Research Enablement · 2026

Standing up a research practice on Apple TV

Timeline2026–Present · first cycle complete RoleResearch lead · wrote the study, coached the designer who moderated MethodsConcept testing · synthetic usability · research enablement Collaboratorsdesign lead · product designer · producer · product manager · legal team stakeholder · engineering
Snapshot

Four designers on Apple TV's enterprise tools were each running their own research with limited shared framework or common tools. I built those, with a process to go with them, using the AI skills my team had already governed, and proved it on a concept test for a new rights management platform. Six findings went into the next design round and the designer who moderated kept the skills.

6
Key findings
10
Recommendations
22
Synthetic findings, six confirmed

01 — ProblemFour designers, four ways of doing research

No researcher had worked in this space before, and the designers had not waited for one. Four of them each covered a different product and each ran their own research, with limited shared framework or common tools.

The design was already moving. The design lead had built a clickable prototype with AI, informed by an earlier contextual inquiry, and the product team wanted their picture of the business process checked against real users. Three things shaped what I could set up:

No precedent: nobody here had worked with a researcher, so skepticism was the honest starting position.
A design already in flight: research had to keep pace with it rather than pause it.
Thin resources: roughly one study's worth of time, across four designers and four products.

So the useful question was never what I could find for this team. It was what they could keep doing once I moved on.

The bet

I could run one good study for them, or leave them with a way to keep running their own. Only the second one survives me moving to the next thing, so the study had to double as training.

02 — RoleWrote the study, someone else ran it

I wrote the plan, the session guide and the findings report, with AI doing the mechanical work to keep us light and lean. A second designer moderated the sessions, deliberately not the design lead who had built the prototype. Coaching her through facilitation was the point rather than a convenience: she wanted the tools for her own research afterwards, and she got them.

03 — ApproachDid our picture of the process match the work?

The product team had documented the business process and wanted it checked. That made the process the subject of the study and the prototype the instrument for getting at it, rather than the other way round.

Two rules held the whole thing together:

Aggressive about saving time on the mechanical work.
Conservative about anything needing judgment, which stayed with me.

The planning and synthesis skills drafted and clustered. I decided what the findings were. Before any of it reached a participant, I ran synthetic users against the same prototype to see what a walkthrough would catch on its own.

Why synthetic first

Running it first was cheap insurance. A stale record or a mislabelled button is findable without a participant, and finding one in a session costs an hour I cannot get back.

How the synthetic pass worked

Background first: what the platform is for, who uses it, and what the team was trying to decide.
Synthetic users: built from the personas and journeys the design lead had already mapped.
The prototype: the HTML, plus screenshots of every modal and dropdown, since Claude cannot click through it.
One walkthrough each: a sales executive and a legal counsel, reacting to the interface in character.

Then four real sessions

Moderated concept testing with a clickable prototype, four participants recruited through the product owner. Each session opened on the documented business process so participants could mark where their own work diverged from it, then moved to the prototype. The product team sat in as observers, which turned out to matter more than the findings did.

04 — FindingsSix findings from four sessions

The last of them is what the product team had asked us to check, and not the answer they expected.

Rights was not top level: it lived inside a deal, and nobody found it.
Everyone had a workaround: all four had already built their own tracking system out of spreadsheets, Airtable and documents.
Deal type labels confused all four: three of the options read as the same thing.
Avails lists were built by hand: hours of manual assembly, with mistakes the user already knew about.
Rights checks went through one team: two participants could not verify anything themselves.
One design, several business models: the platform assumed one licensing workflow, while its users worked in several, with different fields, deal types and approval chains.

What the walkthrough could not see

The synthetic run had produced 22 findings before anyone sat down with a participant, eight from the sales persona and fourteen from the legal one. Set against the findings report:

Six of the 22 held up: every one of them interface level, the kind of problem a careful read of a screen can find on its own.
None of the six above came from it: those needed somebody to describe how the work actually runs, and they came out of the sessions rather than the prototype.
Two it called clean: it logged no friction on the control that tripped all four participants, and it found the section none of them could locate.

The split held all the way through. The two personas it ran as were the roles the design already assumed, so the broadest of the six was never available to it. Synthetic testing is only as good as the persona you can describe, and as a first pass over the low hanging fruit it earned its place, as long as someone with domain knowledge reads the output and catches what it invents.

05 — ImpactTen recommendations, and a process question

6
Key findings from four participants
10
Recommendations, prioritised for the team
22
Synthetic findings, six the sessions confirmed

Ten recommendations went to the team, split into strategic, near term and quick wins, each tied to the findings behind it. The business process gaps went over as change management work rather than design work, which is not what the team expected to be handed.

What it left behind

The skills stayed: the designer who moderated kept them and the method, and can run the next round without me.
Observers who became advocates: sitting the product team in the sessions did more for their appetite for research than any readout would have.
One of several rounds: this was the first, and it set the pattern the rest follow.

06 — EvidenceSelected artifacts

The plan and guide, the findings report, and both synthetic walkthroughs. Complete documents rather than excerpts, but redacted: product and workflow names are generalized, colleagues appear as their function, and the interface screenshots are blurred.

07 — ReflectionHaving a researcher was new, and that was the work

Having a researcher in this program space was new for the product team, designers, and stakeholders. There was some skepticism, and there were learning curves, but involving them as observers in the interviews really opened the product manager's eyes to how enlightening research can be.

The team was already adopting AI, so it made sense to pilot our research AI skills with this study. Since it was a small study, the AI skills melded perfectly to assist in generating an approach and conducting synthesis, all with my oversight. The designer who helped facilitate appreciated the extra tools to help her own research in the future, while the product team was excited to get insightful findings about the business process, especially the need for change management as they introduce a new rights management process.

Want to talk through this work?

Daniel Farooqi · Staff UX Researcher · Product and team names generalized for confidentiality.

{{ modalTitle }}{{ modalSub }}
Shared on request

This is real work for a real employer, so I share the full documents with people I'm talking to rather than posting them publicly.

Email me and I'll send you a code. If you have one already, enter it below and it will remember you on this browser for 30 days.

{{ codeError }}

[email protected]

{{ modalNote }}
{{ pg.alt }}