← Back to work
InsuranceOngoing

Five platforms. One operational picture. Twelve weeks.

Data scattered across five systems. Zero unified customer view. We built the operational picture — and then stayed to build everything else.

5

Data sources unified

12 weeks

Time to operational picture

Ongoing

Status

The challenge

An insurance agency was running its business across five separate platforms. Policy management in one system. Customer communications in another. Claims processing in a third. Commissions tracked in spreadsheets. Alert tasks scattered across email and a legacy CRM that nobody trusted.

Leadership had no unified view of their operations. Every decision required manually pulling data from multiple systems, cross-referencing in Excel, and hoping the numbers matched. They rarely did.

Agents in the field couldn't even close out basic alert tasks because the workflow spanned three different tools. Customer calls generated data in one system that never made it to the system where it was needed.

The engagement

We didn't start with technology. We started with two weeks of embedding — sitting with agents, observing workflows, understanding where information got lost between systems. The operational map we built during those two weeks became the blueprint for everything that followed.

Phase 1: The operational picture (Weeks 1-12)

The first priority was giving leadership a single source of truth. We built a unified data pipeline that pulled from all five platforms into a single warehouse, with reconciliation logic to handle the inevitable data conflicts between systems.

Twelve weeks after embedding, leadership had a real-time operational dashboard. For the first time, they could see policy counts, customer touchpoints, agent performance, and revenue metrics in one place without opening five different tools.

Phase 2: Agent workflows (Ongoing)

With the foundation in place, we turned to the agent experience. Alert task completion — which previously required navigating three systems — was consolidated into a single workflow. Agents could see their full task queue, customer history, and policy details in one view.

Phase 3: Commission tracking (Ongoing)

Commission structures in insurance are complex. Multiple carrier relationships, varying commission schedules, retroactive adjustments, split commissions between agents. The previous approach — spreadsheets built on assumptions — was leaving money on the table.

We rebuilt commission tracking from raw carrier data. No assumptions. No estimates. Direct reconciliation between carrier statements and policy records. The discrepancies we found in the first month justified the entire engagement.

Phase 4: Email intelligence (Ongoing)

The latest phase turns email chaos into structured data. Customer communications that previously lived in individual inboxes now feed into the unified pipeline. Policy changes requested via email get tracked. Renewal conversations get flagged. Nothing falls through the cracks.

The result

What started as a 12-week data unification project has become an ongoing partnership. The operational picture was the foot in the door. Commission tracking, workflow optimization, and email intelligence are where the real value compounds.

We're still there. That's the point.

Lessons

  • The real work starts after the first dashboard ships. The operational picture was necessary but not sufficient. Every subsequent phase built on it in ways we couldn't have predicted at the start.
  • Commission structures require raw data, not assumptions. The spreadsheet approach was wrong in ways that only became visible when we reconciled against carrier statements. You can't estimate your way to accurate commissions.
  • Unified data changes behavior. When agents can see the full picture, they make different decisions. The dashboard didn't just report — it changed how the business operated.
  • Ongoing presence compounds value. Each month we're embedded, we find new opportunities. Not because we're looking for them — because we understand the business well enough to recognize them.

Want results like these?

Start with an honest conversation about your data.