A marketing campaign nobody can read afterwards costs you the money and teaches you nothing. Structure makes the difference, and you decide it before anything runs.
Start with one sentence
Write what the campaign is for, in a sentence naming the outcome and the period. Not "raise awareness". Something you could later say happened or did not.
That sentence prevents the most common failure, which is judging afterwards against whatever number happened to look best. Deciding the goal first turns the review of a marketing campaign into a comparison rather than a negotiation.
Keep the sentence somewhere visible. It is surprisingly easy, six weeks in, to forget what the campaign was originally supposed to do.
How a marketing campaign should be built
Five decisions, in this order, because each depends on the one before.
- The objective, stated as the outcome you actually want.
- Who it is aimed at, described in a sentence.
- A budget and a period, both settled before launch.
- Your message, aimed at that audience and that objective.
- Somewhere to send people, continuing the message rather than restarting it.
Skipping straight to the message is normal and it is why so many campaigns feel busy and produce nothing readable.
The objective is also where a marketing campaign either does or does not commit to a result you can measure. That choice happens here rather than in the reporting.
How many things a marketing campaign should run at once
Fewer than feels productive. Each variation needs enough volume to produce a result you can trust.
Splitting a modest budget across nine audiences produces nine numbers too small to act on, which looks thorough and teaches you nothing. Two or three well fed variations beat nine starved ones every time.
Where the budget is genuinely small, run one thing properly and test the next thing afterwards. A marketing campaign run in sequence still teaches you something. One run in parallel on a small budget teaches you nothing at all.
Setting the marketing campaign period before you start
Judging a marketing campaign in week two usually means judging the setup rather than the performance.
So fix the review date when you fix the budget. Search campaigns can often be read within a month. Anything interruptive, and anything with a long sales cycle, needs longer. Write the date down, and resist the urge to intervene before it, because every intervention resets what you were measuring.
Testing inside a marketing campaign
Change one variable at a time, and decide in advance what result would make you keep the change.
Creative and destination usually move results most, so test those before bidding settings. Meanwhile running two tests at once inside one marketing campaign is only sensible where they cannot affect each other, which is rarer than people assume.
Record what you tried and what happened. Campaigns repeat, and a note from last year is cheaper than rediscovering the same answer.
Budget, and what to do when it is working
Increase gradually rather than doubling. A sudden change resets whatever the platform had learned and makes the before and after incomparable.
Expect efficiency to fall as you scale, because you are reaching further from your best audience. That is normal, and the question is whether the extra volume is worth the extra cost per result, which is a business decision rather than a marketing one.
Agree who makes that call before the marketing campaign reaches the point of needing it. Scaling decisions taken in a hurry tend to get reversed a month later, which costs more than either option would have.
When a marketing campaign is not working
Work down the chain rather than reaching for the budget.
Poor click rate points at the message or the audience. Good clicks with poor conversion point at the destination. Good conversion with poor quality points at who you are reaching. Rising costs with everything else steady usually point at competition rather than at any fault of yours.
Each has a different fix, and applying the wrong one is why campaigns get rebuilt repeatedly without improving. Diagnose before rebuilding, because a marketing campaign started again from scratch also throws away everything it had learned.
Knowing when to stop a marketing campaign
Stopping is a legitimate result and it is cheaper than continuing out of hope.
Stop when the period is over and the objective was missed, after a fair attempt at the obvious fixes. Then write down what you learned before switching it off, because that note is the actual return on a marketing campaign that did not work.
Say so plainly when you report it. A campaign quietly left running at a reduced budget costs money every month and settles nothing, and everybody involved already suspects the answer.
The review that makes the next one better
Half an hour, three questions: did we hit the sentence we wrote at the start, what would we change, and what will we keep.
Do it while the campaign is fresh. A review delayed by a month becomes a summary, and summaries do not improve the next marketing campaign in the way an honest conversation does.
Invite whoever answers the phone. They know which enquiries were worth having, and that knowledge rarely reaches the people choosing where the budget goes.
Frequently asked questions
What should a campaign start with?
One sentence naming the outcome and the period, specific enough that you could later say it happened or did not. That prevents judging afterwards against whatever number happened to look best.
In what order should I make the decisions?
Objective, audience, budget and period, message, then destination. Each depends on the one before. Skipping straight to the message is common and it is why campaigns feel busy while producing nothing readable.
How many variations should I run?
Two or three well funded ones rather than nine starved ones. Each needs enough volume to produce a trustworthy result. Splitting a modest budget widely looks thorough and teaches you nothing.
How long before I can judge it?
Fix the review date when you fix the budget. A search marketing campaign can often be read within a month. Interruptive channels and long sales cycles need longer. Every intervention before that date resets what you were measuring.
Which variable should a campaign test first?
Creative and destination, because they move results most. Bidding settings come later. Change one variable at a time, and decide beforehand what result would make you keep the change, so the review is a comparison rather than a negotiation.
How should I scale a campaign that is working?
Gradually. Doubling resets what the platform learned and makes before and after incomparable. Expect efficiency to fall as you reach beyond your best audience, and decide whether the extra volume justifies it.
What if it is not working?
Work down the chain rather than reaching for the budget. Poor click rate points at message or audience. Good clicks with poor conversion point at the destination. Good conversion with poor quality points at targeting. Rising costs alone usually mean a competitor arrived, which is not a fault in your marketing campaign.
How do I know when to switch a campaign off?
When the period ends and the objective was missed after a fair attempt at the obvious fixes. Write down what you learned before switching it off, because that note is the real return on a campaign that failed.