MCP

Email finding and verification

Last verified: 2026-08-05

TL;DR

  • Verify every list before every send. Bounces are the cheapest deliverability problem to prevent and one of the most expensive to have.
  • Targets: total bounce rate under 2%, hard bounces under 1% (2026 guidance) [1]. A 20M+ email corpus puts the observed average at 5.1%, which tells you how many people skip this [2].
  • Verification tools are accurate but not perfect, and catch-all domains are where they all struggle honestly.

Why bounces cost more than the lost emails

A hard bounce is a mailbox provider telling you that you sent mail to an address that does not exist. Do it at scale and you have demonstrated that you did not check who you were writing to, which is exactly the signal that separates a legitimate sender from a bulk one. The consequence lands on your whole domain, so the mail that would have been delivered gets filtered too.

That is the asymmetry: verification costs a fraction of a cent per address, and skipping it puts the entire campaign's delivery at risk. It also cascades — a bounce spike is a reason to pause a running campaign (Campaign scheduling and pausing).

The numbers to hold yourself to

Current guidance puts anything above a 2% bounce rate in "worthy of attention" territory, with top performers holding hard bounces under 0.5% (Campaign Monitor benchmark data, reported 2026) [1]. A large platform corpus reports 5.1% as the average across everyone sending, with under 2% rated good and under 1.5% excellent (2026) [2].

Read those together: the published target is roughly half the published average. Most senders are failing this, which is a large part of why cold email's reputation is what it is — and a reason a verified list is a competitive advantage rather than mere hygiene.

How verification works, and where it stops

Verification checks syntax, then whether the domain exists and accepts mail, then — where possible — whether the specific mailbox exists, without delivering anything.

The last step is where it gets honest. Many corporate domains are configured as catch-alls, accepting mail for any address and sorting it later, which makes it impossible to confirm a specific mailbox from outside. Tools return these as "accept-all" or "risky", and no vendor can do better than that from the outside — treat a tool that reports certainty on a catch-all domain as less trustworthy, not more.

Independent testing puts leading tools in the high nineties, with some vendors' claimed accuracy running a few points above what testing reproduces (2026 benchmark of eight tools) [1]. That gap is worth remembering: a 97%-accurate tool on a 10,000-row list still leaves a few hundred uncertain addresses.

Practical handling

  • Verify after enrichment, before every send — not once when the list was built. Addresses that verified in March may not in September (List hygiene and data quality).
  • Drop the invalid, decide deliberately about the risky. For accept-all addresses, either exclude them or send them from a separate mailbox so the bounce risk is contained.
  • Never send to role addresses — info@, sales@, support@. They are shared, rarely read by a decision-maker, and disproportionately likely to be spam traps.
  • Feed bounces back. A hard bounce suppresses permanently (Stopping rules and suppression), and a pattern of bounces at one company usually means the naming convention you guessed is wrong.

Guessing patterns

Where no provider has an address, the naming convention is often inferable from known colleagues. This is legitimate and it is also how unverified guesses enter lists — always run a guessed address through verification, and never send to more than one guess for the same person.

References

  1. Instantly — 2026 email verification benchmark: accuracy scores for 8 top tools (vendor benchmark, Jan 2026)
  2. Woodpecker — Cold email statistics (20M+ email corpus, updated Jun 2026)