Who gets credit when a repeat customer books again?
Decide the rule before you read the report, then apply it everywhere. The realistic options are crediting the original acquisition source permanently, crediting nothing and treating repeat work as house revenue, or crediting whatever produced the most recent contact. Each answers a different question. The dangerous option is letting each system apply its own default, which double counts repeat revenue and inflates brand and retargeting.
Why the default is wrong
Ad platforms count conversions, not customers. When an existing customer clicks a branded ad and calls, that is a conversion, and the platform will credit the campaign with whatever value you pass back. Do this for a year and your branded search and retargeting look like the best-performing spend in the business, because they are collecting credit for relationships that already existed.
The operational system has the opposite bias. It knows the customer is not new, so it may credit nothing to marketing at all. Neither view is usable on its own.
Three defensible rules
Any of the three rules below can be defended. What cannot be defended is applying different rules in different reports, or letting the tooling choose silently. Write the rule down, put it on the report itself, and change it only deliberately.
- Original source, permanently. All revenue from a customer credits the source that first acquired them. Best for evaluating acquisition spend and lifetime value.
- House revenue. Repeat work credits no source. Best for judging whether marketing is growing the customer base rather than harvesting it.
- Most recent contact. Credit goes to whatever produced this booking. Best for evaluating reactivation and retention campaigns specifically.
The reporting structure that avoids the argument
Split every revenue report into new-customer revenue and existing-customer revenue as a first cut, before any source breakdown. This single split resolves most attribution disputes because it makes the question explicit instead of implicit.
Then apply the original-source rule inside new-customer revenue and the most-recent-contact rule inside existing-customer revenue. Acquisition spend is judged on the first pool. Reactivation, memberships and email are judged on the second. Nobody is competing for the same revenue twice, which is what makes customer intelligence more useful than campaign-level reporting alone.
Identity resolution is the hard part
All of this depends on reliably recognizing that a caller is an existing customer. Phone numbers change, spouses call, addresses get entered three different ways, and commercial accounts have many contacts. Matching on a single field will misclassify a meaningful share of repeat customers as new.
A practical approach matches on phone plus normalized service address plus name similarity, then reports its own match confidence rather than pretending it is perfect. Publishing the unmatched share as a visible number keeps the report honest and tells you when identity resolution is worth more engineering effort. That is standard work inside an AI systems integration engagement.
Topics: repeat customers · attribution · lifetime value · reporting rules · retention
Have a version of this question about your own business?
The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.