A prospecting list without phone numbers is a list of names. That observation created an entire enrichment market, and with it an inflation of claims that are hard to check: “98 percent coverage”, “millions of verified contacts”, “the best match rate in the industry”.
Those numbers are not fabricated. They are simply measured on something other than what you are about to do. A match rate on emails has nothing to do with a match rate on direct mobile lines. Coverage in the US has nothing to do with coverage in Germany.
This guide gives the figures actually observed, a method for testing a vendor, and the compliance framework that applies now that business contact data is no longer exempt.
What “match rate” means, and what it hides
Match rate is the share of your contacts for which the vendor returned data. That definition is broad enough to cover very different realities.
Three questions to ask before accepting a figure:
A match on what? A work email can often be inferred from a naming convention — first.last@company.com — and then verified automatically. That is why email match rates routinely hit 85 to 98 percent. A direct mobile cannot be inferred from anything: it has to have been collected somewhere. The gap between the two is structural, not commercial.
A match in what geography? Coverage is generally strongest in the US, thinner across Europe, and thinner again outside both. A vendor that is excellent in Chicago can be mediocre in Milan.
A match on what function? The most underestimated variable. Commercial and revenue-facing roles are over-represented in these databases, because they leave the most public traces. Technical, finance and legal profiles are considerably less well covered.
| Data type | Realistic coverage |
|---|---|
| Work email | 85 to 98 % |
| Direct dial or mobile, all functions | 40 to 70 % |
| Direct dial, commercial and revenue roles | Upper end of that range |
| Direct dial, technical, finance and legal roles | Lower end, often below 50 % |
Waterfalls: why one vendor is not enough
B2B databases overlap only partially. Each has its strengths: one built on US corporate filings, another on browser-extension contributions across Europe, a third on industry partnerships.
Hence the waterfall: query vendors in sequence. The first answers, you stop. It misses, you try the second, then the third.
The gain is real. The cost per enriched contact rises, necessarily, since you may pay several lookups for one contact.
The right economic metric is not cost per enriched contact. It is cost per contact you actually reach — and the two can move in opposite directions.
A worked example on 1,000 contacts makes the trade-off legible:
| Single vendor | Three-vendor waterfall | |
|---|---|---|
| Enrichment cost | $300 | $480 |
| Direct dials returned | 450 | 640 |
| Cost per direct dial | $0.67 | $0.75 |
| Contacts actually reached | ~135 | ~192 |
| Cost per contact reached | $2.22 | $2.50 |
The waterfall is marginally more expensive per contact reached, but it produces 42 percent more of them with the same team and the same hours. When your constraint is headcount rather than data budget — which it usually is — the second number decides.
The 100-contact test protocol
This is the only serious way to evaluate a vendor, and it takes half a day.
Build a representative sample
100 contacts drawn from your real target: same company sizes, same functions, same industries, same regions. The temptation is to pick large enterprises and commercial titles, where everybody performs well. That tests the easy case, not yours.
Measure raw match rate
The share of contacts for which data comes back. A starting point, not a conclusion.
Separate direct dials from switchboards
Number by number. A main-line prefix, a number already published on the company website, the same number returned for several people at one company: those are switchboards. This is where most of the gap between promise and reality shows up.
Call the sample
The only test that counts. Measure connect rate and the share of calls that reach the right person. An accurate but stale number, or one belonging to someone who left, counts as a match and is worth nothing.
Compute cost per contact reached
Total enrichment cost divided by people actually spoken to. That is the figure to compare across vendors, and it frequently ranks them differently from the advertised match rate.
The compliance picture, without hand-waving
Most articles on enrichment skip this. It is not complicated, but it changed recently and a lot of teams are operating on outdated assumptions.
Business contact data is no longer exempt
The CCPA originally carved out B2B contact information. That exemption expired on January 1, 2023. Since then, for businesses covered by the CCPA, work emails, work phone numbers and job titles are ordinary personal information: the individual has rights of access, deletion, correction, and opt-out of sale or sharing.
Several other states have since enacted comparable comprehensive privacy laws, with their own thresholds and mechanics. The practical consequence is the same everywhere: a business contact can ask you to delete their record, and you have to be able to do it.
That single requirement has an architecture implication most teams have not thought through:
- You need to know which vendor supplied which record, so you can answer a source request.
- You need a deletion path that reaches every copy — the enrichment tool, the dialer, the CRM, the exports sitting in someone’s spreadsheet.
- You need the opt-out to be honored across future enrichments, not just once. Deleting a record that gets re-enriched next month is not compliance, it is a loop.
TCPA follows the number, not the buyer
This is the point most often missed in B2B. TCPA obligations attach to how you dial and what you dial, not to whether the call is business-to-business. Calls placed to a wireless number using an autodialer or an artificial or prerecorded voice fall within its scope.
Enrichment tends to return exactly that: mobile numbers. So the better your enrichment, the more your dialing configuration becomes a compliance question rather than only a productivity one. Manual-dial workflows, consent tracking and cell-number flagging are the controls that matter here.
Retention is a decision, not a default
There is no single statutory retention period for prospect data in the US. The defensible approach is the same as everywhere: tie the duration to a purpose and enforce it automatically.
A practical rule that also happens to be good commercial hygiene: a list enriched three or four years ago and never reworked is not just legally awkward, it is commercially dead. People changed jobs, numbers moved, companies were acquired. Purging old files protects you and protects your connect rate at the same time.
The 5 mistakes that cost money
- Comparing vendors on advertised match rate. It blends emails and phones, direct dials and switchboards, geographies. Compare on cost per contact reached.
- Testing on an easy sample. Enterprise accounts, commercial titles, major metros: everyone performs there. The test has to look like your target.
- Enriching the entire database at once. Data decays. Enriching on demand, at the moment the contact enters a sequence, costs less and returns fresher numbers.
- No source tracking. The day someone asks where you got their number, “we bought a list” is not an answer you want to give.
- Never measuring afterwards. Connect rate per enriched batch is the only feedback loop that lets you arbitrate between vendors over time. Our connect rate benchmarks give you the reference points.
What to take away
Enrichment is not bought on a match rate. It is bought on a cost per contact reached, measured on your own target. That single change of metric separates offers that looked equivalent.
And keep the real order of magnitude in mind: 40 to 70 percent on direct dials, lower on technical and finance roles. A team that sizes its call targets on the vendor’s marketing claim rather than that number is building a plan that will not hold.
On what happens to the data once it is in the tool, and how it syncs with your CRM, see our guide to phone system and CRM integration.
