Reduce Cost Per Lead

Work out why your cost per lead is high before you try to lower it

A high cost per lead has four common causes and one uncomfortable fifth: sometimes the CPL is correct and the problem is what happens after the lead arrives. Each cause has a different fix, and three of them are not inside the ad account.

Book a Strategy Call See how we work

Start here

Cost per lead is a ratio. You can only move it by changing what you spend or by changing how many leads that spend produces — and most accounts have spent a year attacking the wrong half.

Almost all CPL work happens inside the ad account, because that is where the person looking at the number spends their day. But the denominator is decided by the landing page, the offer, the form and what you count as a conversion. Three of those sit outside the platform entirely.

So the first job is not optimisation. It is working out which of the five causes below is producing your number, because the fix for each one is in a different place and doing the wrong one costs a quarter.

The diagnosis

Five things a high cost per lead usually means

You are paying for the wrong traffic.: The search terms report almost always contains a surprise. Job seekers, students, suppliers, people looking for free versions, and searches for a service you do not offer. Under broad match and automated campaign types this traffic often never appears as a keyword you chose, which is why the account can look tidy while a meaningful share of the budget goes to people who were never going to buy.
The landing page leaks.: The traffic is right and the page loses it. Usually some combination of a slow mobile load, a headline that does not match the ad, a form asking for information nobody will give a stranger, an unclear offer, or no phone number for the large share of people who would rather call. This is arithmetic: halve the leak and the cost per lead halves at the same spend.
You are counting the wrong events.: Newsletter signups, PDF downloads and phone clicks that lasted three seconds all counted as conversions. Or the reverse: a form that stopped firing its event during a site update, so real enquiries are arriving and not being counted. Both distort CPL badly, and both cause the automated bidding to optimise toward the wrong thing, which makes it worse over time.
The market is genuinely expensive.: Some categories have clearing prices that are simply high — legal, insurance, cosmetic procedures, enterprise software. If well-funded competitors can pay more for a customer than you can, no amount of account optimisation closes that gap. The honest answer is to change the mix: organic, referral, partnerships, retention, or a narrower niche where you are the obvious choice.
The cost per lead is fine and the close rate is not.: This is the one nobody wants to hear. If leads are being contacted a day later, or by whoever is free, or with no follow-up sequence, then a perfectly reasonable CPL will look catastrophic by the time it reaches revenue. We check speed to first contact and contact attempts before we recommend a single change to the account.

Before touching a bid

The checks we run before changing anything

Most of this takes a day or two and it decides everything that follows. Skipping it is how accounts end up optimised confidently in the wrong direction.

Read the full search terms report, not the keyword report, and categorise where the money actually went
List every conversion action being counted and decide which ones represent a real commercial enquiry
Verify each event still fires, on mobile as well as desktop, after any recent site change
Check whether phone calls are tracked at all, and whether short calls are being counted as leads
Measure landing page load performance on a mid-range phone on mobile data, not on office wifi
Count the form fields and ask which of them a stranger would genuinely complete
Compare the ad promise to the page headline word for word
Segment cost per lead by campaign, device and match type before drawing any conclusion from the average
Establish average deal value, gross margin and close rate so a maximum viable cost per lead can be calculated
Check how quickly leads are contacted and how many times, before blaming the traffic

Matching cause to fix

Same symptom, five different places to work

This is the table we use to decide where the effort goes. The last column matters most: it is what happens when you treat the wrong cause.

Cost per lead causes, where each one is fixed, and the cost of misdiagnosis
Dimension	How you spot it	Where the fix lives	What happens if you get it wrong
Wrong traffic	Search terms report full of irrelevant intent	Campaign structure, match types, negatives, audience signals	You cut budget instead, and lose good traffic alongside the bad
Leaking page	Good traffic, poor conversion rate, slow mobile load	Landing page, offer, form design, page speed	You raise bids to buy back the volume the page is losing
Miscounted events	Platform conversions do not match enquiries the team received	Analytics and conversion tracking configuration	Automated bidding optimises toward a fictional signal for months
Expensive market	Everything is configured well and the number is still high	Channel mix, positioning, niche, retention	You spend a year optimising an account that was already near its floor
Close rate	Reasonable cost per lead, poor cost per customer	Sales response time, follow-up sequence, qualification	You blame marketing and cut the spend that was working

A worked example

What the search terms report usually shows

These are illustrative searches from a commercial heating and cooling contractor, grouped by what the searcher is actually doing. The shape holds across most lead-generation accounts even though the words change.

Expensive per click, cheapest per customer

High cost per click, high intent, and usually the lowest cost per acquired customer in the account. These are the terms most often paused first to protect a cost per lead target.

emergency ac repair near me
commercial hvac contractor
ac repair open now
rooftop unit replacement quote

Cheap per click, expensive per customer

Cheap clicks and occasional cheap form fills. They flatter the reported number and rarely produce a job. Useful as organic content, rarely worth paying for.

how does a heat pump work
ac not cooling troubleshooting
best hvac brands
furnace filter size chart

Spend you did not choose

Job seekers, trade buyers and students. Under broad match and automated campaigns this traffic arrives without ever appearing in a keyword report, and it is often a meaningful share of the budget.

hvac technician salary
hvac jobs hiring
hvac parts wholesale
hvac apprenticeship

These are examples of how customers in this market search, drawn from keyword research and from the questions that come up on sales calls. They are illustrative, not a volume claim — the actual demand in your area is something we size before recommending anything.

How we work on this

What we will and will not do to a cost per lead number

It is easy to make this metric improve. It is harder to make the business better off, and the two are not the same thing.

What we do

Establish a maximum viable cost per lead from your margins before recommending any target
Report cost per acquired customer alongside cost per lead wherever the data allows it
Tell you when the account is close to its floor and the answer is a different channel mix
Separate traffic problems, page problems and tracking problems before proposing work
Check sales response time and follow-up before concluding the leads are poor

What we will not do

Improve the number by counting newsletter signups and PDF downloads as leads
Pause your highest-intent searches because they have the highest cost per click
Promise a specific cost per lead or a percentage reduction we do not control
Move budget to a cheaper channel that produces cheaper leads and fewer customers
Run tests on a page without enough traffic to produce a result worth acting on

Questions

Cost per lead questions worth answering properly

What is a good cost per lead?

There is no useful benchmark across industries, because a lead worth £200 and a lead worth £40,000 are not comparable. The only number that matters is your cost per acquired customer against the value of that customer over their lifetime.

Work it out in that order: average deal value, gross margin, close rate from lead to customer, then the maximum you can pay for a lead and still make money. Many businesses discover the CPL they have been complaining about is comfortably profitable and the real problem is close rate.

Is cost per lead even the right metric?

It is a useful operational metric and a poor strategic one. It tells you whether an account is drifting; it tells you almost nothing about whether the marketing is working, because it is blind to what the leads are worth.

If you can get sales outcome data back into the ad platform, do that instead and optimise toward acquired customers or revenue. If you genuinely cannot, at minimum split CPL by campaign and by the type of enquiry so you can see which cheap leads are cheap for a reason.

Why did our cost per lead go up when we did nothing differently?

Usually one of four things: a new competitor bidding aggressively, seasonal demand shifting, a landing page change made by someone else in the business, or a tracking change that stopped counting something it used to count.

Check tracking first, because it is the fastest to rule out and the most common. A conversion action that broke during a site update will make CPL rise overnight with no change in the underlying performance at all.

Should we just pause the expensive keywords?

Usually not. In most lead-generation accounts the highest cost-per-click searches carry the highest intent and produce the lowest cost per acquired customer. Pausing them improves the reported CPL and quietly raises the cost of actually winning a customer.

The right test is cost per customer, not cost per lead. If you cannot measure that yet, that is the first thing to fix — otherwise every optimisation decision is being made with the wrong number.

How much can conversion rate work realistically change this?

Cost per lead is spend divided by leads, so doubling the conversion rate of the page halves the cost per lead at identical spend. That is arithmetic rather than a claim, and it is why the page is often a better place to start than the account.

Whether your page has that much headroom depends on where it starts. A page already converting well has little room; a page with a nine-field form, no clear offer and a four-second mobile load usually has a great deal.

Get a diagnosis before you approve another optimisation round

Bring your account, your landing page and your last three months of enquiries. We will tell you which of the five causes you have, and whether the target you are being held to is achievable in your market.

Book a Strategy Call

If we don't deliver the work we agreed to deliver for reasons within our control, you don't pay for the undelivered work. Read our guarantee

Where to go next

Last updated September 7, 2026 · Reviewed by Zubair Afzal