Single customer view
One client record, across a network of stores.
In a network, the same client often exists four times over: once per store they have visited, once on the website, once in the till's customer file. A single customer view turns those records into one, kept properly and usable where the client actually shows up. This page explains what it is, where duplicates come from, and what one record changes for the person standing in front of the client.
Definition
A single customer view: what it actually is.
A single customer view is the base that guarantees a client exists only once across an entire brand, whatever the store, the till or the channel they came through.
What it is
One record per person, built by matching the entries that refer to the same client, then made available to the tools that use it: the store, marketing, customer care, headquarters.
What it is not
Neither a warehouse where data piles up awaiting a purpose, nor a campaign tool. It sells nothing and sends nothing: it answers the prior question — is this the same person? — on which everything else depends.
Where duplicates come from
From the ordinary life of a network. A client pays without producing their loyalty card, orders online under a different address, has their name typed with a typo on a busy afternoon. None of this is a mistake: these are the real conditions of selling.
Who it serves
The advisor first, because they recognise a client a colleague served elsewhere. Marketing, so the same person is not contacted three times. Leadership, to count a client once and a sale once.
The same client, several times
Where a client splits in two.
Pages on this subject talk about silos and a 360-degree view. Here are the concrete scenes, as they happen in a store network, and what one record does with them.
| What the network sees today | What one record changes | ||
|---|---|---|---|
| They pay without producing their card | One more anonymous receipt, and a loyal client who appears nowhere | The payment, the name or the contact details attach the purchase to the existing record | |
| They order online and collect in store | A web client and a store visit, counted as two people | One client, known to buy online and collect in person | |
| They visit two stores in the network | Two partial histories, two advisors each believing they found them | One complete history, and a named person to follow up | |
| Their name is typed two different ways | Two near-identical records nobody has time to merge | A proposed match, and a human decision whenever the case is doubtful | |
| They buy tax free | A separate administrative file, disconnected from the relationship | A purchase attached to the right client, in the same history as the rest |
Every one of these lines happens in a network of two stores as much as fifty. It is not a question of size, it is a question of how many places the same client can appear without being recognised.
How it holds together
From a till entry to one record.
The principle is simple, and worth understanding before choosing anything: what separates a properly kept record base from an ordinary customer file is not the quantity of data, it is the rule applied to every new entry.
01
Data arrives from where it is created
A purchase at the till, an online order, a sign-up in store, an exchange with customer care. Nobody retypes anything: what is already recorded somewhere should not be recorded a second time.
02
Records are matched
Entries that plainly refer to the same person are brought together. On the platform, matching happens automatically; what matters to the brand is that the rule is the same everywhere and that it is written down.
03
Doubtful cases stay decidable
Two namesakes in the same city do not merge on their own. A serious record base can say « I don't know » and leave the decision to a person, rather than mixing two clients — the costliest error of all.
04
The rule is written once for the whole network
What counts as an active client, what counts as consent, what gets matched and what stays separate. Local variants make network figures impossible to compare, and that is what sinks these projects long before the technology does.
At the counter
What it changes for the advisor.
This is the great absentee of the subject: of the ten French pages ranking on this topic today, one mentions the advisor and one mentions the till. The record base is described as a data project. Yet the person who benefits most immediately is the one greeting the client.
- They recognise a client a colleague served in another store, and serve them as if they knew them
- They do not ask again for a size, a preference or an address the house already holds
- They see a complete history rather than the quarter of it that happened in front of them
- They do not create a fourth duplicate by typing quickly on a busy day
What it unlocks
What one record makes possible.
A record base has no value in itself. It has value because it makes four things feasible that are otherwise approximate or plainly wrong.
Following up without duplicating
A client who exists three times receives the same message three times, often on the same day. That is not merely inefficient: it is the kind of detail that undermines a house in front of its best clients.
Recognising a client in any store
Recognition is the very promise of high-end retail. It requires the information to follow the client, not the store where they first happened to walk in.
Assigning client books without overlap
Clients can only be assigned to advisors if each client exists once. Otherwise two advisors follow the same person, with two different histories and two contradictory messages.
Counting a client once
Active clients, purchase frequency, average basket: all of these are calculated on people, not on records. A file holding 40,000 records for 28,000 people is wrong about everything.
The proof
A sale counted once.
None of the ten competing pages analysed says how you know the record base achieved anything. Yet that is the question leadership asks, and it deserves a protocol rather than a promise.
01
Trace the action
Message, appointment, follow-up: every gesture towards a client is timestamped and attributed to the person who made it. Without that trace nothing can be measured afterwards — and it only exists if the client is identified once.
02
Set the attribution window
Seven, thirty or ninety days between contact and sale. It is the brand's decision, written down and known to everyone, not a setting hidden in a tool.
03
Keep influenced revenue separate
A sale falling inside the window joins influenced revenue. It stays distinct from walk-in revenue and is never added to it: a sale counted twice discredits the whole measure, and the record base with it.
04
Reconcile against the till
Store by store, line by line. Across equipped retailers, influenced revenue commonly represents 20 to 25 percent of total revenue — an order of magnitude that means nothing while clients are still duplicated.
The data
A client file kept properly.
Half the pages on this subject cover data protection, and rightly so: a record base brings into one place what used to be scattered. Four points are enough to set the frame.
01
Consent, collected and dated
Who agreed to be contacted, on which channel, and when. Consent belongs in the client record, not in a spreadsheet kept beside it.
02
Access and erasure rights
A deletion request has to run across the whole network. That is precisely what a single record base makes possible, and what a scattered file makes impractical.
03
Field-level permissions
An advisor sees their clients, a manager their store, headquarters the network. Bringing data together does not oblige anyone to open it to everyone.
04
An access log
Who read what, who changed what, when. It protects the house, and it settles internal debates better than a conversation does.
Around the record
What connects to it.
A single customer view is built from what already exists and serves what comes next. It is never an isolated project.
The sources
Where client data comes from
Point-of-sale systems, e-commerce platforms, sending tools: Selekt connects to what is already in place so that purchases and contacts feed the record without retyping.
All integrationsThe measure
A sale counted once
Attribution window, influenced revenue kept separate from walk-in revenue, reconciliation against the till: the protocol that turns a clean base into evidence.
Influenced revenue, explainedWhat you do with it
What this record becomes on the shop floor
Recognising, reaching out at the right moment, following a client book: one record is the condition, client relationships in store are the craft it makes possible.
Clienteling, plainly explainedFrequently asked
Frequently asked questions.
01
What is a single customer view?
A single customer view is the base that guarantees a client exists only once across an entire brand, whatever the store, the till or the channel they came through. It matches the entries that refer to the same person, then makes that one record available to the tools that use it.
02
How is a single customer view different from a CRM?
The record base answers « who is this client, and is it really the same one? ». The CRM answers « what do we do with them? ». The first holds identity and data reliability, the second holds the relationship and the actions. In a store network, the second is worth nothing without the first.
03
Single customer view, golden record, customer data platform: the same thing?
No, and the confusion is common. A golden record is the consolidated record itself; a single customer view is the result of cleaning and matching; a customer data platform is a family of tools often used to build that view and activate it in marketing. The view is an outcome, the platform is one way of reaching it.
04
Do we have to replace our point-of-sale system?
No. The record base is built from the purchases and contacts the till and the website already record; it replaces neither. The till carries on taking payment, managing stock and feeding the accounts. No migration is required.
05
From how many stores does this become useful?
From the second one, and as soon as there is a website. Duplicates do not come from the size of the network but from the number of places the same client can appear without being recognised. The rarer the visit and the higher the basket, the more a lost record costs.
06
How do you know a single customer view returned anything?
By tracing every action taken towards a client, setting an attribution window — seven, thirty or ninety days — and counting the sales that follow separately. That influenced revenue stays apart from walk-in revenue and is never added to it. Across equipped retailers, it commonly represents 20 to 25 percent of the total.
Request a demo
Discover Selekt directly with your use cases.
A guided demonstration, tailored to your boutique network. No commitment.
Page updated September 16, 2026