What to do when a marketing campaign fails

Start by separating the decision from the outcome. They aren't the same thing, and teams that confuse them get more cautious without getting any better. A sound decision can produce a bad result, and a reckless one can get away with it. The review's job is to work out which of those you had, and sometimes the honest answer is that you made a reasonable call and it didn't land.

Then find out which part broke, because "the campaign failed" is never specific enough to act on.

Find the part that broke

Work along the chain in order and stop at the first place the numbers fall off a cliff. It's almost always one link rather than the whole thing.

1. Did it reach anybody? If the impressions or sends were far below plan, nothing after this point tells you anything. The failure was distribution, or budget, or a campaign that never properly started.
2. Did anybody respond to it? If it reached plenty of people and nobody clicked, the message or the offer is wrong, or it reached the wrong people.
3. Did the page do its job? Clicks arriving with nothing happening afterwards point at the landing page, and the usual suspects are the ones in our guide to landing page structure.
4. Did the enquiries turn into anything? If enquiries arrived and nothing closed, that isn't a marketing failure at all, or not only one. Find out what happened to them before you rebuild anything.
5. Was it even measured properly? Check this before you conclude anything, because broken tracking produces exactly the shape of a failed campaign. The two halves people most often miss are the five GA4 reports worth reading and counting the phone calls.

Then ask the three awkward questions

Was the target ever realistic? Where did the number come from? If it was picked to get the budget approved, the campaign didn't fail, the forecast did, and that's a different conversation with a different person.

Did it run as designed? Late sign-off, a shortened run, half the budget, the wrong creative, a landing page that went live a week late: more campaigns are compromised than are wrong, and a compromised campaign tells you nothing about the idea.

Would we have called it a success at a lower number? If nobody agreed beforehand what success looked like, you can't review it now. Agree it next time, in the brief, the way we describe in briefing an agency.

Run the review properly

Hold it within two weeks, while people still remember, and keep to these rules:

• The person who ran it doesn't present the verdict. They present what happened, and the verdict belongs to the group.
• No blame, and no comfort either. "These things happen" ends the conversation as surely as blame does.
• One page comes out of it, saying what you now believe, what you'd do differently and what you're changing as a result.
• One change, not seven. Teams that leave a bad review with a long list of changes tend to make none of them.

Decide what to do with the campaign itself

You have four options, and the second is the one people skip:

• Stop it. That's right when the chain broke somewhere you can't fix, or the audience is wrong.
• Fix one link and run it again. It's usually the cheapest useful thing to do, and teams avoid it because running a failed campaign a second time feels like an admission. If the message worked and the page didn't, you already have most of a working campaign.
• Shrink it and keep testing, when you believe in it but can't prove it yet.
• Shelve it and write down why, so that whoever proposes it again in two years finds your note.

Write the lesson where it will be found

Put it somewhere a person who joins next year will find it, not in a chat thread. The useful record is short: what we thought would happen, what did happen, what we think explains the difference, and what we'd do differently. That record is the whole point of a review.

The tempting mistake afterwards is to become more cautious, which looks like learning but isn't. A team that only runs safe campaigns gets safe results, and safe results are usually just the same results again. The change worth making is better evidence before you commit, not smaller ambitions afterwards.

If you'd rather someone from outside ran the review, which is often easier than running it on yourself, that's the kind of day our fractional and part-time marketing leadership covers.