How to score a lead vendor

Every lead vendor will show you a sample file, a conversion story from a client they will not name, and a price per lead. None of those three things predicts whether you will make money. These four measures do.

Score a vendor on all four before you buy volume, and re-score quarterly — vendors drift, usually in the same direction.

1. Contact rate

The percentage of delivered leads you can actually reach within your normal follow-up window.

Ask for it by source and by age. A vendor quoting a blended contact rate across ten sources is hiding the worst three. Then measure it yourself in the pilot rather than accepting the number: contact rate is the single easiest metric for a vendor to quote optimistically, because their definition of "contacted" may include a voicemail.

What good looks like depends entirely on channel and vertical, so the useful benchmark is your own current source, not an industry figure. If a new vendor cannot beat what you already have, price is irrelevant.

2. Intent verification

How does the vendor know the person wanted to hear from you specifically?

• Ask to see the actual form, landing page, or script the lead came from — not a description of it. Read what the consumer read.
• Check whether your company was named at the point of consent, or whether the consumer agreed to be contacted by "partners". The second is a different product and carries a different risk.
• Ask for the timestamp, IP, and page URL to be delivered with every record. A vendor who cannot supply proof of consent per lead is selling you their compliance exposure along with the file.
• Ask what percentage of records are incentivised — sweepstakes, gift cards, prize draws. Incentivised leads are not worthless, but they are a different economic model and should be priced as one.

3. Disposition honesty

This is the measure most buyers never ask for, and it separates partners from brokers.

Send back your real dispositions — wrong number, not interested, never enquired, already a customer, duplicate — and see what the vendor does with them. A partner uses your dispositions to change what they send. A broker thanks you for the feedback and sends the same file next month.

Test it deliberately. In month two, count how many records you rejected in month one that reappear in a new batch. Anything above a trivial percentage tells you the feedback loop is decorative.

4. Return-rate handling

Read the return policy before the pricing. The three things that matter:

• The window. A 24-hour return window on a lead you cannot reach for three days is not a return policy.
• The definition. Which rejection reasons are accepted, and who adjudicates a dispute?
• The remedy. Credit, replacement, or refund — and whether replacements come from the same source that failed you, which simply moves the problem to next month.

A vendor confident in their file will give you a workable return window without being pushed. A vendor who negotiates hard on returns is telling you what they expect the quality to be.

How to run the scorecard

Buy a small, deliberately unremarkable test volume — enough to be statistically meaningful for your close rate, not enough to hurt. Run it through your normal process with no special handling, because special handling produces a result you cannot reproduce at scale.

Then score cost per qualified conversation, not cost per lead. A cheaper lead that takes three times the dials to reach is a more expensive customer, and the price per lead never tells you that.