Case study · IT services
Case study: how a managed services provider halved its cost per qualified lead by making leads more expensive
Cost per lead doubled. Cost per qualified lead more than halved.
A forty-person managed services provider was generating leads at 16 pounds and converting eight per cent of them to a qualified conversation. After changing the conversion event, the offer and the form, leads cost 34 pounds and thirty-nine per cent qualified. Cost per qualified lead fell from about 200 pounds to about 87, and closed deals went from two to eleven.
By Uzair, Co-founder · Updated
4,000 pounds
Monthly spend
Unchanged across the period
34 pounds
Cost per lead
Up from 16 pounds
39 per cent
Leads qualifying
Up from 8 per cent
11
Closed deals
Up from 2
Context
What was the starting position?
A managed services provider of about forty staff, spending roughly four thousand pounds a month on Meta, and pleased with the reporting until anyone looked past it.
Cost per lead was sixteen pounds, which their previous agency had presented as a strong result, and by the standard of the number itself it was. The dashboard showed lead volume rising quarter on quarter and a cost per lead below anything the sales director had seen from other channels.
The problem appeared in the CRM. Of the leads arriving, about eight per cent reached a conversation the sales team judged worth having. Over the first six months of that arrangement, two deals closed.
So the account was working exactly as instructed and the instruction was wrong. It had been told to find people who submit forms, and it had become very good at it.
Objective
What was actually wrong?
Four things, and they were the four the diagnostic on this site lists, which is the unglamorous truth of most of this work.
The campaign optimised for the standard lead event, so the platform was selecting for form-fill behaviour. The form was an instant form with every field prefilled, so submitting took two taps and required no thought. There were no qualification questions at all. And the creative advertised managed IT support, a category, in language every competitor in the county was also using.
None of those is exotic and none required an audit to find. They were visible in the account within an hour.
What made them persist was that the reporting looked good. A sixteen pound cost per lead is genuinely hard to argue with unless somebody connects it to the CRM, and nobody had.
What changed
What changed?
Three changes over about six weeks, deliberately sequenced rather than shipped together, so each one could be read.
First the conversion event moved from the form fill to a booked call, and closed deals started going back to Meta as offline conversions so the platform could learn which leads had been worth having. Second the instant form was replaced with a landing page carrying three qualification questions: staff count, whether an incumbent was under contract and when it ended, and what had prompted them to look.
Third the creative stopped advertising the category and started naming one trigger. The winning line addressed companies whose contract with an incumbent was coming up for renewal, which is a situation rather than a service.
Nothing about targeting changed. The audience settings at the end were the same as at the start.
Sales outcome
What happened to the numbers?
Cost per lead roughly doubled, which was the intended outcome and still took some explaining.
Leads went from sixteen pounds to thirty-four. Volume fell by a little over half. Both of those look like a campaign getting worse, and for about five weeks the internal conversation was uncomfortable, because the metric everyone had been watching for a year was moving the wrong way.
The qualification rate went from eight per cent to thirty-nine. Cost per qualified lead fell from about two hundred pounds to about eighty-seven. In the six months following the changes the company closed eleven deals against two in the six months before.
Spend did not increase. The same four thousand pounds a month produced a different population, because the platform had been given a different instruction and better information about which leads mattered.
34 pounds
Cost per lead
Before: 16 pounds
39 per cent
Leads qualifying
Before: 8 per cent
about 87 pounds
Cost per qualified lead
Before: about 200 pounds
11
Closed deals in the period
Before: 2
4,000 pounds
Monthly spend
Before: 4,000 pounds
What we got wrong
Not everything worked.
Two things worth recording, because a case study that reports only the parts that worked is an advertisement.
We shipped the landing page and the new creative in the same week, which meant the first four weeks of data could not separate the effect of the form from the effect of the message. If the combination had failed we would not have known which half to revert. It worked, which is luck rather than method, and we sequenced the later changes properly because of it.
We also underestimated how hard the reporting change would be internally. Telling a sales director that the cost per lead is about to double is a conversation that needs to happen before the change, with the cost per qualified lead already agreed as the number that matters. We had that conversation two weeks late and it cost us credibility we did not need to spend.
The account recovered. The relationship took longer.
What cannot be attributed
What else changed in the same period.
More than the numbers above imply, and the honest version of this study has to say so.
The company hired a second salesperson in the same quarter, which almost certainly improved follow-up speed independently of anything we did. Their close rate from a qualified conversation also rose, and we have no way to separate a better-qualified prospect from a better salesperson talking to them.
The eleven closed deals are from their CRM and are real. The claim that all eleven exist because of the ad account is not one we can support, and we are not making it.
What we would claim is narrower: the same spend produced a population of leads the sales team judged worth talking to at roughly five times the previous rate, and that change tracks the campaign changes in time. The revenue conclusion is theirs to draw, not ours.
Want the same channel for your business?
We build paid acquisition systems for companies that need predictable, qualified pipeline rather than cheaper leads. If that is the problem you are solving, let us talk.