Brand Protection Platform

Both legacy platforms had entrenched internal advocates. The consolidation was as much a political challenge as a design one. One platform was powerful but required weeks of training and heavy manual work. The other had design intent but lacked capability. Neither was what analysts actually needed.

Two platforms, two camps, one designer across six teams

Corsearch was the result of an acquisition merging two brand protection companies. Talisman was powerful but unintuitive — analysts spent weeks learning it and routinely completed tasks in Excel and email rather than the platform. Zero had more design intent but significantly less capability. Enterprise clients were split between both, requiring analysts to maintain proficiency in two systems simultaneously.

Internally, stakeholders from each legacy company actively advocated for their own platform. I joined as the only designer, responsible for the entire design output across 6 engineering teams and 4 product owners — covering marketplace enforcement, domain enforcement, AI features, reporting, admin configuration, and platform tooling.

Letting go of existing workflows to find the ideal one

We interviewed 12 users — six from each platform — across different seniority levels. The key question in every session: describe your ideal workflow if there were no technical, legal, or platform constraints at all. This approach bypassed years of accumulated workarounds and produced a workflow that was structurally better than anything either platform had built.

Google Analytics and Pendo data confirmed what interviews surfaced: a significant portion of enforcement work was happening outside the product entirely — in spreadsheets, external databases, and email. The research also uncovered industry fragmentation within the analyst population. Fashion and consumer goods analysts identified counterfeits by image. Regulatory and pharmaceutical analysts worked almost entirely from text. We also found unexpected change resistance — analysts who had used a platform for years were reluctant to learn something new even when the new approach was demonstrably faster.

Using data to end the platform debate

Senior stakeholders from both legacy companies regularly referenced their own platform as the correct approach. We established a consistent principle: include users from both platforms in every research session, present findings to the full stakeholder group, and let the data drive the decision.

Speaking up when senior managers were framing problems incorrectly required conviction. But it built credibility. By the end of year one, research-led decision making had largely displaced legacy-platform-as-reference.

Encoding expertise into the workflow and building for two analyst types

The platform's biggest usability problem was not the interface — it was that effective use required legal knowledge most analysts did not have. Analysts without IP law backgrounds were miscategorising infringement types, causing enforcement letters with the wrong legal basis. The solution: auto-labelling based on scraped data, guided categorisation at decision points, and a review step allowing analysts to send items to a partner queue before committing.

For the industry fragmentation: visual-heavy industries got AI image recognition in the primary view — logo matching, counterfeit image comparison, visual similarity scoring. Text-heavy industries got keyword highlighting, description analysis, and brand name variant detection. Same enforcement workflow, different AI assistance surface. With no inherited components from either legacy platform, I built the design system from scratch — every component matched to the engineering component library to keep design and build in tight sync across six teams. Storybook documentation was underway at handoff, alongside a design-to-code automation pipeline. Token implementation was scoped for the next phase but did not ship during my tenure — a known gap in the system.

Holding the product together when the team fell apart

During year one, we lost several senior PMs and POs who held deep knowledge of both legacy platforms. Incoming PMs were starting from close to zero on platform context. Requirements came in that did not account for global platform logic, and problems were framed at surface level when the structural issue was something deeper.

I stepped into a role that went well beyond design — working directly with engineering leads to validate coherence, flagging when requirements would create downstream problems, and writing design briefs that also functioned as product briefs. Maintaining direct user relationships meant critical questions could be answered in hours rather than weeks.

4.6/5 satisfaction, 42% faster enforcement, promoted within the role

Beta users rated the platform 4.6/5 — a significant improvement on either legacy system. Enforcement time per listing reduced by 42% through automation and AI tooling. NPS reached 42. I was promoted to Senior Product Designer within the role.

Looking back: we built for all industries simultaneously from the start, which required flexibility that added complexity everywhere. The better approach would have been to fully solve one industry first — validate the approach — then extend systematically. I should have pushed harder on that sequencing argument earlier.