Kat McGowan Learning & product design

← UX & product

  • Strategy
  • Discovery
  • AI

Brand intelligence: surfacing hidden patterns to prevent repeat mistakes

My role
Lead designer
Responsibilities
End-to-end UX & UI, discovery, mapping, design system
Collaborators
QA production, production leads, creative directors, customer teams, engineering, PMs
Timeline
3 months

Overview

Our production teams were making repeated mistakes, despite having brand guidelines. One team proposed a solution — a new brand guidelines hub — but I questioned whether that would solve the real issue.

After revisiting the research, I reframed the problem and proposed a smarter approach: surface brand-related insights directly from real project data (feedback, revisions, emails). This concept became the new direction we aligned on as a team.

The final Brand Intelligence proposal shown inside the production tool, with surfaced brand insights in a side panel.
Brand intelligence final proposal.

The problem

Despite having access to brand guidelines, our production teams were still making the same brand mistakes — often only caught late in the process, resulting in repeated revision rounds and low client satisfaction.

The goal

Make sure everyone working on a brand — designer, QA, or creative lead — had access to the most relevant, up-to-date brand information at every stage of the creative process.

Context

This project was handed to me after the initial solution — a revised brand profile — had already been decided on. I was tasked with mocking up the UI. While I pushed back on the value of it, I agreed to move forward in order to have visuals to centre the discussion around.

The initial proposal: a revised brand profile screen listing brand guidelines and assets.
Initial proposal — a revised brand profile.

Reframing the problem

Hesitation from the team around the value of an updated brand profile gave me room to pause and reframe. Instead of assuming ‘we need better guidelines’, I took a step back and asked:

When do mistakes happen, and what context is missing in that moment?

After reviewing internal workflows and project data, I realised the issue was timing and visibility, not a lack of documentation.

The concept

What if we used the data we already collect — revision comments, emails, and phone calls — to surface common brand issues before they happen?

I mocked up a concept that:

  • Flagged recurring issues automatically (e.g. tone, logo use, framing)
  • Integrated directly into production tooling
  • Used AI to cluster and highlight patterns over time
An early concept screen showing recurring brand themes clustered from revision comments, surfaced to production users.
Early concept showing recurring themes for production users.
A revision comment being selected to become the basis of a new brand rule.
Selecting content to add as a new brand rule.
The add-rule interface, turning a recurring issue into a documented brand rule.
Adding a new brand rule.

User feedback

After reviewing the initial concept with users, we identified adjustments that would improve the output and its usefulness:

  • Let teams mark an issue as helpful for this concept, or not
  • Insights should be reviewed and summarised by a creative director, then shared with designers, to avoid ambiguity
  • It should be easy to dismiss or hide irrelevant insights, to reduce noise
  • Teams wanted to see a history of revision insights at any time, including those marked not relevant
  • We’d need to keep tweaking the AI prompt as the feature was used, to ensure the themes stayed accurate
The final Brand Intelligence concept shown in the review stage of the production workflow.
Final concept of Brand Intelligence in the review stage.

Outcome

With a new iteration of the concept, I shared it with the PMs, the original designer, engineers, and stakeholders. The feedback was positive and the value was clear, from both a business and a user perspective. This resulted in:

  • Buy-in to shift direction toward a smarter, insight-driven solution
  • Alignment of two teams around a more scalable, embedded way to prevent brand errors
  • The revised approach now being scoped for implementation

Reflection

This was a reminder that the first solution isn’t always the right one — and that sometimes your real value as a designer is in helping teams slow down, reframe, and re-align.