note
The decision.
On the intern app, we ask for a mood rating at check-in. Five faces, from "Struggling" through to "Excellent." If an intern picks one of the bottom two — "Struggling" or "Not great" — a follow-up question appears: what's going on? A short, optional cause prompt. If they pick "OK", "Good", or "Excellent" — nothing. We move on. We don't ask them to explain why.
That's a small design decision. It feels like nothing. But it's one of the decisions we get asked about most when other practitioners see the product.
Why it matters.
Many of our interns are autistic. Several have lived years with adults — well-meaning, well-trained adults — asking them to verbalise emotional states they don't easily articulate. "How are you feeling?" "Can you explain that to me?" "Use your words."
If we ask an intern to tell us why they're "OK", we're implying that being okay is a state that requires explanation. It's not. It's the default. It's what we should hope for, not interrogate.
What we ship.
Cause prompts only fire on the two lowest ratings. And even then, they're optional — the intern can skip them, and the rating still logs. The coach still sees that the day started rough, but the intern hasn't been compelled to perform an explanation they may not have words for.
For "OK", "Good", and "Excellent", the working day just continues. No question. No prompt. No friction.
The principle underneath.
Every interaction on the platform costs the intern something — attention, time, sometimes emotional load. We try to spend that budget where it matters: on the things that genuinely help. On a rough day, the offer to share what's going on is a useful thing. On a fine day, it's an unnecessary tax on someone who's already navigating a busy workplace.
This is the lens we try to apply everywhere. Default to trust. Don't make people prove their state. Build the platform around the assumption that interns are competent adults having a working day, not subjects whose internal states need to be audited.
What got us here.
The first version wasn't this. The first version asked everyone, every time. It came from a good instinct — coaches wanted rich data, parents wanted reassurance, funders wanted evidence. But within the first week of the cohort, we noticed something: interns were either skipping the prompt (so the data was thin anyway), or they were writing the kind of thing they thought the coach wanted to read. The data wasn't reliable, and the question itself was costing trust.
We changed it the next week. It's been better since.