Counting customers, not leads

How long a customer takes to pay you back sets how fast you can grow. Pay too much for each one and it drains your working capital quickly.

A retailer selling by catalogue and online, in two countries, set out to grow by advertising for new customers. In two years the spend on winning customers more than tripled.

The customers did not keep pace. The total finished behind target, and the cost of each extra customer had risen faster than the number of them. I was one of the business leaders who delivered the work that followed.

What we measured

Three numbers, kept separate.

Leads: people who signed up or asked for a catalogue.
Conversion: the share of those leads who went on to buy.
Cost per customer: the spend divided by the customers who actually bought.

We also wrote down exactly who counted. Someone who signed up before a campaign and bought during it was not credited to that campaign. Someone who signed up during it and bought later was. Without that rule, every channel claims the same customer.

Cheap leads were hiding expensive customers

The first television campaigns asked people to phone for a free catalogue. They beat their lead targets. Only about a third of those leads ever bought.

Asking for the sale worked better than asking for a sign-up

Over three campaigns the advertising changed what it asked people to do.

Campaign	What the ad asked	Leads against target	Leads who bought
First	Phone for a catalogue	Above	About a third
Second	Go to the website for a catalogue	Above	Over half
Third	Go online and shop now	Below	Six in ten

The third campaign brought in fewer leads than planned. The cost of winning a customer had still more than halved in the larger market, because six in ten leads now bought.

What else it showed

The new customers behaved differently. Once advertising was pushing people to sign up, more of them signed up and never bought, and fewer went on to a second and third order. The questions to ask about that are on the customers page.

Search was doing more than it was paid for. It brought the largest number of new customers in both countries. When the search budget was cut in one of them, new customers there stalled straight away.

Catalogues sent cold were the dearest way to buy a customer, at several times the cost of the best television campaign.

What changed

The plan for the following year set a ceiling on what a customer could cost in established markets, and a higher, agreed figure in new ones, to come down as the brand became known.

What you can take from it

Count customers, not leads. A campaign can beat its lead target and still lose money.
Write down who counts as acquired, or every channel will claim the same person.
Ask what the next customer costs, not the average one.
When a budget is cut, watch which number moves first. That is the channel doing the work.

The marketing page lists the questions to put to your own numbers.

The free 45-minute chat ends with a one-page readback of where I think the money is most likely being lost in your business. Ask for it here.

Counting customers, not leads: what it really costs to win a new customer.