The short answer
A single customer view (SCV) is one record per customer that brings together everything your business knows about them: what they browsed on your website, which stores they visited, what they were quoted, what they bought and what your team knows about them. In retail it's also called a unified or omnichannel customer profile, and it's the foundation that clienteling, personalisation and honest attribution are built on.
Storing the data is the easy part. The hard part is the joins — linking an anonymous browser to a named person, recognising one person with two email addresses, attaching till and concession orders to the right record — and the rules for those joins decide whether anyone can trust what the record says.
Picture one customer's year. A first order on your website in March. A visit to your flagship in April, where they buy a footstool and the till opens a fresh record for them. A lamp from the counter you run in a department store in June, which reaches you as a line on the host's monthly statement. Twenty minutes browsing your new collection, signed out, on their phone in August. In November, a win-back email offering 10% off their "first order". Five touchpoints, three systems, and a regular who looks like three strangers.
What goes into a single customer view
| Source | What it adds | How it usually joins |
|---|---|---|
| Website browsing | Products viewed, baskets, searches, return visits | A first-party cookie on your own site, attached to a person when they identify |
| Campaign arrivals | Which ad, email or link brought them, and when | Recorded on arrival, attached along with the browsing |
| Store visits | Which showroom, when, and what they were shown | Appointments, codes scanned in store, quotes raised in the room |
| Quotes | What they were offered, when they opened it, what they decided | Raised against the customer directly |
| Orders | What they bought, where, for how much, and what came back | Email address on the order, from your checkout, till or imports |
| The team's knowledge | Notes, preferences, who they buy with, what they're waiting for | Written against the customer by whoever served them |
The joins that make or break it
Anonymous to known
Most of your website visitors never sign in. When one finally does identify — by subscribing, enquiring, booking or asking for a quote — the question is whether their earlier browsing joins their record or whether the record starts from that moment. For a considered purchase, the weeks before the email are where the decision was made, so a single customer view that starts at the email misses most of the story.
One person, many addresses
People use a personal email to browse and a work email to order, or a partner's address for delivery. An SCV needs a way to say "these addresses are the same person" and route future orders to the existing record rather than creating a duplicate. When two records turn out to be one person, merging them should move everything across and leave an audit trail of who decided.
Two people, one browser
The family iPad is the classic case: two people identify from the same device. A careless system silently reassigns the browsing history from one to the other, and now one customer's record contains the other's shopping. The safer rule is to refuse to reassign automatically and to raise it as a question for a person to answer.
Guest orders, tills and concessions
Some sales arrive with no email at all, and some — like a concession inside another retailer — arrive as a spreadsheet from somebody else's till. Keep them anyway, mark where they came from, and join them to a customer when the information allows. Importing the same file twice should never double-count.
Deletion that doesn't break the thread
Records get tidied: a test quote deleted, a duplicate customer removed, a store closed. None of that should sever a customer's history. If deleting a quote also forgets that the browser behind it belonged to that customer, you'll meet them as a stranger next time.
Keep the sequence, not a tally
It's tempting to summarise: "viewed 6 products, 3 visits, 2 orders". But six views in four minutes and six views spread over five weeks are different customers — one skimming, one deciding. A good single customer view keeps the actual sequence, newest first, with searches in the customer's own words, and puts the summary on top rather than instead.
Build it or buy it?
| Approach | Good at | Watch out for |
|---|---|---|
| Customer data platform (CDP) | Collecting and unifying data from many systems at scale | Needs a data team to run it, and doesn't put anything on the shop floor by itself |
| Data warehouse and BI | Analysis and reporting across everything | Built for looking back, not for the associate serving a customer now |
| CRM as the hub | Contacts, activity and campaigns | Rarely sees website browsing, and depends on people typing things in |
| Record-first clienteling | Assembling the record from a website tag, order imports and store codes, and putting it on the floor | Built for one brand's retail customers rather than an enterprise's every data source |
The deciding questions are who maintains the join rules, how long until an associate sees a real customer's record, and whether the record reaches the people who serve customers or only the people who report on them.
Privacy and trust
A single customer view concentrates personal data, so it deserves care. The principles aren't complicated:
- Respect consent. Website tracking should follow the choices customers make in your cookie banner, including browser-level opt-outs such as Global Privacy Control.
- Keep only what helps. If you wouldn't be comfortable showing the customer a field, question why you have it.
- Scope access. An associate in one showroom rarely needs every customer in the business.
- Don't store what you don't need. A visitor can be recognised without their IP address ever being written down, for example by storing a salted hash instead.
This isn't legal advice; take that for your own situation. But a record built on these lines is also one customers recognise and don't resent.
Five tests for the view you have
Pick a real customer who has bought from you twice, through different channels, and try to answer these:
- Can you see both orders on one record?
- Can you see what they browsed before their first order — including before they gave you their email?
- Can you see which of your stores they visited, and who served them?
- Can you see which campaign first brought them to you?
- Could an associate see all of that on a tablet in the time it takes to say hello?
Most businesses get one or two. Each "no" is a join that isn't happening.
Where Stitchwork fits
Stitchwork builds a single customer view for showroom retail without a data project. One script tag on your website, behind your existing consent banner. Your orders by CSV or from your own checkout, with repeat imports safe. Every store, pop-up, franchise and concession counter as a channel with its own staff, country and currency. Browsing joins a customer retroactively when they identify; a second email address can resolve to the same person; a browser that identifies as two people raises a question for your team instead of guessing; and deleting a quote, a customer or a store doesn't break the thread.
Each customer's timeline shows web browsing, campaign arrivals, store visits, quotes and orders as one stream, newest first. The client IP is never stored, only a salted hash, and every workspace's data is isolated at the database. See how the record builds itself, or read about customers and identity in the documentation.
Questions people ask
What is a single customer view?
One record per customer containing everything the business knows about them across every channel: online browsing, store visits, quotes, orders, returns and notes from the team. It's the difference between knowing a customer and knowing three unconnected halves of them.
What's the difference between a single customer view and a CDP?
A single customer view is the outcome: one record per person. A customer data platform is one kind of software for producing it, typically by collecting and unifying data from many systems. You can have a CDP without a usable single customer view, and you can get a single customer view without a CDP.
Is an omnichannel customer profile the same thing?
Yes, for practical purposes. "Omnichannel customer profile", "unified customer profile", "customer 360" and "single customer view" all describe one joined-up record per customer across online and offline channels.
Do I need a data warehouse to build one?
Not for a single retail brand. A warehouse is useful for analysis across many systems, but a showroom business can get a working single customer view from a website tag, order imports and in-store identification, without building or maintaining a warehouse.
How do you handle customers with more than one email address?
Keep one primary record and record the other addresses as belonging to it, so orders under either land in the same place. When two existing records turn out to be one person, merge them deliberately, move everything to the surviving record, and keep an audit trail of who made the decision.