Customer data · Compliance · Activation

Farmers Insurance · via Capgemini

Real-time suppression and consent logic across 9M+ customer records

An urgent customer-data sprint connecting consent, identity, activation, and measurement across a complex enterprise stack.

9M+customer records in the suppression pipeline
20+concurrent workstreams structured
Real timesuppression and ad-platform activation
1governed taxonomy ownership model

01 · The challenge

What had to change

The environment spanned Invoca, Twilio Segment, Google Ads, GA4, Snowflake, FullStory, OneTrust, Braze, and multiple delivery teams. The immediate requirement was suppression, but the deeper issue was an operating model for trustworthy identity and activation.

02 · The system

Architecture and operating components

ARCHITECTURE VIEWFarmers Insurance · via Capgemini operating system
01Consent + identity30+ domains
Farmers InsuranceCapgemini
02Customer dataCanonical events
Twilio Segment
03Governed pipeline9M+ records
SnowflakeSQL / PostgreSQL
04SuppressionReal-time controls
Google
05ActivationPersonalization
Braze
Open the implementation sequence
  1. 01

    Introduced an Agile delivery rhythm and structured more than 20 concurrent workstreams in a shared prioritization system.

  2. 02

    Designed and deployed an incremental Snowflake pipeline for 9M+ records, with CCPA/LDU logic and automated S3 export batching into Meta and Google Ads.

  3. 03

    Audited the CDP across source mapping, consent, identity resolution, MTU cost optimization, and layered data architecture.

  4. 04

    Created a Master Taxonomy ownership and change-control model after a taxonomy break affected Power BI channel reporting.

03 · The outcome

What became possible

  • Connected compliance controls directly to paid-media activation instead of treating them as a downstream review step.
  • Restored a clear path to reporting integrity through governed taxonomy ownership and pipeline controls.
  • Produced a practical roadmap for consent, identity, personalization, and data-cost optimization.
Forward signal

The strongest customer-data systems make consent and taxonomy operational inputs—not documentation that sits beside the pipeline.

Next case

Verix.io

A MarTech operating system rebuilt from CRM taxonomy to activation