I’ve been guilty of starting advertising campaigns and trusting the dashboard.

You spend money, the platform reports sales, and everything looks great.

But what happens when you start questioning whether those adverts are actually making a difference?

I’ve been there too.

Sometimes you need to do something drastic to find out.

Make a big enough change to learn something

If you really want to understand the impact of your advertising, turn it off for 30 days.

If that sounds too frightening, double your budget instead.

The point isn’t that cutting or increasing spend is automatically the right commercial decision. It’s that tiny adjustments often make it difficult to distinguish the impact from normal fluctuations.

Make a meaningful change, establish what you’re measuring, and give the experiment time to produce useful evidence.

And when you get that uncomfortable feeling that you’re leaving sales on the table, resist the temptation to interfere.

We cut a client’s advertising budget in half

I recently worked with a business that had been running Meta advertising for around five years.

This wasn’t a new account trying to find its feet. It was an established advertising operation with a substantial monthly budget.

We spent two weeks recording everything we could reasonably measure.

Website traffic, online orders, telephone enquiries, emails, footfall into the physical store. Every customer touchpoint we could track.

Then we cut the advertising budget in half.

And we left it alone.

The results were interesting.

Telephone enquiries dropped. Emails dropped. Shop visits dropped.

But online sales barely moved outside their normal variation.

If we’d only looked at online revenue, we might have concluded that half the advertising budget was being wasted.

That would have been the wrong conclusion.

Customers weren’t following the call to action

The adverts had been designed to generate online sales.

But a significant part of the audience wasn’t behaving the way the advertising team expected.

They were seeing the adverts, then choosing to telephone, email or visit the shop instead.

The advertising was influencing customers. It just wasn’t necessarily influencing the metric we were optimising for.

So we rebuilt the campaigns.

Some adverts were designed specifically to encourage offline enquiries and visits. Others were rebuilt to do a better job of converting customers online.

When we restored the original advertising budget, offline activity improved beyond the previous baseline.

Online sales also increased beyond the old baseline, with both improvements exceeding the variation we’d normally expect.

We hadn’t simply discovered whether the adverts worked.

We’d discovered how customers actually wanted to buy.

Your dashboard doesn’t understand your entire business

An advertising platform can report what happens within the information it receives.

It doesn’t automatically understand the customer who sees an advert, thinks about it, then walks into your shop the following morning.

Or the customer who sees a product online and telephones to ask a question before ordering.

If you’re only measuring online purchases, those customers can disappear from your assessment of advertising performance.

That doesn’t mean every untracked sale should be credited to advertising either.

It means you need to measure the business, not just the platform.

Don’t optimise for the wrong outcome

I wouldn’t recommend everyone immediately halving their advertising budget.

But I would challenge anyone spending serious money on advertising to ask what they’re actually trying to achieve.

If the goal is business growth, measure business growth.

If the adverts are generating enquiries, measure enquiries.

If they influence physical store visits, don’t ignore those visits because they don’t appear in your online conversion report.

And don’t assume a campaign is failing simply because customers aren’t clicking the button you wanted them to click.

The most valuable thing we learned wasn’t how much to spend. It was what the advertising was actually doing.

Sometimes your ads aren’t broken.

Your measurement is.