Context
While running user research at Discovery Education, I ran into a set of challenges that kept coming back, study after study. Some of them apply well beyond EdTech, but I can only speak to what I've seen. Working through them with some excellent peers, we landed on approaches that held up, and I want to share not just what we did, but the circumstances that made it necessary.
The focus here is direct user engagement: interviews, usability testing, and prototype feedback sessions.
Challenge 1: Locked-down school devices
Teachers, and almost always students, use school-provided computers that restrict what can be installed. That conflicts with most of the tools used to run remote research.
For years, most of our interviews were phone calls, with no way to observe, assist, or interact with the participant. Video tools often required software users couldn't install, which led to more technical problems and more compromised data than we were comfortable with. It also made it much harder to build a real relationship with participants.
Group sessions suffered most. When nobody could see each other, participation caused anxiety, something we only learned by talking to users over time. Moderating was harder too, since we had no control over the call.
What worked: browser-based video tools
We moved to tools like appear.in that didn't need an elaborate permissions model. Later, hybrid platforms like RingCentral let us consolidate back onto standard tools. One constraint never went away: screen sharing almost always requires the participant to install something. That's what pushed us toward Figma.
What worked: mailing cameras to participants
For studies where observation was essential, we bought action cameras, built custom stands, and mailed them out with prepaid return postage. It was slow and expensive, so we reserved it for research where the stakes justified it.
What worked: Figma prototypes in multiplayer mode
Figma's prototype and multiplayer modes sidestepped installation and security restrictions entirely. We could follow and record exact cursor movement and prototype interactions, and give participants far better support during the session.
Challenge 2: Distractions and unplanned participants
You sit down for an interview and suddenly there are three people instead of one. A true usability test is off the table, and the session will inevitably turn into a discussion. You have to change your moderation plan on the spot while still getting to your research goals and keeping a good relationship with the participants. This isn't unique to EdTech, but I saw it far more often with schools, where some districts and administrators are much more hands-on than others.
How much it hurts depends on the study. It's a significant setback for usability testing, or when you only have a handful of sessions planned.
What worked: have a plan, be explicit, stay flexible
There's no elegant way to tell someone they're not invited, so the answer is preparation. We'd lean on the open-ended questions we normally saved for the end of a session. Having that fallback ready is essential.
Challenge 3: Choosing the right content
Of all the user groups I've researched with, teachers have the hardest time setting aside their mental models to participate in a session.
Teachers work within a specific range of grades and subjects every day. In a test, they assume the content will match what they already teach. When a history teacher opens the prototype and finds Language Arts content, they get thrown off. Even within the same subject, being a grade or two off creates friction you have to work through.
In early project phases we rarely had access to production user data, so we couldn't build prototypes around real user profiles and scenarios. We could recruit teachers by subject, but when content isn't what you're testing, that shrinks an already small participant pool.
What worked, with caveats: custom-content prototypes
We tried building a version of the prototype for each teacher, with content broad enough to feel familiar. It did improve sessions, but the extra effort only paid off at high participant volumes.
What worked: framing the scenario before the session
This is standard practice, but it matters more here because each teacher's curriculum is so specific, and different subjects are taught in different ways. Put simply, in our use cases, content and layout can't be evaluated independently.
Challenge 4: Limited availability
Asking for anyone's time during the day is a big ask, and teachers are responsible for their students' days as well as their own. But being overbooked doesn't stop them. Teachers sign up at a higher rate per invite than any user group I've worked with, and they're happy to meet early in the morning or late in the evening. I've seen teachers book slots they knew they couldn't make because they were excited to take part. That pushed us to rethink scheduling for remote research altogether.
What worked: self-serve scheduling
With Calendly, we opened time slots across entire days to fit teachers' schedules and our own. Automating scheduling saved time and stress, which made us more willing to run studies in the first place.