Reframing the problem
If we only ask: How can a large set of test records become a report?
The more useful question: Which moments show where a problem occurs and support a concrete product decision? The unit of organization becomes a traceable issue.
What needed solving
Twenty-four days, six cities, 60 tests and interviews, and 3,600 recorded minutes. After collection, the product team still needed a clear answer: what should change, and why?
I contributed to research design, fieldwork, organization of the records, and analysis. The 254GB of recordings and roughly 600 pages of working material were inputs. Turning them into useful grounds for a product decision was the next task.
How I reframed it
A page-by-page list can help fix a release, but it does little to explain recurring difficulties. I worked from events within a task: what state the person encountered, what they did, and what followed.
Financial tasks need careful interpretation. A pause may mean confusion, or it may be a deliberate check of the consequences. Treating both as inefficiency can lead to the wrong change.
What I did
I separated observable events, possible explanations shared across events, and implications for the product. Observation, interpretation, and recommendation stayed distinct.
The study produced 86 decision-ready issues, organized under navigation, feedback, visual design, consistency, terminology, and software faults. Each issue connected the task, observed behavior, evidence, participant explanation, and priority.
Delivery and limits
The output was a set of locatable issues and a basis for discussing priorities. Public material does not include post-release business measures, so I cannot claim a conversion or revenue improvement.
The 60 tests and interviews count research encounters, not the effective sample behind every issue. Aggregate figures come from my 2019 public talk and the existing project-record account. Client identity, interfaces, and individual findings remain private.
- Source recordsKeep tasks and context
- Tasks & observationsSeparate actions and interpretations
- Traceable issuesEach finding has a source
Records → Tasks & observations → Traceable issuesConceptual model