The dashboard shows cost per lead. Your sales team experiences lead quality. Those two things are measured by different people, reported in different meetings, and almost never compared, which is why campaigns can look successful for months while quietly wasting everybody’s time.
Cheap leads are easy to buy. Cheap leads that actually close are the whole job.
This is for you if your ad reporting looks fine and the people answering the phone disagree.
What lead quality actually means
It is the share of enquiries that turn into something real.
Two numbers cover it for most businesses. What percentage of leads become qualified conversations, meaning someone with the need, the budget and the authority to proceed. And what percentage become customers.
That is it. You do not need a scoring model to start, and starting without one is better than waiting six months for a system nobody uses.
The reason this goes unmeasured is structural rather than lazy. Your cost figure arrives automatically in a dashboard. The quality figure requires somebody to go and count, which means it only exists if a person decides it should.
Why cheap leads are often the expensive ones
Cost per lead is a spending measure, not a quality one, and it can always be improved by lowering standards.
Broaden the targeting, shorten the form, soften the offer, and your figure falls within days. More people respond because you asked less of them. The report looks excellent and the sales team spends the month on people who were curious rather than committed.
There is a mechanical reason behind it too. If your campaigns optimise towards form submissions, the platform builds a model of people likely to submit forms and finds more of them. Willingness to fill in a form and willingness to buy are correlated, not identical, and the gap between them is where poor lead quality lives.
The number that fixes this: cost per qualified lead
One metric replaces both and settles most arguments.
Take your spend, divide by qualified leads rather than by all leads. Suddenly a channel producing cheap junk and a channel producing expensive prospects can be compared honestly.
A worked example. Two sources at $30 a lead. One qualifies 60% of them, so its real cost is $50. The other qualifies 15%, so its real cost is $200. Identical on the dashboard, four times apart where it counts.
That single calculation reverses budget decisions regularly, and it takes an afternoon to set up once you can identify which enquiries qualified.
Lead quality differs by channel, predictably
Where an enquiry came from tells you a lot about how likely it is to be worth having.
Ruler Analytics, working from 110 million sessions and £33.8 million of tracked spend across 13 industries, measured paid search converting at 5.4% against 2.11% for paid social. Roughly two and a half times.
That is not a criticism of social. It reflects intent. Somebody who typed their problem into a search box is further along than somebody who saw you between holiday photos, and no amount of targeting closes that gap entirely.
The practical consequence is that a single quality expectation across channels will always make one of them look broken. Set a separate expectation per channel, judge each against its own history, and compare them only on cost per qualified lead.
When high cost is correct lead quality
Worth stating because the reverse mistake is just as common.
The best example on my books is 906 leads at $91.18 each for a US automobile accessories business on Google, running roughly 80% of the spend through Search and the rest through Performance Max. Against any benchmark that is expensive. Against a market where one closed order is worth thousands, it was excellent.
A $91 lead is not expensive if it is worth $4,000. A $9 lead is expensive if it never answers the phone.
The point is not that dear leads are better. It is that price on its own tells you nothing, and a figure only becomes meaningful next to the qualification rate and the value behind it.
How to improve lead quality, in order
Signal first. Send closed-won data back to the ad platform so it learns which enquiries became customers, then optimises towards those people rather than towards form fillers. Nothing else comes close to this, and most accounts have not done it.
Then the offer. Vague offers attract vague interest. “Free consultation” gets you people collecting free consultations. Something specific about what happens and who it suits filters at the top of the funnel, which is the cheapest place to filter.
Then one qualifying question on the form. One, not four. It should be the question your sales team asks first, and it will cost you volume, which is the point.
Then follow-up speed, because a good lead badly handled becomes a bad lead. Quality is partly created after the enquiry arrives, not just before.
Filter at the form, not after it
This is where I differ from a lot of the advice, and it is worth being direct about.
I filter before submission and almost never after. That means qualifier questions on the form, and for anything running through Meta instant forms, a two-step phone verification so the number has to be real before the enquiry counts.
What I do not do is sit between the client and their enquiries, marking some as junk after the fact. Those go straight through.
Two reasons. Filtering after submission is judgement applied to somebody else’s business, and I am not the person who knows which odd-sounding enquiry turns out to be a customer. And a lead you reject after it has already counted as a conversion has still taught the ad platform to go and find more like it. The damage was done at the moment it counted, not at the moment you binned it.
Filtering at the form stops both problems at once. Fewer enquiries arrive, they arrive already checked, and the platform only ever learns from the ones that passed.
The form format decides how much of this you have to do
Worth knowing before you pick one, because the two common formats fail in opposite directions.
A form on your own website produces very little spam and a high cost per lead. People had to arrive, read something, and type their details in.
Meta instant forms produce the cheaper leads and more of the junk, because the fields arrive pre-filled and somebody can submit in two taps without fully registering that they did. Lower cost, lower quality, more spam.
Neither is wrong. They are different trades, and the instant form trade is only a good one if you close the gap deliberately, with a qualifier question and phone verification, rather than hoping the sales team absorbs it.
What lead scoring is for
Scoring is useful when it changes behaviour and decorative when it does not.
If a score decides who gets called first, or which enquiries go to a senior person, it earns its place. If it populates a field nobody looks at, it is admin.
Start with two tiers rather than a hundred-point model. Worth calling now, and worth calling eventually. Most businesses never need more than that, and the ones that do will know because the simple version stopped being enough.
Whose problem this is
Lead quality falls between two chairs in most businesses.
Marketing owns cost and reports on it monthly. Sales experiences quality and complains about it informally. There is no shared number, so the conversation becomes a matter of opinion and whoever is more senior wins.
The fix is one accountable metric: cost per qualified lead, owned by whoever buys the traffic, using qualification data from whoever answers the phone. It forces the two halves of the argument into the same sentence.
If nobody owns it, the account will drift towards cheap leads every time, because that is the number being watched.
What to change this week
Four steps.
Count the last 90 days of leads and mark which ones became qualified conversations. Do it by hand if you have to, since the first number matters more than the system that produces it.
Work out cost per qualified lead by source, and compare that ranking against your cost per lead ranking. Where they disagree, the second one is wrong.
Check whether closed-won data flows back to your ad platforms. If it does not, that is the highest-value thing on this list.
Then agree who owns the qualified figure, so it appears in a report rather than in a corridor.
The wider piece on judging lead cost covers the ceiling, working out what a lead is worth gives you the value behind the quality, why your close rate sets the target covers the same idea from the other end, and fixing Performance Max lead quality is the most common place this goes wrong in a Google account. For what the numbers look like across real accounts, the case studies lay them out. To have your own reviewed, get in touch.