Health tech · Part 2 of 11 · November 2026 · 6 min read
Health tech · Part 2
Your best product research is in the Q&A box
Every expert session is a focus group you forget to keep.
Every live expert session is a focus group that most companies forget to keep. The questions people ask a sleep scientist, a dietitian or a menopause specialist show exactly where the product's explanations are failing, in their own words, before anything has gone wrong.
The Adelo teamAdelo Insights
Published 3 November 2026
An Adelo Insights essay on health tech and the expert voice
The clearest description of your product is in the questions people type when an expert on your live session asks, "What would you like to know?" Your website describes what you hope the product means; those questions show what people think it means. Most companies let them scroll past and never look again.
1
What questions show that complaints miss
Every health tech company already has feedback channels: support tickets, app store reviews, a community forum, the replies to a reorder email. People use all of them after something has gone wrong. A ticket says the sync failed; a one-star review says the score felt unfair or the capsules did nothing. Neither tells you what the customer believed in the first place, which is what the product team most needs to know.
A question asked to an expert comes from curiosity rather than annoyance, and it arrives in the customer's own vocabulary. "Is a glucose spike after fruit bad?" tells you someone reads a normal rise as a warning. "Which magnesium is best for sleep?" tells you they want the choice made for them. "Why is my sleep score low after a good night?" tells you the score contradicts how they feel. Researchers have documented exactly this gap: in one detailed account of 18 nights with a sleep tracker, the author repeatedly found the app's scores and wording contradicting her own experience, and learned to distrust the app's explanations rather than her own.1 Sleep clinicians have a name for the extreme version, orthosomnia, where the pursuit of a perfect sleep score becomes its own sleep problem.2
Companies want this input badly enough to build for it: Oura created an in-app hub, Oura Labs, to test ideas with members.3 And as Andreessen Horowitz's David Haber argues, the most valuable context in an organization now sits in recorded conversation rather than in documents.4 A live expert session is that conversation, held with your own customers.
2
Build a question graph
A question graph is simple to describe and valuable to own. Take every question from every expert session. Tag each one with the topic it is really about, which is often not the one it names; with who asked it, by segment; and with the moment in the session it was asked, so you can see what prompted it. Then make the whole thing searchable across the program.
Fig. 1The graph
What a month of member questions tells the product team
Illustrative data
Read down the list and you get a product brief. The most common glucose question rests on one misunderstanding, that any rise after a meal is a problem, so the fix is an explainer on what a healthy curve looks like rather than a new chart. Supplement questions ask for a decision, so the fix is one line of guidance on form, dose and timing. The two short pink bars matter most: people asking whether a reading means they have a disease, or whether a supplement is safe with their prescription. There are only a handful of them, and each one needs to reach product and compliance the same day.
3
Analyze it like a researcher
Questions are qualitative data, and the methods for qualitative data are well established. The most widely cited is thematic analysis, set out by Virginia Braun and Victoria Clarke: read everything, label each item, group labels into themes, check the themes against the raw data, and write them up with how often each appears.5 AI can do the first pass across thousands of questions in seconds. A person who knows the product should decide the themes.
How much data do you need? Usually less than people expect. In a well-known 2006 study, Greg Guest and colleagues coded 60 interviews and found the themes were essentially complete and stable after 12.6 A month of weekly sessions, each with a few hundred questions, will surface the main themes. The value after that is in watching the themes move: which ones fade once the product explains them, and which ones grow.
Tag the topic the question is really about, which is often different from the one it names. "Why am I so tired?" is usually a sleep question, sometimes an iron one.
Keep the customer's words. Onboarding and label copy written in their own phrasing lands better than anything the product team writes.
Segment before you conclude. Women in perimenopause and endurance athletes describe sleep very differently.
Flag anything that reads as diagnosis or treatment. Send it to the product team and the compliance team the same day.
A question asked to an expert shows what a customer believes the product does.
Adelo Insights · Part 2
4
How Oura turned member testing into product
Oura offers a working example of a company treating member reactions as research. In 2024 it opened Oura Labs, an in-app space where members can try features before they are finished.3 Symptom Radar, which compares resting heart rate, skin temperature, sleep and breathing rate against a member's own baseline and suggests rest when they drift, was tested in public that year and left beta in December 2024. The beta feedback added a wellness trend graph and a breakdown of each input to the final version.7 Oura Advisor, its AI companion, followed the same route: it started with a small set of Labs users and was rolled out to every Ring Gen3 and Ring 4 member at the end of March 2025.8
Labs gathers reactions to features Oura has already built. The questions members ask an expert in a live session come earlier, before anything is built, and point at what to build next. A company that keeps both has a research program most of its competitors are paying an agency to imitate.
5
What to do with it
The graph shows
What you change
One metric or ingredient gets most of the confused questions
The explainer the content team makes next, and the first candidate for a redesign
The same question in different words across segments
Onboarding copy per segment, in their vocabulary
Questions spike at one moment in a session
The next session's agenda, and a clip for everyone who missed it
People asking if a reading is a diagnosis, or a product a treatment
Wording and claims review, before a regulator asks the same question
A theme that fades after an explainer ships
Proof the explainer worked, in a way a dashboard cannot show
Most health tech companies pay for surveys and panels to learn what their own live sessions already tell them every week, in the customers' own words. What they lack is a way to keep it.
Frequently asked questions
What is a question graph?
Every question from every expert session, tagged by the metric it is really about, who asked it and when, and searchable across the whole program. It shows which metrics, ingredients or features customers misunderstand, how they describe them, and when a wellness product is being read as a diagnosis or a treatment.
Why are member questions better research than support tickets?
Support tickets show what broke and app reviews show who is angry, but neither tells you what a customer thinks a score or an ingredient does. A question asked to an expert in a live session does, in the customer's own words, before anything has gone wrong.
How many sessions do you need before the themes are reliable?
Fewer than you would think. Research on qualitative saturation found that themes were essentially complete after about 12 in-depth interviews. A month of weekly sessions with a few hundred questions each will surface the main themes; the value after that is in tracking how they change.
How do you analyze questions from a webinar or live session?
Use thematic analysis: read everything, label each question, group labels into themes, check the themes against the raw questions, and report each theme with how often it came up and who asked. Let AI do the first pass; a person who knows the product decides the themes.
What should a health tech company do with the question graph?
Use it to write onboarding, labels and in-app explanations in customers' own words, to decide the next expert session, to catch a wellness product being read as a diagnosis or treatment before a regulator does, and to brief the product team on what to fix next.
Adelo Insights are written by the Adelo team, the professional video event platform for health. They draw on the sessions pharma, agency and consumer health teams run on Adelo every week, and on the research cited in each essay.
Adelo Insights · Health tech and the expert voice
