CommBroker

Challenge

A critical platform had quietly become dependent on bookmarks, memory, and support calls.

Brokers used CommBroker to find policies, manage applications, and keep up with compliance-heavy processes. But the platform had become increasingly difficult to navigate. Content was duplicated. Information lived in the wrong places. The right answers often existed, but brokers could not reliably find them when they needed them.

In practice, the product had stopped behaving like the source of truth. Brokers were relying on personal workarounds and institutional knowledge instead.

Research

The key insight was simple: the problem was navigation, not the quality of the information.

I grounded the redesign in a multi-layered research program: heuristic review, content audit, Treejack IA testing, moderated concept evaluation, and usability testing. The data consistently pointed to the same thing. Brokers were not confused by the policies themselves. They were frustrated by how much work it took to locate them.

Strategy

Two concepts went into testing. The decision was made by evidence, not internal taste.

I developed two directions. One stayed closer to the existing structure. The other made a more decisive IA shift, consolidating fragmented destinations into clearer hubs. Testing revealed a sharp difference: the bolder structure delivered much higher discoverability and broker confidence, even if it asked users to learn a new model.

Solution

The redesign treated information architecture as the product, not a layer beneath it.

The final experience reorganised content around broker intent rather than internal ownership. I introduced a central Resource Centre as the primary entry point for policy, process, and forms, reduced navigation noise, and created more direct routes to applications, loans, and supporting tools.

Outcomes

The platform moved closer to doing what brokers assumed it should have done all along.

The headline number is 98 percent discoverability in concept testing, but the deeper outcome was a more reliable self-service experience. Brokers could find what they needed without escalating to support, relying on out-of-date bookmarks, or checking with peers to confirm policy details. That moved the platform closer to the self-service model the business actually needed.

Reflection

The most critical part of this work was resisting the urge to start with the UI.

This project reinforced that information architecture does not merely support the experience in enterprise products. It is the experience. The platform got better because the structure got better, not because individual screens became more visually polished.

It also created the conditions for later work. The Gen AI pilot that followed would not have been trustworthy on top of inconsistent, fragmented content. This redesign made future capability possible.