Lead list quality metrics that actually matter
The metrics that tell you whether a lead list is good: match rate, citation coverage, reveal hit rate, real cost per verified contact, bounce handling, and decay.
The metrics that tell you whether a lead list is actually good are match rate, citation coverage, reveal hit rate, real cost per verified contact, bounce handling, and measured decay. Notably absent from that list is row count, which measures the vendor's export, not your outcome. This guide defines each metric, what a healthy value looks like, and how to audit any list or vendor against them.
Row count is the number the category sells because it is the number it can inflate. A 10,000-row export and a 100-brief working set are not two sizes of the same thing; they are different products. Measure the one you are buying.
1. Match rate: did the list answer your ICP?
Definition: leads kept after evidence review ÷ leads delivered.
A run delivers a set of candidates; you review each one's evidence and keep the ones the evidence supports. The share you keep is the match rate, the clearest signal that the ICP-to-candidate matching is doing real work rather than pattern-matching keywords.
- Healthy: a majority of delivered briefs survive an honest read. If you are keeping 10% of what a run returns, either the ICP was vague or the matcher padded.
- Watch for: a tool that discourages review. Bounded output (50-100 briefs per run in CitedLeads) exists precisely so match rate stays measurable. Nobody reviews 10,000 rows; that is the point of returning them.
Match rate is also the metric that improves with your input: tighten the ICP, re-run, and watch it move. See ICP filtering before discovery.
2. Citation coverage: can you check the reasoning?
Definition: share of leads carrying inspectable evidence (which ICP criteria matched, which provider-backed signals support it, what is missing, what counted against).
A lead without evidence is a claim; a lead with evidence is a record. Citation coverage below 100% means some of your list cannot answer "why is this person here", and those rows will be the first to embarrass a rep.
- Healthy: every delivered brief carries its match evidence. In CitedLeads this is structural: the brief is the unit, and the citation section is part of the record, not an upsell.
- Watch for: "confidence scores" substituted for evidence. A score without a visible rubric is an answer to a question nobody asked.
3. Reveal hit rate: how often does spend produce a contact?
Definition: reveals that return a usable email ÷ reveal attempts.
This is the metric credit-based pricing hopes you never compute. It splits cleanly by state:
| Reveal outcome | Counts toward hit rate | CitedLeads cost |
|---|---|---|
| Verified | Yes | 1 credit |
| Catch-all (returned, flagged lower-confidence) | Yes, discounted appropriately | 1 credit |
| Invalid / unknown / miss | No | 0 credits |
- Healthy: the vendor's incentive points the same direction yours does; under a verified-reveal meter, the vendor eats the cost of misses.
- Watch for: "verification rates" quoted without the catch-all share broken out. A large share of B2B domains accept mail to anything; a hit rate that counts catch-alls as full verifieds is measuring the probe, not the mailbox. The mechanics are in verify a B2B email before outreach.
4. Real cost per verified contact: the only price metric
Definition: total spend ÷ verified contacts delivered.
Every other pricing number is an intermediate. A $0.10 credit at a 40% hit rate costs more per verified contact than a $0.60 credit that only moves on a verified outcome, and that is before counting the search and verification meters other tools charge upstream of the reveal.
- Healthy: you can compute it from a ledger, not estimate it from a pricing page. CitedLeads writes every grant, reveal, and adjustment to an append-only credit ledger for exactly this reason.
- Watch for: bundles that make the denominator unfalsifiable: seats, meters, and quotas arranged so no single number survives scrutiny.
5. Bounce handling: what happens after the metric fails?
Post-reveal bounce rate is the audit metric for everything above it: if verified contacts bounce at real rates, "verified" was a label, not a check.
- Healthy: a vendor that treats a hard bounce on a verified contact as their data problem. In CitedLeads, report a bounce and support credits the reveal back manually; the adjustment is a ledger entry like the original spend. There is no automatic refund path; the manual step is deliberate, because a credit-back should be a recorded decision, not a silent counter reset.
- Watch for: bounce remedies that are marketing copy but not mechanics: a policy page with no ledger behind it.
6. Decay: is the list a snapshot or a maintained set?
Definition: change events per period on kept leads (role changes, signal refreshes, contact-validity changes).
People change roles; companies change shape. A list that cannot tell you when that happened decays silently and bills you for the silence. Monitoring that re-verifies kept briefs on a cadence (at 0 credits, since cost should never push you toward stale data) turns decay from a surprise into a log.
- Healthy: the list files change events you can review; monitoring does not compete with growth for budget.
- Watch for: the re-purchase model: a vendor whose only freshness mechanism is selling you the list again.
A quick audit checklist
Before buying or renewing, ask the vendor for, or compute yourself:
- Match rate on a sample run, reviewed honestly.
- Share of leads with checkable evidence (not scores).
- Reveal hit rate, with catch-alls broken out.
- Cost per verified contact, computed from ledger data.
- The bounce remedy, and where it is recorded.
- What the tool tells you when a kept lead's situation changes.
A vendor that can answer all six is selling you a measurable product. One that cannot is selling you rows.
FAQ
What is a good hit rate for email reveals? Compare within the same cohort and insist the catch-all share is broken out; a "90% verified" number that quietly includes catch-all domains is not the same product as one that flags them. The honest comparison is verified-only rates.
Is list size ever the right metric? Only after the other six pass. Volume is what you scale a good list into; it is not evidence of one.
How do I measure match rate if the tool gives me no evidence? You cannot, which is itself the measurement. A tool that cannot show why a lead matched is asking you to trust the number it chose.
Does monitoring really cost 0 credits? Yes. Re-verification of kept briefs runs on a cadence and never touches the credit balance. Costs that punish freshness produce stale lists.
Where does CitedLeads stand on these metrics today? The product is in early access on synthetic demo data while provider contracts clear, so the mechanics above are the printed rules it is built around (verifiable on the pricing and trust pages) rather than measured production statistics. Contact us for early access.