What Every Marketing Pound Should Prove: The Owner’s Invoice Test

Most small-business owners are not short of marketing activity. They are short of a straight answer. Money goes out to Google, Meta, a leaflet printer, a local magazine, maybe a freelancer. What comes back is harder to pin down. The result is a familiar feeling: you suspect some of it works, but you are not sure which part.

The invoice test is a way to fix that without buying a dashboard or hiring an analyst. The idea is simple. Every marketing channel has to submit an invoice — not the one it sends you, but the one it earns. What did it cost, what did it bring in, and did it cover its own bill?

Start with the true cost, not the headline price

A £500 ad spend is rarely a £500 marketing cost. Add the management fee, the software subscription, the offer discount and the hours you or your team spend replying to enquiries. If you would not do the work for free, your time belongs in the sum.

Take a typical small firm spending £600 a month on Google Ads, £150 on management, £80 on tools and six hours of owner time at £25 an hour. The real monthly cost is not £600. It is £980. That changes the maths. A channel that looks cheap at the invoice stage can be expensive once the whole picture is on the table.

Write down the true cost for each channel you use. Keep it boring. A spreadsheet is fine. The point is to compare like with like.

Count what comes back, not what looks busy

Likes, impressions and reach are not payments. They can be useful signals, but they do not pay the wages. For the invoice test, count the things that lead to money: enquiries, bookings, quotes, sales and repeat custom.

You do not need perfect attribution. You need a consistent habit. Ask every new customer how they found you, and write it down. Use a separate phone number for a leaflet drop. Give each campaign a simple code or booking link. Check your Google Business Profile insights. If a channel cannot be tracked at all, treat that as a warning, not a personality trait.

Then build three columns for each channel: cost, new customers, and first-sale value. If you can, add repeat value over the next three to six months. A £40 first sale that turns into a £400 customer is not the same as a £40 one-off.

The invoice test in three lines

For each channel, answer three questions:

1. What did it cost, including time and extras?
2. How many paying customers did it bring in during a clear period — say 90 days for direct-response activity?
3. What was the revenue from those customers, including repeat business you can reasonably trace?

If a channel brought in £2,400 of revenue and cost £800, it returned £3 for every £1 spent. That is a useful sentence. If it brought in £300 and cost £900, that is also a useful sentence, even if it is an uncomfortable one.

Not every channel should be judged on the same clock. Local SEO may take six to nine months to show its value. A leaflet drop should show something much sooner. The test is not about demanding instant results from everything. It is about agreeing the payback window in advance and then checking whether the channel met it.

A worked example from the real world

Picture a joinery firm in Stoke-on-Trent. It spends £800 on Google Ads and gets 20 enquiries, six jobs and an average first job of £1,800. That is £10,800 from £800, before repeat work. The channel is pulling its weight.

It also spends £300 boosting Facebook posts. It gets five enquiries, one job and £1,800 of work. The return looks decent, but the cost per job is £300. If the owner has capacity, it may be worth fixing the targeting or the landing page. If the diary is full, it may be a channel to pause.

A local magazine costs £350 and produces no traceable jobs in three months. The owner likes the advert. That is not a business case. It goes on the cut list unless a clearer offer or tracking method can be tested.

An email to past customers costs almost nothing and brings four repeat jobs worth £2,000. That is the quiet winner. It rarely gets the attention of paid ads, but the invoice test does not care about excitement.

Keep, fix or cut

The output of the invoice test should be a decision, not a report. Put each channel into one of three boxes.

Keep: it pays for itself within an agreed window, and the numbers are not a one-off fluke.

Fix: the channel has promise but a clear leak. Maybe the landing page is slow, the offer is vague, the follow-up takes three days, or the targeting is too broad. Fix one thing at a time and check again.

Cut: it has had a fair run, the numbers do not add up, and there is no specific fix worth testing. Cutting a weak channel is not failure. It is how you fund the stronger ones.

Be careful with the word brand. Brand matters, especially for small local firms. But brand spending still needs a job to do. If the job is awareness, decide how you will know it worked — more direct searches, more profile views, more people mentioning the advert at the counter. If you cannot describe the result, you cannot invoice for it.

The awkward bits are worth it

Attribution will never be perfect. Someone may see your van, then search your name, then click an ad. Someone may get a leaflet and ring the number they already had. That is normal. The invoice test does not require perfect data. It requires a consistent, honest method that is good enough to spot the obvious winners and the obvious money pits.

Do it once a quarter. Give yourself 45 minutes. List the channels, add the true costs, count the customers, and ask each channel to pay its own invoice. If it cannot, either fix it or stop renewing it by default.

Marketing is not magic. It is a series of small financial bets. The invoice test simply makes sure you know which bets are working before you place the next one.