Guide · Retail
The single customer view.
The most-bought, least-delivered thing in retail. Everyone wants to recognise the same shopper online and in store; almost nobody can. The way in is not a platform. It is the few moments where recognition actually changes the experience.
This guide, in 5 parts
One person, one record, on both sides of the counter.
Definition
A single customer view is one record for one person, the same whether they buy on your website or hand over a card at the till.
Online and in store, one identity and one history, readable from both directions. Not a tidy database for its own sake. A way to recognise a shopper you already know, at the moments where treating them as a stranger costs you the sale, the loyalty, or the argument at the return desk.
It is the most-requested capability in retail, and one of the least delivered. The reason is not that the idea is unclear. It is that the work gets scoped as a platform, when the value lives in a handful of moments. Name those moments first and the build becomes finite. Skip them and you buy a suite that recognises everyone slowly and no one well.
« Only about 12% of retailers can recognise the same shopper across online and offline. The rest meet customers they have already met, and see strangers. »
Retail TouchPoints, study of retailer identity capability (24% cannot identify a shopper across channels at all; 26% only once a purchase is made).
Meanwhile the customer assumes you already know them. About 70% expect every representative they deal with to have the same information about them, and roughly 79% expect consistent interactions across a brand. The gap is plain: the shopper is one person, and most retailers can only see two halves.
Recognition only earns its keep in a few places.
A single view does not pay off everywhere at once. It pays off at the specific moments where a missed identity is felt by the customer. There are usually two or three. Find them, and you know exactly what to stitch first, with a pass-test you can check in plain language.
The till, when a known shopper looks like a stranger
Someone who buys from you every month walks up to the counter and starts from zero: no history, no size, no past basket, nothing the website already knows. Recognise them here and the associate can finish the sale the way the site would, with the same context.
Pass test · Recognised: the person at the till is the same record that logged in online, before they say a word.
The return desk, when proof of purchase lives in another system
A web order comes back to a store and nobody can see it, so the customer has to argue, dig out an email, or get sent away. One identity that carries the order makes the return a lookup, not a negotiation, at any door.
Pass test · Recognised: a store sees the online purchase, refunds and restocks it, with no call to head office.
An associate with a tablet, when the screen knows less than the customer
Give an associate a tablet and it only helps if the screen knows what the customer knows: the order in flight, the item reserved, the last conversation. A tablet on top of two systems is a slower way to say I cannot see that.
Pass test · Recognised: the associate's screen shows the same shopper, basket and history the customer is holding in their hand.
These three cover most retailers. Yours may have a fourth, a service booking, a click-and-collect handover, a clienteling message. The discipline is the same: name the moment, write the felt test, stitch identity there. Everything else can wait for the platform.
Why the most-wanted view stays unbuilt.
It crosses the hardest internal border
The website team owns an email and a login. The store team owns a loyalty card and a face. The single view lives exactly on the line between them, which is the one line nobody was given to own.
It was scoped as a platform, not a moment
Bought as a suite, it becomes a multi-year programme that boils the ocean and recognises nobody in the meantime. The moments where recognition pays off are narrow, and they get lost inside the platform.
Offline data never gets onboarded
Most of the value is in joining in-store behaviour to the online profile, and that is the step most often skipped. A view that only sees the website is not single, it is half.
Nobody set the test in plain language
Without a felt pass-test, recognised or not, success becomes a dashboard nobody trusts. The customer's experience is the only honest measure, and it is rarely the one being tracked.
Stitch identity for the moments first, let the platform follow.
The order matters as much as the work. Start from the moments, join them with a key you already have, bring in the offline side, and only then decide on a platform. Which moments to take first, and what to defer, is a Hub Map call. The right question is not what to buy, it is where recognition is felt soonest.
Name the few moments where recognition changes the experience
Start from the customer, not the schema. Pick the two or three places a missed identity is actually felt: the till, the return desk, the associate with a tablet. These are the moments worth stitching first, and most retailers have fewer of them than a platform pitch implies.
In shortTwo or three felt moments, named before any tooling.
Choose one identity key both sides already hold
You do not need a perfect universal ID to start. Pick the key both the website and the store can actually capture today, a phone number, an email, a loyalty scan, and make it the join for those moments. Perfect identity is a programme; a working key is this quarter.
In shortOne key both channels can capture now, not the ideal one.
Onboard the offline side, because that is the missing half
The online profile usually exists already. The work is bringing in-store behaviour to it: purchases, returns, what an associate logged. This is the step most skip and the one that turns a web-only view into a single one, so it is the step that earns the name.
In shortBring store behaviour to the profile, not the reverse.
Let the platform follow the moments, not lead them
Once the named moments recognise the customer with a real key and both sides feeding it, the platform decision gets easy, because you are buying to extend something that already works rather than to start. Stitch first, scale second; the suite is how you grow the win, not how you find it.
In shortBuy to scale a working join, never to start one.
The single-view checklist.
It is on the right track when it can tick all five. Until then, it is a platform looking for a moment.
- The two or three moments where recognition is felt are named, before any tooling.
- One identity key both the website and the store can capture today is chosen as the join.
- Offline behaviour, purchases and returns and floor notes, is onboarded to the online profile.
- Each moment has a felt pass-test: recognised, or not, in plain language.
- The platform decision comes after the moments work, to scale the join rather than start it.
Frequently asked.
What is a single customer view, in plain terms?
It is one record for one person, the same whether they buy on your website or hand over a card at the till. Online and in store, one identity and one history, readable from both sides. The point is not a tidy database; it is that a shopper you already know is recognised at the moments where being a stranger costs you, the counter, the return desk, the associate with a tablet. Most retailers want it, and only about 12% can actually do it across channels.
Do we need to buy a CDP or a big platform first?
No, and starting there is the common mistake. A platform bought before the moments are named becomes a multi-year programme that recognises nobody while it runs. Begin with the two or three moments where recognition is felt, join them with one identity key both channels already capture, and onboard the offline side. Buy the platform after that works, to scale a join that already pays off rather than to go looking for one.
Where do we actually start?
With the moments, not the schema. Name the few places a missed identity is felt by the customer: a known shopper treated as new at checkout, a web return nobody at the desk can see, a tablet that knows less than the person holding it. Stitch identity for those first, with one key both sides hold today. The narrow start is the point: fewer moments, recognised properly, beat a universal view that recognises everyone slowly.
Is this just personalisation by another name?
No. Personalisation is what you do once you can recognise someone; the single customer view is the recognition itself. Without it, personalisation runs on half the picture, the website half, and misses everything the customer did in store. Get recognition right at the moments that matter and personalisation has something true to stand on. Skip it and you are personalising for a stranger you have actually met many times.
Can our existing teams build this, or do we need help?
Often your teams can, and where they can, they should. If you have engineers who know both your e-commerce stack and your store systems, the named moments and one identity key are a clear programme to hand them. We are useful when the two sides were built as separate companies and the identity join is nobody's job, when one side is strong and the other is not, or when you want it shipped without a long hire. We have built both sides, which is rarer than it should be.
Much of this you can run yourselves. If you have engineers who know both your e-commerce stack and your store systems, the named moments and one identity key are a clear programme to hand them.
Bring us in when the two sides were built as separate companies and the identity join is nobody's job, when one side is strong and the other is not, or when you want it shipped fast without a long hire.
We are the rare team that has built both sides. This is the how-to under our flagship offer, The Whole Way, where we build and stay until recognition holds.
Meeting customers you have already met?
No deck, no gate. A working session on the real moments where a known shopper looks like a stranger, with the operators who would stitch the join.