Case study

Fragmented customer data, moved to one warehouse that watches churn

Internet and IT

An internet company whose customer data was spread across business units. It now sits in one Snowflake warehouse, with dashboards that monitor usage and churn.

Read the engagement brief(opens in a new tab)
Period
9 months
Company
250–500 people
Churn
20% lower
A schematic: the separate reports this engagement started from, the one consolidated reporting layer that replaced them, and the technologies underneath.

Scattered data, replaced by one central warehouse

Customer data sat separately in each business unit, so nobody saw how customers used the product as a whole. We moved it into one Snowflake warehouse and built dashboards that show churn as it develops.

lower customer churn
20%
central warehouse
1
months of engagement
9

Anonymised by sector and size, because the confidentiality that protects this client would protect you too. Sector and outcomes are real; identifying detail is not. The attrition figure is as recorded at the end of the engagement.

Before and after

Before and after the engagement

Scattered data

  • Data spread across business unitsEach unit with its own copy
  • No single view of usageChurn seen after it had happened

One warehouse

  • One Snowflake warehouseThe central reporting environment
  • Usage and churn dashboardsRefreshed automatically
From customer data held separately in each business unit to one warehouse the reporting reads.

The engagement in full

Client
An internet company of 250 to 500 people with several business units.
Situation
Nobody could see how customers used the product as a whole, so churn was counted after it had happened rather than watched while it developed, which leaves nothing to act on.
Mandate
Migrate fragmented customer data into one cloud warehouse as the central reporting environment, and build the usage and churn reporting on it.

From usage to action

Churn monitoring turned usage data into a signal the business could act on, which contributed to 20% lower churn.

  1. Usage dataFrom every unit, in one warehouse
  2. Behaviour dashboardsHow customers use the product
  3. Churn monitoringCustomers at risk become visible
  4. ActionTaken before customers leave

The architecture

Business units: Unit data · Customer records. Ingestion: Azure Data Factory · Databricks. Warehouse: Snowflake. Reporting: Usage dashboards · Churn monitoring
Data from each business unit loads through automated pipelines into one Snowflake warehouse, which the usage and churn dashboards read.

Architecture decisions

DecisionRejected alternativeWhy
One cloud warehouseReporting per business unitCustomer behaviour can only be read across the whole business from one place.
Automated ETL pipelinesExports gathered by handLoading every unit the same way keeps the warehouse current.
Churn as a monitored measureChurn counted after the factA dashboard that shows it as it develops gives time to act.

What the client has today

  • One Snowflake warehouse as the central reporting environment, in place of data held per business unit.
  • Automated ETL pipelines loading data from every business unit.
  • Automated Power BI dashboards for customer usage and behaviour.
Where it left them
Customer behaviour is read from one warehouse, and churn is watched as it develops rather than counted afterwards.
Stack
  • Snowflake
  • Databricks
  • Azure Data Factory
  • Power BI
  • SQL
All case studies

The other engagements

  • Internet and IT

    One set of subscription metrics for finance, product and customer teams

    An internet company of 50 to 250 people with subscription revenue, whose finance, product and customer teams each kept their own spreadsheets.

    • Power BI
    • DAX
    • SQL
    • Dataflows
    • Azure Data Factory
    • Semantic modelling

  • Retail

    Foundation then Activation then Acceleration alongside Stewardship

    Store revenue, executive questions and forecasting on one set of definitions

    A retailer with Microsoft Fabric already in production, semantic models over Direct Lake and store revenue streaming in, who wanted forecasting, Copilot questions and anomaly detection on top of it.

    • Microsoft Fabric
    • Power BI semantic models
    • Direct Lake
    • Copilot in Power BI
    • Fabric AI Functions
    • Activator

  • Industrial manufacturing

    Foundation then Activation then Acceleration

    Governed reporting across a production network

    An industrial manufacturer reporting across a production network of sites, each with its own view of what was running.

    • Microsoft Fabric
    • OneLake
    • OneLake Security
    • Direct Lake
    • Spark
    • MES
    • OPC UA
    • Power BI
    • DAX
    • Azure Key Vault
    • Purview
    • Azure DevOps

  • Hospitality and real estate

    Foundation then Activation

    A CFO cockpit over a group that closed its books separately

    A holding group across hospitality and real estate, with entities that each closed their own books and a CFO who had to answer for all of them.

    • Microsoft Fabric
    • OneLake
    • Direct Lake
    • Databricks Mirrored Catalog
    • Fabric Data Factory
    • Power BI
    • DAX
    • Calculation groups
    • Row-level security
    • Azure DevOps
    • Azure Key Vault
    • Purview
All case studies

Free consultation

Know where your data stands in an hour.

A free 60-minute consultation. No pitch, no invoice. You’ll leave knowing whether your data is ready for AI, and whether we’re a good fit.

Book your free consultation(opens in a new tab)