Knowing When to Turn It Off

Why campaigns outlive their usefulness

Because continuing is passive and stopping is a decision somebody has to make and defend.

Nobody has to justify the campaign that ran again this month. Somebody has to justify cancelling it.

Which produces a systematic bias towards continuation, entirely independent of whether it is working.

The remedy is deciding the stopping condition in advance, when nobody is yet invested in the answer.

The three honest reasons to stop

• Cost per enquiry exceeds what an enquiry is worth.
• The target you set was missed over the period you agreed.
• The money would do more elsewhere, demonstrably.

The first is arithmetic and should end the discussion. If an enquiry is worth four hundred and you are paying six hundred, the campaign is losing money every month regardless of how many enquiries it produces.

The second requires having set a target, which is why that step matters.

The sunk cost trap

The reasoning that keeps bad campaigns running, and it is worth naming precisely.

We have spent this much already, so stopping now wastes it.

Which is exactly backwards. Money already spent is gone in every scenario, and the only question is whether the next month's spend will return more than it costs.

The variant specific to advertising: it needs more time to learn, which is sometimes true early and becomes an indefinite excuse after six months.

Setting the period in advance is what distinguishes a genuine learning phase from an excuse.

A worked example

A business three years into a campaign that had never been formally evaluated.

Working out the numbers for the first time: the annual spend, the enquiries attributable to it, and the proportion that became jobs.

The cost per job was slightly above the average job's gross margin, which meant the campaign had been losing money for three years while appearing to work because enquiries were arriving.

They stopped it and redirected the budget to a person answering the phone more reliably, plus a systematic review request after every job.

Enquiries fell and jobs won stayed roughly level, because a larger proportion of a smaller number converted.

The uncomfortable finding was that nobody had ever done the calculation, in three years, because nothing had forced it.

Doing the calculation

Simpler than it sounds and requiring only figures you should have.

Total spend over a period, divided by enquiries attributable to it, gives cost per enquiry.

Divide again by the proportion that became jobs, and you have cost per job.

Compare that against the gross margin on an average job, not the price.

If the cost per job exceeds the margin, the campaign is losing money on every job it produces, which is a state that can persist for years while looking busy.

The month after stopping

Worth anticipating, because it feels worse than it is.

Enquiries fall immediately, since advertising stops producing the day it stops running.

Which produces an instinct to restart, before the actual question has been answered.

The question is not whether enquiries fell. It is whether the jobs won fell, and by enough to justify the spend that was producing them.

Give it a full quarter, since a fall in enquiries with no fall in jobs means the campaign was producing enquiries that never converted.

Where continuing is right despite poor numbers.

A campaign genuinely still in a learning period, which for a small budget is a few months rather than a few weeks, and which should have a stated end date.

Seasonal work where the annual figure matters and a quiet quarter is expected.

A long sales cycle where enquiries today become jobs next year, and the current period's figures understate the return.

And situations where the campaign is doing something other than producing enquiries, such as defending a brand term, which should be judged separately.

Who is making the decision

Worth being aware of, since the incentives are not aligned.

Where an agency manages the campaign and is paid a proportion of the spend, the recommendation to continue comes from somebody whose income depends on it.

That does not make them dishonest, and it does mean the evaluation should be done by the business rather than delegated.

The numbers required are ones the business has: what a job is worth, what the margin is, and how many jobs came from the campaign.

Asking a supplier whether their own work is worth continuing is a question with a predictable answer, however good they are.

Stopping partially

Frequently better than stopping entirely, and less often considered.

A campaign is rarely uniformly bad. One service, one area, or one set of terms usually accounts for most of the return and another for most of the waste.

Which means the decision is not always continue or stop, but which parts to keep.

Turning off the weakest half and leaving the rest running at the same total budget concentrates the spend and frequently produces a viable campaign from an unviable one.

That requires enough data to tell the parts apart, which is another argument for a concentrated budget in the first place.

What to do with the money

The question that makes stopping easier.

Stopping without a plan feels like retreat. Stopping in order to do something specific feels like a decision.

The alternatives that compound: collecting reviews, writing the service pages that were never finished, photographing work properly, or improving how quickly enquiries are answered.

That last one is frequently the highest return available. A business answering enquiries within an hour converts noticeably better than one answering within two days, and the improvement applies to every enquiry from every source.

Which is a considerably better use of a monthly budget than buying more enquiries that you then answer slowly.

Setting the condition now

1. Work out what an enquiry is worth.
2. Work out your current cost per enquiry.
3. Write down the figure at which you would stop.
4. Write down the review date.
5. Check the parts separately before stopping the whole.
6. Decide where the money goes if you stop.

The third and fourth together are what turn an indefinite spend into something with a defined end, and writing them down is the whole of the discipline involved.

A campaign with a written stopping condition is one that will be evaluated. One without will simply continue.

Setting that target properly is covered in deciding what success looks like first.

Frequently asked questions

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.