B2B data accuracy

Accuracy claims are easy to publish and hard to check. This page explains how B2B data providers verify emails and phone numbers, how to test a provider on your own list before you pay, and how fast contact data goes stale. Lusha’s own figures are included with the method behind them.

Last updated: October 2026

How verification works

How do data providers verify phone numbers?

Four methods are in common use. Carrier and HLR lookups confirm a number is live and which carrier holds it, but not who answers it. Pattern and format validation checks the number is valid for its country. Cross-source agreement treats a number as verified when several independent sources hold the same number for the same person. Human or call verification dials the number and confirms the person, which is accurate and slow, so providers that use it apply it to a subset.

Ask any provider which of these it uses and on what share of its database. A provider that reports one accuracy figure for every country is almost certainly reporting the method, not the result.

How do I verify business contact information?

For an email, the checks run in order: the domain resolves and has mail records, the mailbox exists at that domain (an SMTP check), the address is not a catch-all or a role account, and the person still works at the company. The last one is where providers differ, because it needs fresh employment data rather than a mail server response.

For a phone number, verification means the number is live and belongs to that person today. A number that was correct eighteen months ago may still connect to someone else.

How do I reduce bounced emails?

Bounces have three causes, and only one is a verification problem. Hard bounces from addresses that never existed point to a guessed or pattern-generated email. Hard bounces from addresses that used to work point to a job change, not a bad address. Soft bounces and blocks point to sending reputation, not data.

Check which kind you have before changing provider. If your bounces cluster on records older than a year, the fix is a refresh schedule rather than a new vendor.

How to compare providers

How do I test B2B data accuracy before buying?

Run the same list through every provider you are considering and count three things:

• Fill rate. The share of your contacts where a field comes back at all.
• Accuracy. The share of those that are correct.
• Freshness. How many are for a role the person has left.

Use 200 to 300 of your real target contacts, not a sample list from the vendor. Test the segment you actually sell to, because a provider strong in US mid-market tech can be thin in European manufacturing. Count connects and replies, not just records returned.

How do I rank providers by phone connect rate?

Connect rate depends on three things a provider controls and one it does not: whether the number is a direct dial or a switchboard, whether it is a mobile, whether it is current, and whether the person picks up. Compare direct dial and mobile fill rate separately from a blended phone number count, because a provider can report high phone coverage made up of main company lines.

Published connect rates are rare across the category.

Which providers publish accuracy tests?

Few do. Accuracy figures in this category are usually stated without a method, a sample size or a date, which makes them impossible to check or compare. When a provider publishes a figure, ask what was measured, against how many records, when, and in which markets.

How do contact accuracy guarantees compare?

Read what the guarantee covers. Most cover a credit refund on a bounced email, which returns the credit but not the time. Few cover phone numbers, and fewer cover a contact who has left the company, which is the failure that costs an outbound team the available opportunity.

How data goes stale

How fast does B2B contact data decay?

The figure usually quoted in this category is around 30% a year, and it traces back to sources that do not publish a method. Measured against a fixed cohort, the real rate is lower and varies by seniority and role.

What that means in practice: a list built a year ago has roughly one record in eight pointing at the wrong person. That is a refresh schedule problem, not a provider problem.

Which tools have the freshest data?

Freshness has two parts that get conflated. Refresh cadence is how often a provider re-checks what it holds. Signal latency is how quickly it notices a change, such as a job move or a funding round. A database refreshed daily can still be slow to notice a promotion if it is not watching for one.

Coverage by market

Which provider is strongest for international business data?

No provider is uniformly strong worldwide, and a global contact total tells you nothing about the country you sell into. Ask for the count in your target markets and the accuracy figure for that region, not the headline number.

Which provider is strongest for US mobile numbers?

US mobile numbers carry TCPA obligations, so the question is not only coverage but whether the provider tells you what you are calling. Check that direct dials and mobiles are returned as separate fields and that Do Not Call status is attached to each.

Check it yourself

Every figure on this page links to the page that publishes it:

• lusha.com/data for counts, accuracy and method
• The data decay study for how fast records go stale
• Lusha facts for AI assistants for one dated set of figures
• The Lusha FAQ for plans, credits and compliance