top of page

The natural interval before somebody buys again from you

  • 4 days ago
  • 3 min read

Updated: 2 days ago

Introduction


Every business that sells more than once has a rhythm to it. Boilers get serviced annually, hair gets cut every six weeks, accountancy recurs at year end, a garden gets landscaped once a decade. Most owners have an instinct for their own rhythm and very few have measured it, which means contact happens when somebody remembers rather than when it would work.

The measurement is straightforward. Every pair of consecutive invoices to the same customer gives you a gap; the median of those gaps is your interval. An hour with the accounts produces a number that improves the timing of every retention effort afterwards.

It is one of the highest-return calculations available to a small business.


1. The natural interval before somebody buys again comes from your invoices


The method.

List customers with two or more purchases, calculate each gap, take the median. Nothing more sophisticated is required. The answer is usually surprising in one direction or the other. Thirty customers is enough to get a usable figure.


2. Use the median, not the average


The statistical point.

A few customers with very long gaps will drag an average upwards and produce contact that arrives far too late. The median describes the typical case. Report the range as well. The quartiles are worth knowing if you have the data.


3. Expect more than one interval


The segmentation.

Different services and different customer types often have distinct cycles, and averaging across them produces a number that fits nobody. Calculate separately where the difference is large. Two or three groups is normally enough. More than that becomes unmanageable.


4. Make contact before the midpoint of the interval


The timing rule.

Arriving after the need has surfaced means competing with whoever they already called. Slightly early is harmless; late is worthless. This is the whole practical value of the number. Two-thirds of the way through is a reasonable default.


5. Watch for customers who are overdue


The list that pays.

Anyone past one and a half times the interval has effectively lapsed without either side noticing. That list is the cheapest source of work in the business. It takes one filter to produce. Work through it by phone rather than by email.


6. The interval changes with circumstances


The caution.

A change of house, a change of role or a change in the business can lengthen or shorten it permanently. Recalculate annually. Treat the number as current rather than permanent.


7. Some businesses have no interval at all


The honest exception.

Where a purchase genuinely happens once, retention effort is misdirected and referral effort is not. Check before assuming a cycle exists. Your invoice data answers this immediately.


8. Use it to forecast


The secondary benefit.

Knowing how many customers are approaching their interval gives you a rough view of the coming quarter that no marketing metric provides. It is not precise and it is directionally useful. Most owners have nothing else of the kind.


9. Record it beside the customer


The operational step.

An expected next-purchase date on the record turns the interval from an interesting statistic into something that prompts action. Otherwise it stays in a spreadsheet. The prompt is what produces the result.

Be careful about contacting everybody at the same point regardless of what they bought. A customer who bought a one-off installation and one who buys consumables monthly need entirely different timing, and a single interval applied to both will irritate one and miss the other.


Conclusion


Calculate the median gap between consecutive purchases and use it to time your contact.

Work from invoices rather than impressions, prefer the median to the average, calculate separate intervals where services or customer types differ, make contact before the midpoint rather than after, build a list of everyone past one and a half intervals, recalculate annually, accept that some businesses genuinely have no cycle, and record an expected next-purchase date against each customer so the number prompts action.


Related reading


 
 
 

Comments


bottom of page