Most organisations do not have a customer data problem. They have a customer data agreement problem. The CRM holds one version of the account, the e-commerce platform another, the service desk a third, and the marketing automation tool a fourth — all technically correct, none of them reconciled. Ask a simple question such as “how many customers do we actually have?” and four systems return four answers.
A single customer view (SCV) is the discipline that resolves this. It consolidates every fragment of customer information scattered across the IT landscape into one certified record that all systems can consume — and, crucially, keeps that record current as the underlying sources keep changing.
What a single customer view actually is
A single customer view is a governed, continuously maintained record that represents one real-world customer across every system that touches them. It is not a report and not a dashboard. It is an operational asset: a golden record for the customer domain, complete with survivorship rules, a stable identifier, consolidated contact details, interaction history and the business rules that determine which source wins when two systems disagree.
That distinction matters. A 360-degree view that only exists inside a BI tool tells you what happened. A single customer view feeds the systems that act — the campaign engine, the service portal, the billing run, the AI agent drafting a reply. The value comes from distribution, not from visualisation.
SCV, CRM, CDP or MDM — where the boundaries sit
These four terms are routinely used interchangeably, which is where most SCV projects go wrong before a line of configuration is written.
| Capability | What it is built for | Where it falls short as an SCV |
|---|---|---|
| CRM | Managing and recording the commercial relationship — pipeline, activities, opportunities. | It is one source among many, not an arbiter between sources. Its data model serves sales, not the whole organisation. |
| CDP | Assembling behavioural and marketing data for activation across channels. | Optimised for marketing use cases. Records rarely flow back into ERP, billing or service systems as authoritative data. |
| Data warehouse / BI | Analysis, reporting, historical trends. | Analytical, not operational. Nothing downstream consumes a corrected record from it. |
| MDM | Creating and distributing certified master records across all domains, customer included. | Nothing — this is the architecture an SCV needs. It simply requires governance to be taken seriously. |
Put plainly: a CDP or a CRM can host a customer profile, but only master data management produces a record that every other system is contractually obliged to accept. If your single customer view lives inside a marketing tool, it is a marketing view — a good one, but not a single source of truth for the enterprise.
Build a trustworthy data foundation to enable AI-driven automation and reliable decision-making. Download the whitepaper to learn how to establish Data Governance as a strategic capability — and unlock scalable automation and AI.
The work an SCV performs on your data
Behind the concept sits a sequence of unglamorous data operations, each of which has to run continuously rather than once at go-live.
- Restructuring — records from different sources are remapped onto one shared customer model, so that “account”, “contact”, “client” and “party” stop meaning four different things.
- Normalisation — formats, address structures, country codes, name conventions and identifiers are brought into line with a defined standard.
- Cleansing and enrichment — incomplete records are corrected and completed from other internal sources or verified external references.
- Matching and deduplication — probabilistic and rule-based matching identifies records that describe the same person or organisation, and survivorship rules decide which attribute values survive the merge.
- Certification and distribution — the resulting golden record is approved, versioned and synchronised back to every consuming system, in real time, on an event or on a schedule.
The last step is the one most frequently underestimated. Deduplication produces a clean record; only reliable integration makes it useful. A perfect golden record that never reaches the call-centre agent’s screen has changed nothing.
What the business gets in return
The technical outcome is a unified customer profile. The business outcomes are more concrete.
Segmentation that holds up. Segments built on deduplicated data stop double-counting. Campaign reach figures become defensible, and the same customer no longer appears in two mutually contradictory segments.
Coherent contact across channels. A customer who has just raised a complaint should not receive an upsell email an hour later. That kind of failure is almost never a strategy problem; it is a data problem, caused by two systems that were never told they were dealing with the same person.
Personalisation that scales. Consolidating purchase history, browsing behaviour, service interactions and declared preferences into one profile is what makes genuinely contextual scenarios possible — a renewal offer timed to an individual consumption cycle, or e-commerce content that adapts to a customer’s actual relationship with the brand rather than to a broad persona.
An AI foundation that can be trusted. Generative models and AI agents are only as reliable as the records they read. Pointing an agent at four conflicting customer databases produces four confident, inconsistent answers. A certified customer record is the precondition for putting AI anywhere near a customer conversation — a point we develop further in our piece on data governance as a business capability.
Compliance that is demonstrable rather than asserted. GDPR obligations — right of access, right to erasure, consent management, retention limits — are almost impossible to honour reliably when customer records are duplicated across a dozen systems. With one authoritative record and documented lineage, a deletion request becomes a controlled operation instead of an archaeology project. Role-based access and full audit trails apply consistently rather than system by system.
See how certified customer records are created and distributed in practice.
How to implement a single customer view
An SCV is not deployed so much as negotiated. The technology decisions are comparatively easy; the agreement on what a customer is is where the effort goes.
- Map the systems that hold customer data. Every source, every attribute, every undocumented spreadsheet. A data catalog turns this from a one-off audit into permanent, maintained transparency about where each attribute originates.
- Define the use cases before the model. Deduplicating for a marketing campaign and deduplicating for regulatory reporting are different problems with different tolerances. Decide which ones the SCV has to serve first.
- Agree the customer data model and the survivorship rules. Which system wins on an address conflict? What counts as a match? Who signs off a merge? These questions are business decisions, and answering them late is the single most common cause of delay.
- Select the platform against a written specification. Evaluate a shortlist of vendors in depth rather than a long list superficially, and test with your own data rather than the demo dataset.
- Roll out with governance in place. Configuration, stewardship workflows, approval paths and a genuine test phase — not a pilot that quietly becomes production.
- Make the SCV the default. Every customer-related initiative — segmentation, reporting, new application onboarding — routes through the SCV. The moment a project is allowed to build its own customer table, fragmentation restarts.
- Extend it. Real-time API access for consuming systems, AI-assisted matching for low-confidence cases, and monitoring so that data quality degradation is visible before the business notices it.
Embedded in a CDP, or independent?
It is entirely possible to run a single customer view from within a customer data platform or a CRM, and for organisations whose customer data challenge is genuinely confined to marketing, that can be the right call.
The limits appear as soon as the record needs authority outside that perimeter. A profile assembled inside a CDP is optimised for activation; it rarely governs what the ERP believes, what appears on an invoice, or what a service agent sees. An independent MDM layer inverts the relationship: the customer record is maintained centrally and the CDP, CRM and every other application become consumers of it. One version of the truth, feeding every process with the same quality, regardless of which application asks.
The architectural options behind that choice — registry, consolidation, coexistence, centralised — are covered in detail in our guide to data hub models for master data management.
Building your single customer view with SoftProject
A working SCV needs three things at once: certified records, transparent data foundations, and integration that actually moves the record to where decisions are made. The SoftProject platforms cover all three.
- Master Data Management creates and maintains the customer golden record — matching, deduplication, survivorship rules, stewardship workflows and automated synchronisation with CRM, ERP and domain systems.
- MyDataCatalogue documents the systems and sources feeding the SCV, so that the origin of every attribute is traceable — essential for audits and for GDPR evidence.
- dataspot. adds business-oriented data governance and end-to-end lineage, so that a golden record traces back to clearly defined, well-understood data rather than to an undocumented mapping.
- X4 BPMS connects the applications through an integrated ESB and more than 200 ready-made adapters, and lets you model and automate the surrounding processes graphically with BPMN 2.0 — the stewardship approvals, the onboarding flows, the service portals.
- Phoenix keeps those integrations and processes running under enterprise-grade execution governance as volumes and systems grow.
The result is a single customer view that does not stop at unification. It covers the full lifecycle of customer data — from the moment it enters the organisation to the moment it is legitimately erased — and it treats the business teams who rely on that data, and the architecture that carries it, as part of the same problem.
Edouard Cante is responsible for the strategic direction and further development of SoftProject’s product portfolio as Chief Product Officer. With a strong understanding of the market and a high level of innovative drive, he advances customer-centric solutions and ensures the company’s long-term competitiveness.
FAQ single customer view
What is a single customer view (SCV)?
A single customer view is one certified, continuously maintained record representing a customer across every system in the organisation. It consolidates data from CRM, e-commerce, service, billing and marketing sources into a golden record with a stable identifier, and distributes that record back to the systems that need it.
What is the difference between a single customer view and a customer data platform?
A customer data platform assembles customer data for marketing activation. A single customer view built on master data management produces a record that is authoritative for the whole organisation — including ERP, billing and service systems — and distributes it as the reference version rather than as one profile among several.
Is a single customer view the same as a customer 360?
They are closely related but not identical. Customer 360 usually describes the completeness of the view — all relevant data about a customer in one place. A single customer view adds the requirement of uniqueness and authority: not just all the data, but one agreed version of it that other systems consume operationally.
How does a single customer view support GDPR compliance?
It provides one authoritative record with documented lineage, which makes access requests, erasure requests, consent management and retention rules executable as controlled operations. Role-based access control and audit logging then apply consistently across the customer domain instead of being configured separately in each system.
How long does it take to implement a single customer view?
It depends far more on organisational agreement than on technology. Mapping sources and configuring matching rules is measured in weeks; agreeing the customer data model, the survivorship rules and the ownership of each attribute across departments is what determines the overall timeline. Scoping the first release around one or two concrete use cases is the most reliable way to shorten it.
Do we still need a CRM if we have a single customer view?
Yes. A CRM manages the commercial relationship — activities, pipeline, interactions. The SCV governs the identity and reference data underneath it. In practice the CRM becomes both a source for and a consumer of the golden record, working from certified data rather than from its own isolated version.
