Most organizations hold the same customer in five systems and treat them as five people. Customer Insights does two related jobs: it unifies those fragments into one profile, and it orchestrates journeys that respond to what the customer actually does rather than to a schedule set six weeks ago.
Customer Insights combines a customer data platform with real-time journey orchestration. These are the four problems that combination exists to address.
Different identifiers in the CRM, the e-commerce platform, the service system and the events tool. Nobody can say how many customers you actually have, let alone what any one of them is worth.
A batch goes out on Tuesday regardless of what the recipient did on Monday. The message that would have mattered arrives a fortnight after the moment it mattered.
Every list requires a data request, which takes days, which means campaigns are built around what is easy to extract rather than around who should receive them.
Different systems hold different opt-out states. Somebody who unsubscribed still receives messages from another channel, which is both a poor experience and, increasingly, a compliance problem.
Where the return actually comes from: The direct return is in campaign performance: relevance improves response rates, and triggered messaging outperforms scheduled messaging consistently enough to be predictable. The larger and slower return is in retention — knowing which customers are disengaging while there is still time to act on it. And there is a real cost saving that rarely gets counted: the data requests, list-pulling and reconciliation work that disappears when marketing can build its own segments against a profile everybody trusts.
Two products working together: the data platform and the journey engine.
Ingestion from transactional, behavioural and demographic sources with identity resolution, producing one profile per customer rather than several records that resemble each other.
Segments and calculated measures built by marketers in the interface, with Copilot allowing segments to be described in natural language rather than specified as queries.
Trigger-based journeys reacting to customer behaviour across email, SMS and push, so the message follows the action rather than the calendar.
Voice and SMS journeys where an AI agent built in Copilot Studio handles the conversation — proactive outreach for things like a cancelled appointment or a delivery confirmation, escalating to a person when needed.
Consent, quiet hours and channel preferences applied centrally so every journey respects them, which matters for both experience and regulation.
A unified interaction timeline passed to sellers and service agents, including predicted value and engagement history, so the next conversation starts informed.
The notable development is the convergence of marketing and service: journeys that integrate directly with Contact Center, so an automated outreach can become a live conversation with an AI agent and then a person, without the customer starting again. That is a genuine change in what is possible, and it makes the data foundation more important rather than less.
We baseline from your existing campaign performance, duplication rate and retention data.
Identity resolution across your source systems
Improvement from relevance and triggered timing
Marketing self-service instead of a data request
Improvement where disengagement is detected early enough
Preferences respected across every channel
Shared by marketing, sales and service
How to read these: How to read these: the figures above are typical ranges we plan and measure against, not guarantees. In your first engagement we agree the baseline, the target and the measurement method in writing, then report against them.
Microsoft's documentation frames the application's value in the following terms.
Where this comes from: Where this comes from: these themes follow Microsoft's Dynamics 365 Customer Insights documentation on learn.microsoft.com, covering both Customer Insights - Data (the customer data platform) and Customer Insights - Journeys (real-time orchestration), including conversational journeys with Contact Center. The numeric ranges above are ours and are planning figures rather than Microsoft benchmarks.
The definition of a customer differs, and so does what you do once you can see one.
One customer across store, web and phone, with lifetime value and next-purchase modelling rather than order-level thinking.
Household and entity relationships assembled across product systems, supporting pricing, cross-sell and concentration decisions.
Patient outreach for recall, preventive screening and care-gap closure, with consent and communication preferences respected.
Prospective students, current students, alumni and donors as one constituent, with communication coordinated across offices.
Donors, members, volunteers and participants in one profile, so an appeal reflects everything a supporter already does for you.
Client relationships across practice groups, with engagement history informing cross-sell rather than duplicating outreach.
The data foundation comes first. Orchestration built on unresolved identities produces personalised messages to the wrong person.
The unglamorous foundation that determines whether anything above it works.
The marketing capability, designed so the team can operate it without a data specialist.
Closing the loop between engagement and commercial outcome.
Implementation, customization, support and integration — with the data foundation treated as the majority of the work, because it is.
Environment design, tenant and licensing setup, configuration, data migration, testing and go-live — scoped to a fixed price and a fixed date, against outcomes agreed in writing before we start. For Customer Insights that means proving identity resolution against a sample your team can manually verify before anything is orchestrated on top of it.
Where the product stops short of your process, we extend it inside the platform rather than beside it, and we build it as configuration you can maintain wherever that is possible. Match rules, segment definitions, measures and journey templates configured as assets your marketing team owns and adjusts.
Managed support after go-live: a named team, agreed response times, release management for Microsoft's update cadence, and a backlog we work through with you. Ongoing match-rule tuning, journey performance review and the deliverability work that email programmes quietly depend on.
Connecting this platform to the systems you are keeping, with monitored, re-runnable interfaces and a documented contract for every field that moves. Ingestion from your CRM, e-commerce, service, events and finance systems, plus activation back into the channels you actually send from.
Identity resolution is where this succeeds or fails, and it is not a purely technical decision. Whether two records are the same person is a business rule with commercial consequences, and we will insist your team verifies a sample by hand before we build anything on top of the result.
The temptation is to start with the journey because it demonstrates well. The result is a beautiful journey delivered to a fragmented audience.
We map every system holding customer data and how identity is represented in each.
We build and test identity resolution, then have your team verify a sample manually.
We design journeys around real triggers, starting with one high-value scenario.
Consent, preferences, quiet hours and data retention configured and validated.
Journey performance and commercial attribution tracked against the baseline.
We ask your marketing team to manually verify a sample of resolved profiles before go-live. It is tedious and it is the only way to know whether the match rules are right, because an over-aggressive rule merges two customers and an over-cautious one leaves you exactly where you started.
Customer data platforms fail on the data, then get blamed on the platform.
Identity resolution is the majority of the work and the part most likely to be rushed. We treat it as the deliverable that everything else depends on, because it is.
Segments that require a technical request will not be used. Self-sufficiency for the marketing team is a design goal, not a training afterthought.
Preference handling across channels is both a customer experience question and a regulatory one, and it is easier to configure correctly than to retrofit after a complaint.
The unified profile is only valuable if sales and service can see it. Implementing marketing alone recreates the fragmentation you were trying to fix.
Two illustrative engagements showing the shape of the work.
The same shopper existed separately in the e-commerce platform, the loyalty system, the service desk and the events tool. Campaign targeting was based on whichever list was easiest to pull.
Renewal reminders went out on a fixed schedule regardless of engagement. Members who had stopped attending anything received the same message as the most active ones.
Practical pieces on unification, orchestration and measurement.
Why match rules need commercial judgement, and how to test them before you build on them.
Read moreWhat changes when messaging follows behaviour instead of a calendar.
Read moreCentralising consent so an unsubscribe means what the customer thought it meant.
Read moreGive us extracts from two or three systems that hold customer records. We will run identity resolution against them, show you how many duplicates exist and what a unified profile looks like, then demonstrate a triggered journey built on it.
Describe the situation in your own words.
