Stop Optimising for Cost per Lead. Optimise for Cost per Booked Lead.

Why does optimising for cost per lead tend to reward junk leads?

Cost per lead is calculated the same way regardless of what happens after the form is submitted: ad spend divided by number of form fills. That formula cannot distinguish between a lead who was ready to book and a lead who submitted a form out of curiosity, filled in a phone number that does not connect, or was never in the market at all. A campaign that broadens its targeting, loosens its qualifying questions, or simply casts a wider net will almost always lower cost per lead, because more people will fill in almost any form when the barrier to doing so is low enough. None of that has any relationship to how many of those leads actually turn into a paying customer. Optimising for the cheapest number of form fills is optimising for the wrong side of the funnel.

What should a business track instead, and where does it come from?

MAJ Leads' dashboard carries cost per qualified lead alongside the rest of a client's lead metrics. In V1 this is calculated against ad spend a client enters manually for each source, there is no live spend API pull yet, set against the number of leads from that source that were actually called and reached a qualified or booked outcome through the RNR cadence. The formula itself is simple, spend divided by qualified leads rather than spend divided by total leads, but the number only means something because it is anchored to outcomes that came from a real phone call, not from a form field alone.

What does the difference look like in practice?

The figures below are a worked example only, invented to illustrate the arithmetic. They are not a MAJ Leads benchmark, an industry average, or a claim about what any campaign should expect to spend or convert.

On cost per lead alone, Campaign A looks far more efficient, AED 50 against AED 125. Once the same spend is divided by qualified leads instead, Campaign B is actually the better spend, AED 417 per qualified lead against AED 500, because a larger share of its leads were genuinely in the market. A manager watching only the first column would keep scaling Campaign A and could easily cut Campaign B for looking expensive, the exact opposite of what the underlying quality of each lead source justifies. Again, these numbers are hypothetical, the point is the arithmetic, not the figures.

Is this just another way of describing return on ad spend?

It is related, but it answers an earlier question. Return on ad spend usually needs a closed sale value attached, which for a service business can take weeks to materialise, a signed contract, a completed treatment, a closed deal. Cost per qualified lead is available far sooner, as soon as a lead has been called and its outcome recorded, which for most sources is within hours rather than weeks. It will not replace a proper revenue-based ROAS figure once deals actually close, but it gives a much earlier read on whether a campaign is likely to earn a good ROAS at all, without waiting for the full sales cycle to play out. A campaign with a poor cost per qualified lead in its first few days rarely turns into a strong ROAS performer later, and a manager does not need to wait to find that out.

How does a business calculate this for its own campaigns?

1. Record ad spend per source or campaign for the period being measured, entered manually against that source on the dashboard.
2. Count leads from that source that reached a qualified or booked outcome inside the RNR cadence, not leads that merely submitted a form.
3. Divide spend by that qualified count, not by the total number of leads received.
4. Compare the resulting figure across sources and campaigns over the same period, rather than judging any single campaign's cost per qualified lead in isolation.

Why can a business not calculate this without a cadence that actually reaches every lead?

Cost per qualified lead is only honest if "qualified" reflects a real attempt to reach every lead, not just the leads a busy team happened to get around to calling. If half of a source's leads never get a call at all, the qualified count understates what that source could have produced, and the cost per qualified lead figure looks worse than the source actually deserves. This is the direct link back to speed to lead and the RNR cadence covered elsewhere in this cluster: a lead that is called and genuinely evaluated, whatever the outcome, produces a trustworthy qualified or not-qualified data point. A lead that sits uncalled produces nothing but a gap in the data.

What actually changes in ad bidding once conversions are booked-based rather than form-based?

This is where cost per qualified lead on the dashboard connects to the closed-loop reporting covered in our companion post on closed-loop attribution. Once Google Ads and Meta are receiving Qualified Lead and Booked events back from actual outcomes, rather than only ever seeing a form submission, their own bidding algorithms start weighting the audiences, placements and creatives that produced those outcomes more heavily. In practical terms, a campaign that was previously optimised purely to lower cost per lead can, once outcomes feed back in, start drifting toward a higher cost per lead alongside a lower cost per qualified lead, because the platform is now chasing a smaller, better-matching audience rather than the cheapest possible click. A rising cost per lead is not automatically a bad sign once this loop is closed, it depends entirely on what is happening to cost per qualified lead alongside it.

Is this just a way of making campaigns look better on paper?

The opposite. Cost per lead is the number that is easiest to make look good on paper, because it is disconnected from whether anything real happened after the click. Cost per qualified lead is harder to inflate artificially, because inflating it requires actually producing more qualified conversations, not simply loosening a form's fields or widening an audience. A business that switches to reporting cost per qualified lead internally, even before touching the ad platform's own bidding, will often find its most "efficient" campaign by the old metric is not its best campaign at all.

What does a business need in place to track this properly?

Ad spend entered per source, a cadence that actually attempts every lead so the qualified count reflects reality rather than a partial sample, and, where the goal is to change bidding itself rather than just reporting, the closed-loop reporting to Google Ads and Meta described in our companion post. See how leads are captured and attributed by source on the WhatsApp CRM page, compare running the calling cadence with AI Mode against Team Mode, and see current pricing, Team Mode starts from AED 299 per month, on the pricing page.