model does.

The red-flag number

Price the thing you actually want: an address you can send to without burning the domain you send from. By that measure, anything much below a cent per sendable address should worry you. At that price someone is reselling stale data or skipping the verification that keeps your bounce rate survivable, and either way the cost has not disappeared. It has moved into your domain reputation, which is the one thing in this whole stack you cannot buy back.

What our waterfall does with all this

Order by effective per-found price and stop at the first verified hit, so the cheap tier sees everything and the expensive tiers only meet the residue. Cache every result globally, so nobody pays twice for the same person. Learn the domain's address pattern from every found email, whoever found it, because the second person at a company should cost one verification, not a finder fee. And gate the last-resort provider on a known pattern, so its requests are high-odds instead of hopeful. None of this is secret now. All of it compounds.

If you are assembling this yourself

The arithmetic above assumes you already hold accounts with two finders and two verifiers, each with its own minimum, and that you enjoy gluing CSVs together at midnight. That is the actual price of DIY: not the credits, the subscriptions and the plumbing between them. It is also, transparently, why our product exists: GTM for One (gtmforone.com) runs this exact waterfall, with both verification layers, inside one flow. We are biased. The arithmetic is not.

What we can't tell you

Prices were re-verified against the providers' published pages on 5 September 2026 and will drift; check them before quoting this page at anyone. Hit rates vary by audience more than by provider, and our illustrative rates are exactly that. What holds regardless of the tools you pick: model the cascade, verify everything, treat catch-alls as unfinished work, and count your unused credits as part of the price, because they are.