What Should Revenue Operations Teams Evaluate in Lead Enrichment? Three Scenarios, One Decision Tree

In 2022 I signed off on a $14,200 enrichment contract. The vendor demo showed an 87% match rate. Our real number, once the data hit our actual CRM, came in at 61%. That gap cost roughly two SDR headcount-months chasing dead addresses, plus a lot of internal credibility I haven't fully earned back.

I've built a checklist since then. But the honest version of the lesson isn't "here's the framework." It's that there is no single framework. What you should measure depends on what you're actually trying to do — and I've watched three very different setups get this wrong in three very different ways.

Scenario A: You're a 1–3 person outbound team (or founder-led sales)

If this is you, the temptation is to optimize on price per record. I did this for the first two years. Let me save you the time — it's the wrong metric at your scale.

When you're sending 500 to 2,000 emails a month, you're buying maybe a couple thousand records a quarter. A $0.05-per-record difference is $100. That's not a real number on your P&L. What is real is whether those 2,000 records get into your sequence without a separate CSV-shuffling ritual every Monday morning.

What to evaluate at this stage:

• Free tier or pay-as-you-go. If a vendor insists on an annual contract with a 10K-record minimum, walk away. You don't have the volume to justify it, and the renewal negotiation is going to be miserable.
• Native integration with whatever you're sending from. HubSpot, Pipedrive, Smartlead, whatever's in your stack. If it requires a Zapier bridge to work, you'll spend more time babysitting the sync than the tool is saving you.
• What's NOT included. I've learned to ask this before I ask the price. Is verification extra? Is intent data a separate SKU? Is export capped at 500 rows a day? These are the fees that quietly turn a $49/month plan into a $180/month plan by the third invoice.

One thing I'd flag here, and it cuts against a lot of what's written about this: don't buy the biggest list you can afford. Buy what you'll actually send to in 30 days. Bought data decays faster than you think — B2B contact turnover runs somewhere in the 20–30% per year range depending on industry and seniority, which means your "freshly verified" file from January is a coin flip by July.

Scenario B: You're running RevOps or GTM engineering for 5–15 SDRs

Different game entirely. You're not shopping for a tool. You're shopping for a data layer.

The metric that matters here isn't coverage. It's match rate against your actual CRM records. Coverage is a marketing number — it says "we have 200M contacts." Match rate says "of the 80,000 accounts you already own, we can return verified data for X%." Those are not the same thing, and vendors love to blur the two.

The other trap at this stage is treating price-per-record as the deciding factor. It isn't. If you're enriching 80,000 records a quarter, you have three cost centers:

1. The enrichment spend itself.
2. The engineering hours to wire it up and keep it clean.
3. The SDR time lost when bad data slips through.

Number 3 is the one that never shows up on the procurement spreadsheet until it's a problem. I want to say I've lost roughly a week of one SDR's time per quarter to what I now call "ghost data" — records that pass a basic email-format check but bounce on send. Don't quote me on the exact number, but it's not small.

What to actually measure during a pilot:

• Run 1,000 records from your worst list. Not the vendor's sample file. Pull the 1,000 messiest contacts out of your CRM — stale titles, missing direct dials, old company domains — and test those. Anyone looks good on clean data.
• Field-level fill rate, not just record-level. A record "matches" if the email populates. That says nothing about whether job title, direct dial, or company size got filled. You'll find vendors who look 90% on the first and 40% on the second.
• How updates write back. If the enrichment lives only in the vendor's dashboard, it's not enrichment — it's a second list you now have to reconcile. It should push into your CRM without a monthly manual sync.

The vendor who lists every fee upfront — even when the top-line number looks higher — usually costs less in the end. Not always. But usually. The hidden-fee pricing models are the ones that get weird at renewal.

Scenario C: You're an agency or a high-volume outbound shop

This is the only scenario where unit economics genuinely drive the decision. At 50,000+ records a quarter, a $0.03 vs $0.05 spread is $4,000. That's real money, and it's worth sweating.

But here's the counterintuitive part: at this volume, buying the cheapest source is almost always the wrong move. We tried it. Cheaper sources tend to have lower match rates, and the SDR time to fish valid emails out of a low-quality list eats the savings by the end of the first week. The spend number looked great. The productivity number didn't.

The play at agency scale is a waterfall — cascade through multiple sources, stop at the first acceptable match, then layer in intent signals so you're not just blasting the whole list. The evaluation questions shift:

• How does the waterfall terminate? First-match or best-match? First-match is faster but usually worse. Ask the vendor to say which, explicitly, on the call. Half of them won't be able to.
• Intent signal recency and coverage. Intent data from 90 days ago is historical data, not intent. Ask what the freshness window actually is on their signal feed.
• API rate limits and webhook behavior. If you can't trigger enrichment from your own queue, you're going to build a fragile workaround anyway. Better to know that on day one.
• Human-in-the-loop review. At agency volume you can't eyeball everything, but you should be able to sample-review a batch before it goes out. If there's no preview step, that's a deal-breaker for me.

One more thing at this tier: don't skip verification. Sending to a 3% bounce rate across 50,000 sends a quarter will wreck your sending domain inside 60 days, and rebuilding reputation is a longer project than most agencies want to admit to clients. Verification is worth the extra half-cent per record, every time.

Also, remember that CAN-SPAM (enforced by the FTC) is not suspended for B2B lists. Per FTC guidance, unsubscribe requests must be honored within 10 business days, and commercial email rules apply whether the recipient is a person or a company inbox. Enrichment quality affects deliverability; list hygiene and honoring opt-outs affect whether you're still allowed to send at all.

This is the tier where agent-native tooling — I've been watching what okki-go and similar platforms are building — genuinely starts to make sense. But you still have to run the same evaluation questions first. The tool doesn't decide whether you need it. The scenario does.

How to figure out which scenario you're actually in

The labels above are useful, but the real question is which cost hurts you more: bought (money out the door on data you never use) or burned (SDR hours on data that was wrong). Whichever one stings more is where you should place your evaluation weight.

Ask yourself three questions. Answer honestly, not aspirationally.

1. How many records will you realistically send to in the next 30 days? Under 2,000 — you're Scenario A, no matter what your title says. Over 20,000 — you're C.
2. Does someone have "keeping the CRM clean" as an actual job responsibility? If yes, B. If it's a side-of-desk duty, you're A pretending to be B, and the evaluation criteria will fight you.
3. Is enrichment a fixed operating cost or a Q-by-Q experiment? Fixed → B or C. Experiment → A. Always A.

If I remember correctly, the last time I ignored question three was in Q1 of 2024. I signed a $9,000 annual contract to run a "small pilot." You can guess how that went.

The bottom line: pick your scenario first, then pick the metrics. Reverse the order and you'll build an RFP that scores vendors on values your team doesn't actually operate by. That's the sequence that produced my $14,200 mistake in the first place. The checklist is useful. The order is what saves you.