Back to all questions
Google Ads Strategy
At £40 a click, how do you test anything at all?
N
Neil Ashworth
Asked Jul 15, 2025
3
1596
Personal injury. Our average CPC is around £40 and the good keywords are north of £60. A
hundred clicks is four thousand pounds, and a hundred clicks is not a test, it is barely a
hint.
Every piece of testing advice assumes you can afford to be wrong for a while. Run the
experiment, wait for significance, keep the winner. At our costs the experiment itself is a
line item the partners ask about.
So what do people in expensive verticals actually do? Specifically:
Do you test landing pages rather than ads, on the grounds that the traffic is already paid
for by the time it reaches the page and the test is therefore free?
Do you accept lower confidence and make decisions on data that would not pass anywhere else?
Or is the honest answer that at this CPC you copy what worked for someone else and only test
the things that are cheap to be wrong about?
Genuinely open to being told my framing is wrong. I have been doing this for three years and
I still do not have a testing process I would defend.
Replies (3)
P
Priya Deshmukh
·
Jul 15, 2025
Your instinct about landing pages is right and it's the single biggest lever available to
you. The click is a sunk cost by the time they land, so a page test costs you nothing extra
and moves the number that matters most at high CPCs, which is conversion rate rather than
cost per click.
On confidence: yes, you accept less, and you should be explicit about it rather than
pretending otherwise. Decide in advance what you would do at 80 percent confidence versus
95, write it down, and stick to it. The failure mode in expensive accounts is not making
decisions on thin data - it is making them on thin data while telling yourself the data is
thick.
One thing to add to your list: test with the cheap traffic and apply to the expensive.
Broad-intent keywords in your vertical are a fraction of the cost of the high-intent ones.
Ad copy angles, page layouts, form length - learn those where clicks cost a fiver, then
deploy on the £60 terms.
G
Greg Pallister
·
Jul 16, 2025
The third option in your list is the correct one and nobody wants to say it out loud, so I
will.
In a £40 CPC vertical the majority of your decisions should be judgement informed by
experience, not tests. You don't have the volume, you'll never have the volume, and
pretending otherwise produces a lot of significance calculations performed on noise...
What you can test cheaply: pages, forms, call handling, follow-up speed. What you cannot
afford to test: match types, bidding strategies, campaign structure. For those, copy the
consensus and spend your attention elsewhere.
M
Marta Silva
·
Jul 17, 2025
Worth adding that at your CPCs the biggest wins are usually not in the account at all. We
had a client in a similar bracket where response time to a form fill went from four hours
to twenty minutes and the effective cost per case dropped by a third, with nothing changed
in Ads.
Depressing if you're the ads person. Excellent if you are the client.