Retail customer data scorecard
Answer honestly and you’ll get a score, a tier, and the two gaps costing you the most. No sign-up or shared data required.
What each tier looks like from the inside
Same discount depth, same margin, just fewer orders sold on promotion. You currently sell 45% of orders on promotion.
Scattered: 0 to 5
Stitched: 6 to 10
Joined up: 11 to 14
Customer-led: 15 or 16
Book a demo to close your two biggest gaps.
We’ll take your answers and show you what closing the top two looks like on your own data: what gets joined, what a marketer can then build unaided, and what it changes in the first ninety days.
Why joined-up customer data decides everything downstream
Most retail marketing problems that look like strategy problems are actually plumbing problems. A team that cannot tell whether the person who bought in Chadstone last Tuesday is the same person who abandoned a basket online this morning is not going to fix its retention rate with better subject lines. It is going to keep running campaigns to half a customer.
The scorecard above is built around four things that matter more than the rest, and it is worth saying why.
Identity across channels
If one person can appear as two records, every number built on that record is wrong in the same direction. Purchase frequency halves. Recency looks worse than it is. Lifetime value understates.
Worst of all, exclusion lists stop working, which is how a customer ends up being emailed a discount for something they bought in store the week before.
Margin by customer
Almost every retailer knows margin by product and almost none know it by customer. That gap is why discount dependency creeps: without customer-level margin, the busiest cohorts look like the best ones, and the promotional calendar gets built to serve them. The customers who quietly buy at full price and come back anyway are invisible in a product-level view, and they are the ones funding the business.
Segment latency
How long it takes to get a new audience is the most honest single measure of whether a customer data setup works, because it tests the whole chain at once: ingestion, identity, definitions, access and trust. Retailers who answer “a sprint” have usually stopped asking for the segments they actually want, which means the campaign calendar has quietly reshaped itself around what is easy to build. That reshaping is invisible on any dashboard.
FAQs
What makes retail customer data joined up?
Four things: one person’s online and in-store purchases resolve to a single record; gross margin is known per customer, not just per product; a marketer can build an audience without raising a request; and the view is current enough to act on this week.
Why does in-store purchase data matter for email marketing?
Because for an omnichannel retailer most behaviour happens where the email platform can’t see it. If store purchases never arrive, every segment and exclusion is built on the online fraction of the relationship, which is how people get emailed things they already bought.
How long should it take to build a new customer segment
Same day. If it takes a sprint, the calendar ends up planned around the audiences that already exist rather than the ones you need, and most teams stop asking long before they admit it.
The scoring
Four questions are marked high impact: in-store visibility, segment latency, margin by customer, and identity across channels. That weighting decides which gaps we surface first, not how the score is calculated. Every question counts the same towards the total, so the number stays something you can verify yourself rather than an index only we can compute.
Where the questions come from
| Figure | Source | Status |
|---|---|---|
| The four high-impact questions | Bluecore’s finding that retailers recognising over 40% of customers ran repeat purchase rates 53% higher than average, and those under 10% ran 33% lower. Identity and visibility are the strongest observed differentiators | Sourced (Bluecore, April 2024, 100+ retailers, calendar 2023) |
| The remaining four questions | Lexer’s own experience of what blocks retail marketing teams. Not derived from published research | Assumption |
| Tier boundaries (5 / 10 / 14) | Chosen to split the range into four readable bands | Assumption. A designed instrument, not a validated one |
| Category median | No published data exists on retail customer-data maturity | Not shown. We’d have to invent it |
What we’re assuming
This is a scorecard, not an audit. It’s self-reported, and generous answers produce generous scores. It’s useful for placing yourself roughly, not for reporting upward.
The tier boundaries are a design decision. They split sixteen points into four bands people can hold in their head. They aren’t derived from a distribution, because no distribution has been published.
We show no category median, deliberately. Every comparison figure we could put there would be invented. When enough retailers have completed this, the median becomes real and we’ll publish it with the sample size attached.
Half the questions come from experience rather than research. The four high-impact ones line up with what Bluecore observed about customer recognition. The other four are what we see blocking retail teams, and we’ve labelled them as such.