Log inRequest access

Poor lead quality

Criticallead_quality_poor6 min read

In shortLead quality is not visible in any advertising metric — it only exists once somebody has called the leads and said which were real, and the number that matters afterwards is the cost per qualified lead, which is often several times the headline figure everyone has been optimising against.

Every guide to this problem will give you the same four causes: bots, forms that are too easy to fill in, the Audience Network, and not feeding conversions back to the algorithm. They are all real. They are also all answers to a question nobody has actually established yet, which is whether your leads are bad and by how much.

Lead quality is not an advertising metric. It does not appear in Ads Manager because Meta cannot see it. It exists in one place only: in whether somebody picked up the phone and the person on the other end was worth talking to.

The complaint nobody can check

“The leads are rubbish” is the most common thing a sales team says about marketing, and in its usual form it is unfalsifiable. How many were called? Of those, how many were real? What would have counted as good?

Without answers, the complaint cannot be acted on and cannot be dismissed. So the campaign keeps running because the cost per lead looks fine, the sales team keeps not calling because the leads are rubbish, and both positions are defensible from the evidence each side has.

A dashboard can tell you what a lead cost. Only a person who called one can tell you what it was worth.

The number that matters

Once leads have been checked, the useful figure is not cost per lead. It is cost per lead that was actually real — spend divided by the qualified ones only, with the rest treated as what they were, which is nothing.

The arithmetic is unkind. A campaign at a comfortable price where one lead in four is genuine is really costing four times what the report says. That is frequently the difference between a campaign that looks like the best on the account and one that is quietly the worst.

This is also why a cost per lead that suddenly falls is a warning rather than a win. Cheaper people are easier to reach precisely because nobody bidding properly wants them.

Before you blame the form

The standard fixes — harder forms, qualifying questions, excluding the Audience Network — all work on the assumption that the wrong kind of person is seeing the right ad. Often the simpler explanation is that the right kind of person is nowhere near you.

Geography is the cause almost nobody lists. A radius reaching forty kilometres past the catchment produces people who will fill in a form and never visit, at a price that looks excellent because there is no competition for them. Check where the money actually landed before rewriting the form.

What to do

  • Review a fixed sample and record a verdict per lead. Twenty calls with an outcome written down beats a thousand leads and an opinion.
  • Agree what qualified means before you start, in writing. Otherwise the standard moves to fit whoever is arguing.
  • Recompute cost per qualified lead and compare it against what this account normally pays, not against an industry figure.
  • Feed the verdicts back. An algorithm told which leads were real optimises toward people like them; told nothing, it optimises toward whoever fills in forms.

When Admetriq raises it

The actual rule, so you can hold this page against what you see rather than take its word for it.

  • This rule will not fire on advertising data alone. It requires leads to have been reviewed and marked qualified or not, because nothing in spend, clicks or cost per lead can tell you whether a person was worth calling.
  • Once that verdict exists, the raw price is ignored entirely. If none of the reviewed leads qualified, the finding says so in those terms: this much money bought nothing worth calling back, whatever the headline cost says.
  • Otherwise it divides spend by the qualified leads only. Where that figure lands above what the campaign was meant to beat — the client’s own band ceiling, or 1.5 times their median — it is raised as critical.
  • There is an opposite finding for the same calculation. When the qualified cost holds up, it is reported as a confirmation rather than silence, because a campaign proven to produce real leads at a real price is the one you want to scale.

Most often mistaken for

Cost per lead above normalcpl_above_normal

A high cost per lead is an advertising problem: you are paying too much for the same thing. This is a different problem wearing a better number — the price looks fine, or unusually good, and the leads are worthless. They can point in opposite directions, and the cheap-looking campaign is the more dangerous of the two because nothing on the dashboard asks you to look at it.

All findings · Longer writing