Formlyy Journal
CPL is going down but sales are not: how to fix the gap in 2026
May 21, 2026 · 9 min read · By Arthur Goudard
Seeing your CPL drop is always nice.
The dashboard becomes greener. The campaign seems to learn better. Cost per lead makes it look like acquisition is finally working in the right direction.
Then comes the weekly commercial point.
Sales are not increasing. Salespeople find weaker leads. The CRM is filling up, but the pipeline is not keeping up.
This is not uncommon. A falling CPL could be good news, or simply a sign that you're buying leads that are easier to generate, but less easy to convert.
Why a CPL can drop without creating more sales
The CPL measures the cost of a generated contact. It's useful, but incomplete.
A lead can cost less because:
- the audience is wider;
- the form requires less effort;
- the promise is more aggressive;
- the algorithm optimizes towards the easiest submissions;
- the leads are colder, less mature or less close to purchasing.
So the problem is not the CPL. The problem is reading it alone.
If you want a clear basis, the article on CPL: definition + calculation method + limits in 2026 sets the framework well. Here, we go further: what to do when the CPL falls, but the business does not follow?
The classic trap: optimize ease of conversion
Advertising platforms learn from the objective you give them.
If the main objective is the submission of a very simple form, they will look for people likely to submit this form. Not necessarily people likely to buy.
On Google Ads, lead forms can even send leads directly into the CRM via CSV, email or webhook, as explained in the lead form assets documentation. It's powerful. But the more fluid the entry, the more precise you have to be about what happens next.
A good system doesn't just ask: "How much does the lead cost?"
He also asks:
- how many leads actually respond;
- how many become SQL;
- how many book an appointment;
- how many show up;
- how many end up in opportunities or sales.
The CPL should open the analysis, not close it.
The 5 symptoms of a lying CPL
1. Volume increases, but sales reps call back less
When leads are too weak, dirty teams fail mentally. They call back less quickly, follow up less cleanly and end up treating the source as a secondary queue.
The problem becomes circular: less perceived quality, less effort, less sales.
2. The qualification rate drops
If the CPL goes down while the SQL rate goes down too, you're not really winning.
You just shift the cost to the sales team.
This is why cost per SQL often becomes more useful than CPL as soon as a company sells with a real decision cycle.
3. The cost per appointment rises silently
Let's imagine:
| Scenario | CPL | Leads | Qualified appointments | Cost per qualified appointment |
|---|---|---|---|---|
| Before | 45 euros | 100 | 20 | 225 euros |
| After | 25 euros | 180 | 12 | 375 euros |
The CPL drops. The actual cost of the appointment increases.
On the media dashboard, everything looks better. In business, everything is more expensive.
4. Business objections are changing
Prospects say more often:
- “I just wanted to see”;
- “I didn’t think I would be called back”;
- “it’s not for now”;
- “I don’t have the budget”;
- “I don’t really understand the offer”.
These are signals of discrepancy between the ad, the form, the intention and the follow-up.
5. The CRM fills up, but the pipeline remains flat
A full CRM is not a pipeline.
A pipeline begins when there are reliable statuses, next actions, dated opportunities, and reasonable business probability.
The method to correct without blindly reassembling the CPL
The wrong reaction is to say: “we’re going to add friction everywhere”.
Sometimes yes. But not always.
The goal is not to make the journey difficult. The goal is to put the right friction in the right place.
1. Separate leads, MQL, SQL and qualified appointments
Before touching campaigns, clean up the statuses.
A lead is not an MQL. An MQL is not an SQL. A SQL is not always a qualified appointment.
If everything is put in the same column, you will never know if the problem is with the audience, the message, the form, the callback or the sale.
2. Add 2 useful qualifying questions
Not ten.
Two real questions may suffice:
- the main need;
- the deadline or level of urgency;
- the current volume;
- the budget if the market accepts it;
- the size of the team or company.
The right question is the one that changes the next action. If it doesn't change anything, it is decorative.
3. Import offline conversions
Google recommends measuring objectives like qualified or converted leads in its offline imports and advanced lead conversion tracking. The Google Ads documentation on advanced lead conversion tracking emphasizes this point: CRM data can help with better attribution and optimization.
In short: don't just feed the algorithm with submitted forms.
Feed it with what really matters.
4. Track cost per step
Watch:
- CPL;
- cost per qualified lead;
- cost per SQL;
- cost per qualified appointment;
- cost per opportunity;
- cost per sale.
The right choice is not “low CPL or high CPL”. The right trade-off is: what cost produces the best final commercial return?
5. Reconcile sales feedback with data
Salespeople's feedback is valuable, but it must be structured.
Ask them to qualify the reasons:
- off target;
- no need;
- not the right timing;
- not reachable;
- false contact;
- no budget;
- good lead, but bad follow-up.
The article on lead quality details this logic: quality should not remain an opinion. It must become usable data.
Concrete example
An agency lowers the CPL Meta Ads from 42 to 24 euros.
On paper, it's clean.
But the client complains:
- fewer prospects respond;
- appointments booked are more fragile;
- salespeople spend more time sorting;
- the closing decreases.
The correction is not about cutting the campaign.
It consists of:
1. distinguish raw leads from contactable leads;
2. add a question of intention;
3. create a “qualified appointment” CRM event;
4. compare creations to the cost per appointment, not just to the CPL;
5. send weekly feedback to campaigns.
Often, the CPL rises a little. But the cost per sale is falling.
And this is where the subject gets serious.
What to remember
A low CPL is not a victory if sales do not follow.
It can even be dangerous, because it gives an illusion of performance.
Good reading in 2026 consists of linking:
- the acquisition cost;
- sales quality;
- processing speed;
- the appointment rate;
- the final revenue.
Your goal is not to generate the cheapest lead. Your goal is to generate the next client at the most controlled cost possible.
FAQ
Frequently asked questions
Is a low CPL always a bad sign?
No. A low CPL is excellent if it maintains a good qualification, appointment and sales rate. It becomes suspect when it falls while business indicators deteriorate.
Should the form be lengthened to improve quality?
Not necessarily. Above all, we must add questions that change business prioritization. Two well-chosen questions are often better than a cumbersome form.
What KPI to look at after the CPL?
The cost per SQL, the cost per qualified appointment and the cost per opportunity give a reading closer to business reality.
About the author
Arthur Goudard
My name is Arthur Goudard. I share what I see in the field when a marketing strategy needs to turn warm interest into a useful conversation, then into a clear appointment.
Sources
Keep reading
Read next