I built a reusable platform that performs a detailed usability test on any intranet site. It turns site content, staff feedback, and institutional constraints into evidence-based findings, giving stakeholders practical results they can review, question, and implement at scale.
The input was broad and uneven: workshop notes, post-it comments, email suggestions, likes, contributor names, and follow-up additions. Some comments were strategic, some were tiny navigation fixes, and some described missing knowledge assets.
Stakeholders needed to see what the feedback meant without losing the original voice of staff. A simple summary would have been too thin; a raw database would have been too hard to act on.
I created a repeatable usability-testing platform: capture site and feedback evidence, assess it against usability standards, translate findings into practical intranet changes, and package the result as a detailed test report plus implementation roadmap.
I kept contributor names, original wording, vote/like signals, and session context. This makes every usability test credible: stakeholders can inspect the evidence behind each finding.
Rather than organizing the work by department or page ownership, I grouped notes around jobs staff were trying to accomplish: finding tools, locating experts, accessing publications, learning from projects, and understanding processes.
Each cluster was translated into concrete implementation patterns: knowledge-base entries, filtered lists, quick links, FAQ structures, dashboard summaries, content ownership, review cadence, and search/AI discoverability checks.
The solution deliberately has different layers: raw notes for auditability, synthesis cards for comprehension, a timeline for execution, and AI querying for exploration. That prevents one page from trying to do everything.
The AI pieces are designed to help stakeholders ask questions of the feedback and generate briefings. The underlying evidence remains visible so the AI output can be challenged and checked.
The final artifact is a scalable usability-testing platform, not just a mockup. For any intranet site, it identifies what to fix, why it matters, what evidence supports it, how hard it is, who needs to own it, and how the work can be sequenced.
| Layer | What it does | Why it matters |
|---|---|---|
| Evidence database | Stores raw staff notes, contributors, themes, and like signals. | Keeps the synthesis honest and lets reviewers trace recommendations back to evidence. |
| Cluster synthesis | Turns many small comments into nine coherent design problems. | Helps stakeholders understand patterns without reading every note individually. |
| Implementation roadmap | Sequences the work by timing, effort, priority, and dependencies. | Moves the conversation from feedback collection to delivery planning. |
| AI inquiry layer | Lets reviewers interrogate usability-test results and generate evidence-grounded briefings. | Supports consistent, detailed usability tests across many intranet sites while preserving the underlying source material. |
The important shift is from “we collected feedback” to “we can run a detailed usability test at scale.” The platform applies a consistent testing method to any intranet site and gives stakeholders a practical operating model for prioritizing, assigning, and maintaining improvements.