GUIDES
Single customer view: what it is and why retailers need one
.png)
A single customer view is one profile per real person, built by merging every record that belongs to them: online orders, in-store purchases, loyalty membership, email and SMS activity, service conversations and reviews. Instead of six partial records scattered across six systems, the business holds one.
Single view of customer benefits are extensive and cover a wide range of services and scenarios in which organizations may require deeper insights to improve the customer experience. The primary advantage of this approach is that it allows organizations to quickly access customer data, eliminating the need for them to track down and verify the information. This is a great way to streamline data across your entire company and ensure that everyone’s on the same page with the same information.
Likewise, maintaining a single customer view can make it easier for you to understand customers with more precise segmentation. Data is most valuable when it’s specific, and by maintaining accuracy via a single view, you can develop a more detailed classification system over time, allowing you to better meet the needs of different types of customers. Another key benefit of the single view approach is that you can provide better service to customers. By performing deeper data analysis, and connecting information from various systems, you can get a complete look at the customer experience and identify areas for improvement.
What sits inside a single customer view
The profile is only as good as what feeds it. In order to have accurate single customer profiles, you need to ingest and unify every source of data that captures customer information. While this varies for different brands, the most common data sources for retail companies are the following five.
Transactions
Every order, online and in store, with line items rather than just a total. Line items matter because they carry product, category, size, discount and margin. A customer who spends $400 a year at full price is a different customer from one who spends $400 a year at 40% off, and only the line items tell you which is which.
In-store and point of sale
The hardest source, and the one that separates a real single customer view from a website view. Point of sale (POS) data arrives without an email address most of the time. Linking it takes loyalty scans, receipt capture, staff-entered details or card tokens.
Loyalty
Usually the cleanest identity anchor available, because the customer volunteers who they are. Loyalty is often the bridge between the in-store shopper and the online one.
Marketing and web
Email opens and clicks, SMS replies, site sessions, abandoned carts. Rich, but anonymous until it's tied to a known person.
Service and reviews
Support tickets, returns history, product reviews. Low volume, high signal. A customer who has lodged two complaints and a return reads differently from one who hasn't.
The benefits, in the order they show up in revenue
You stop counting the same person several times
Deduplication sounds like housekeeping. It isn't. A retailer with a 20% duplication rate is reporting a customer base 20% larger than it has, an average spend lower than reality, and a repeat purchase rate that understates loyalty. Every number downstream is wrong, and every plan built on those numbers is wrong with it.
Segments start matching reality
Segmenting on ecommerce data alone finds your online customers. It misses the person who browses online and buys in store, who in most omnichannel retailers is a large and disproportionately valuable group. Once the profile is whole, segments describe actual shopping behaviour rather than one channel's version of it.
Churn becomes visible before it happens
Churn in retail is a slow fade, not an event. You can only spot it by comparing a customer's current gap since last purchase against their own established purchase cycle. That calculation needs a complete purchase history. Split the history across channels and the maths breaks.
Margin enters the conversation
With product cost attached to line items, the profile carries gross margin per customer, not just revenue. That surfaces the uncomfortable cases: the high-revenue customer who only ever buys on promotion and contributes very little.
Store teams get context at the counter
Once the profile is whole, the person serving a customer in store can see what they bought online last month. That's a service difference customers notice.
Why most retailers still don't have one
Four reasons, roughly in order of how often they're the culprit.
- The systems were bought separately. Ecommerce platform, POS, loyalty and email were each chosen by different people at different times for different reasons. None was chosen for how well it shares data.
- Identity is genuinely hard. Same person, three email addresses, two phone numbers, a maiden name and a work address. Matching that correctly needs rules, not a spreadsheet.
- Nobody owns it. Marketing wants the segments, IT owns the pipes, finance owns the margin data, and retail operations owns the POS. The single customer view sits between all four.
- The first attempt was a data warehouse. Warehouses store data well. They don't resolve identity, and they don't hand marketers a usable audience.
How identity resolution actually works
- Deterministic matching uses exact shared values. Same email, same phone, same loyalty number. High confidence, low reach. It catches the easy cases and nothing else.
- Probabilistic matching scores likely matches on combinations of weaker signals: name plus postcode plus card token, for example. Broader reach, and it needs a confidence threshold you're comfortable defending.
- Survivorship rules decide which value wins when two records disagree. If one says the customer lives in Fitzroy and the other says Carlton, something has to break the tie. Usually the most recent verified value, but the rule should be explicit and auditable.
Get these three right and the profile holds. Get them wrong and you either merge two different people, which is a privacy problem, or fail to merge one person, which is the problem you started with.
What can build a single customer view
Enterprise customer data platforms
Segment, Tealium, mParticle and Amperity all resolve identity well. They're built for data engineers, implementations commonly run 12 to 18 months, and at the end you have infrastructure rather than retail answers. Strong fit for large organisations with a data team already in place.
Marketing automation platforms
Klaviyo, Emarsys, Bloomreach and Dotdigital hold a customer profile, and for ecommerce-only brands it may be enough. The gaps are in-store purchases and product cost. They're built to send, not to unify.
Data warehouses with a modelling layer
Snowflake or BigQuery with dbt and a reverse ETL tool such as Hightouch. Flexible and durable. It also needs engineering to build and engineering to change, which puts a queue between marketing and every new segment.
Retail customer data platforms
Narrower by design. Fewer industries served, more retail logic built in: POS ingestion, loyalty matching, product hierarchies and margin.
Master data management tools
Built for product and supplier data originally, sometimes stretched to customers. Rarely a comfortable fit for marketing use.
Where Lexi fits
Lexi is the agentic customer data platform built for omnichannel retail. Identity resolution sits at the centre of it: Lexi ingests transactions, inventory, POS, loyalty, reviews and web signals, resolves them to one customer, and holds the product cost data that makes margin per customer possible.
What's different is what happens next. Ask Lexi a plain-language question about the unified profile and it builds the answer from the data, shows the working, then builds the segment and pushes it to the tools your team already uses. No SQL, no ticket, no waiting on an analyst.
One furniture retailer's CRM manager spent several days each month pulling board reporting and building segments by hand. That work now happens in a single conversation.
Lexi runs inside AWS Bedrock. Data never leaves the platform, no personally identifiable information goes into AI processing, and Lexer is SOC 2 certified.
How to tell it's working
Although every retail brand will have their own measures of success, these four metrics often indicate that a single customer view is having a positive impact.
- Duplication rate. Profiles merged as a share of raw records. Expect a large one-off drop, then a low steady rate.
- Match rate on in-store transactions. The share of POS transactions attached to a known customer. This is the honest test of whether the view is genuinely omnichannel. Under 40% and you have a website view with some shop data bolted on.
- Segment overlap. How much your online-only segments and your true segments differ. A big gap means you were marketing to the wrong people.
- Time to build a segment. From question asked to audience live in the sending tool. This is the number the marketing team feels.
Related Articles
📄 Customer Data Platform (CDP)
📄 Customer Experience Platform
📄 Digital Transformation in Retail
📄 Customer Intelligence Platform
📄 Omnichannel Customer Experience
📄 Customer Data Platform Architecture
📄 Customer Segmentation in Retail
📄 Customer Experience in Retail
Common questions
What is a single customer view?
A single customer view is one profile per customer, built by merging every record that belongs to that person across ecommerce, point of sale, loyalty, email, SMS and service. It replaces the partial versions held in each separate system with one complete history of what someone has bought, browsed and asked for.
How long does it take to build a single customer view?
With a purpose-built retail platform, first unified profiles usually appear within weeks, because the connectors and matching logic already exist. Enterprise customer data platforms and warehouse builds typically run 12 to 18 months, since identity rules, pipelines and the modelling layer all have to be built before anything is usable.
What is the difference between a single customer view and a CRM?
A CRM records your relationship with a customer: contacts, conversations, and often sales activity. A single customer view records their behaviour, merging purchases and interactions from every channel into one profile. The CRM is usually one of the sources feeding the single customer view rather than a replacement for it.
See how Lexi helps retailers drive more sales.
Leading retailers unify their customer data, build high-value audience segments, and grow lifetime value with Lexi.
Book a demo