Dynamics 365 Customer Insights: Data Setup for Customer Zero

Purpose of a “Customer Zero” Setup in Customer Insights – Data

When a sales organization needs a working customer dataset but has limited or incomplete sales history, the priority is not advanced analytics. The priority is building a reliable unified customer view that can support internal segmentation, account assignment, and basic targeting. In Dynamics 365 Customer Insights – Data, this starts with connecting source data from Dataverse and then configuring unification and deduplication so the same customer is not fragmented across multiple records.

This approach is especially relevant for an internal “Customer Zero” sales team, where the immediate goal is to make the data accurate enough for day-to-day decisions, even if real transaction history is minimal.

High-level workflow: From data sources to a unified customer view

The setup flow typically follows a strict order:

Add data source
Confirm change tracking
Select tables
Set primary keys
Define deduplication (entity resolution) rules
Configure match rules (linking)
Build the unified data view
Run unification and complete

Following this sequence matters because unification accuracy depends on the quality and structure of the ingested tables and the consistency of the matching keys.

Connecting Dataverse as the data source

After the Customer Insights – Data installation is complete in the target environment, the first practical step is to connect customer-related data to the Data module.

Within the Customer Insights – Data interface:

Open the Data experience and select Add data source
Choose Dataverse as the source type when the customer entities live in Dataverse (common for Dynamics 365 Sales implementations)
Provide a data source name and the correct Dataverse server/address
Proceed to the next step to select the tables to ingest

Selecting a minimal table set for internal sales readiness

For a “Customer Zero” team, it is usually better to ingest only what supports identity resolution and near-term sales operations. A practical starting selection can include:

account (company or organization master)
opportunity (sales pipeline context, if available)
contact (primary identity for people)
activitypointer (activity events)
task (task-level interactions)

An initial full sync can take time depending on volume and environment performance. The key validation step is confirming that records appear correctly in the `data view` after ingestion.

Unification configuration: primary keys first

Unification is where Customer Insights – Data consolidates records into unified entities. The process begins by defining primary keys per table. This ensures that the system can track identity across refreshes and apply deduplication rules on a consistent basis.

From the left navigation, select Unify, then map primary keys for each ingested entity.

Deduplication for Accounts: building trustworthy company identity

Deduplication rules define when two records should be treated as the same entity. For the Account entity, a typical starting point is to compare a combination of fields that describe the company, such as:

Company name match (often accountname)
Phone number match
Email domain match
Address match

Because name-based matching alone is vulnerable to variations (punctuation, legal suffixes, casing, spacing), the setup should also include text normalization rules.

Normalizing account names to improve matching accuracy

Normalization pre-processes text before comparison so that visually different but logically equivalent values can match. For the account name field, recommended normalization settings typically include:

Unicode to ASCII (helps standardize full-width versus half-width differences, which matters in multilingual datasets)
Symbols removal (for example, removing punctuation characters found in names like “Co., Ltd.”)
Text to lowercase (reduces case sensitivity issues)
Whitespace normalization (removes extra spaces)
Type normalization for legal entity suffixes (such as Co., Ltd., Inc., Corp.)

These settings reduce false negatives in entity resolution and improve stability across data sources and manual entry styles.

Match rules (linking): connecting entities into a unified customer picture

After deduplication, match rules are configured to link related entities. This step is essential for internal sales workflows because it determines how contacts, activities, and opportunities connect to the same organization identity.

Well-designed match rules ensure the unified profile supports realistic questions such as:

Which contacts are associated with a specific account?
Which tasks or activities belong to that customer identity?
Which opportunities should appear for the unified account-person relationship?

Running unification and validating outcomes

Once primary keys, deduplication rules, and match rules are defined, the unified data view can be created. The final step is running the unification process and completing the build.

Validation should focus on measurable outcomes:

Data sync health: data loads correctly and refresh behavior is active
Unification quality: duplicates reduce compared to raw source tables
Usability for sales: created segments or views return expected record counts

Best-practice rollout for a “Customer Zero” sales team

A pragmatic rollout approach reduces risk:

Ingest customer identity entities first (contacts and leads are often the core starting point)
Configure unification around stable identifiers like email and phone where available
Create a small set of baseline segments and confirm they populate correctly
Only then expand ingestion to additional tables and behavioral signals

This approach helps ensure the internal sales team can act on consistent, unified customer profiles even before deep historical sales data exists.

Key takeaways

Start with a controlled Dataverse ingestion scope suitable for identity resolution.
Define primary keys before building deduplication and matching rules.
Use normalization to improve company name matching and reduce duplicates.
Validate unification results by checking record counts and duplicate reduction.