Bruised, Not Broken: Seven Questions to Ask When a Campaign Isn't Landing

Let's start with the thing nobody says out loud in the review meeting: three weeks is enough to know something is off. It is not enough to know what.

By week three you've got enough data to see that the trend line is wrong. You have not got enough to tell whether the audience is wrong, the creative is wrong, or the reporting is wrong. And yet the instinct — almost every time — is to reach for one of two levers: put more money in, or pull the whole thing.

Both are expensive. Neither is diagnostic. So here's the triage we'd run, in order, before a penny more goes out the door.

Question one: are these the right people?

Check who actually saw the thing, not who you asked for. Look at the delivery report, not the media plan. If your target was 35-54 ABC1 in the Midlands and you've bought a broad national reach build because it was efficient, you haven't got a creative problem — you've got a targeting problem wearing a creative costume.

The tell: strong engagement rates but terrible conversion, and an audience profile that doesn't match your customer base. Fix the targeting, leave the creative alone.

Question two: is the offer doing any work?

Not the creative. The offer. Is there an actual reason to act now, or are you asking people to feel warmly about a product with no prompt attached? Free delivery that isn't free. A "limited time" that's been running since March. A discount that undercuts your margin and still doesn't beat the supermarket own-brand price two feet away.

The tell: people click, people look, people leave. Interest without intent almost always points at the offer, not the ad.

Question three: can anyone repeat your message back?

Ring three people who saw it. Ask what the brand was selling. If the answers diverge wildly, the message is muddled — usually because two departments wrote it and nobody chose. Three propositions in one ad means zero propositions in the customer's head.

There's a red apple test for this, and it's blunt: if the message can't survive being described badly by a stranger, it was never clear enough.

Question four: is the creative actually getting seen?

This is where most "the campaign isn't working" conversations end up, and often rightly. But check the boring mechanics first: was the cutdown made for the format, or did the 30-second TV ad get squashed into a 6-second vertical? Is the first frame doing any work, or is it a logo reveal? Is the audio necessary?

The tell: high impressions, low three-second view rate. That's a creative cut-through problem, and no amount of extra budget fixes it — you'll just pay more for the same skip.

Question five: is the budget so thin it's invisible?

A £6,000 budget across six channels and four audiences is not a campaign. It's a rumour. Frequency matters: people need to see something more than twice before it registers, and splitting a modest pot across every platform on the deck guarantees nobody sees anything enough times for it to matter.

Concentrate. One channel, one audience, enough weight to be unmissable for a fortnight, then add the second channel. The tell is reach that looks respectable and frequency that's essentially zero.

Question six: did you flight it into a brick wall?

Timing has a habit of being nobody's job. Launching a payday-dependent offer on the 2nd of the month. Running a summer product in a wet fortnight. Overlapping with a category giant's big spend. Going live in the week everyone in your sector goes live.

Check the calendar before you check the creative. Sometimes the campaign is fine and the diary isn't.

Question seven: do you trust the number you're staring at?

Be careful here, and be unalarmed. Attribution is imperfect by design — cookie loss, walled gardens, view-through windows doing arithmetic you didn't ask for, a CRM that hasn't spoken to the ad platform since 2024. We've seen campaigns declared dead on the strength of a dashboard that was double-counting and under-reporting in the same week.

So before you act: reconcile platform-reported conversions against your actual order data. If the two don't roughly agree, the number is the problem, not the campaign.

Creative problem or media problem?

Here's the shortcut. Look at the funnel:

Plenty of impressions, low engagement — creative or targeting. Decent engagement, no conversion — offer or landing experience. Nobody saw it enough times — media weight or audience. The numbers contradict themselves — measurement.

One of those four is true. Work out which before anyone opens the budget spreadsheet again.

Refresh one element, or pull the whole thing?

Refresh, not rebuild, is the default. Swap the opening frame. Change the offer. Narrow the audience. Re-time the flight. One variable at a time, with a clean period of data after it, and you'll learn something.

Pull the whole thing only when two or three of the seven questions come back bad at once — when the audience, the message and the offer are all misfiring. At that point you don't have a campaign that needs tuning, you have a strategy that was never quite settled, and no amount of media spend will paper over it.

A bruised apple is still lunch. Most underperforming campaigns are a single fix away from working, and the discipline is finding that single fix before the money runs out.