Measuring the post purchase stage with numbers you already have

4 days ago
3 min read

Introduction

Businesses measure acquisition in detail and retention not at all. There are figures for enquiries, conversion and cost per customer, and nothing whatsoever for what happens afterwards, despite the fact that in most firms the majority of profit comes from customers who return.

The reason is not difficulty. Every number needed is derivable from invoices you already hold. It is that nobody has decided which numbers matter, so none are calculated, and the stage stays invisible in the reporting even while it dominates the accounts.

Six figures cover it, and they can all be produced in an afternoon. None of them require software you do not already own.

1. Measuring the post purchase stage starts with the repeat rate

The first number.

The proportion of customers who ever buy a second time. It is the single most informative retention figure and very few owners know it. Calculate it over a period long enough for a second purchase to be plausible. Two full purchase cycles is the minimum.

2. The interval between purchases

The second number.

The median gap between consecutive purchases. It tells you when contact should happen and when a customer has effectively lapsed. Everything about timing depends on it. Use the median rather than the mean.

3. Revenue from repeat customers

The third number.

What share of income comes from people who had bought before. This is usually much higher than expected and reframes where effort belongs. Track it annually. Compare it against what you spend acquiring new customers.

4. The lapsed count

The fourth number.

How many customers are past one and a half intervals with no further purchase. It is a list as well as a number. Both are actionable. The list is the more useful half.

5. Time from order to delivery

The fifth number.

Days between agreement and the work being complete, which covers the gap where deals still fail. Long values here predict problems later. It is countable from existing records. Watch the longest cases as well as the median.

6. Problems raised after completion

The sixth number.

A simple count, by month, of issues raised after a job closed. Rising counts indicate a process problem rather than bad luck. Nobody records this and everybody should. One line per issue is enough.

7. Add a use measure if that applies

The conditional one.

Where the customer has to do something for the purchase to work, count whether they did. Non-adoption is invisible otherwise. This is essential in subscriptions and training and irrelevant in some trades. Decide once whether it applies to you.

8. Compare against acquisition cost

The reason it matters commercially.

Retention figures only become decisions when set against what a new customer costs. The comparison usually justifies substantial effort on the existing base. Put both on the same page. A retained customer typically costs a fraction of a new one.

9. Review annually

The cadence.

These move slowly and quarterly review produces noise. Once a year, properly, against the previous two years. That is enough to see a trend. Book the afternoon rather than intending to find it.

Be careful about measuring retention over too short a period. A business with a two-year purchase cycle cannot assess repeat rate from twelve months of data, and doing so will make a healthy relationship look like a failure.

Conclusion

Spend an afternoon with your invoices and produce six retention figures.

Calculate the proportion of customers who buy a second time, the median interval between purchases, the share of revenue from returning customers, and the count of those now lapsed. Add the time from agreement to delivery and a monthly count of problems raised after completion, include a use measure where the customer has to act, compare all of it against what a new customer costs, and review the whole set once a year.