We asked 92 women with ADHD what they rely on to run their days. Then we asked, cold, what their hand actually reached for the last time everything fell apart. Two thirds of the time, those are two different things.
The day does not fall apart on schedule. It tips over at 3pm on a Tuesday, and in that moment the setup on her phone is not what her hand goes to. This study was built to measure that distance.
Page one, first screen, no warm-up, so nothing could pull the answer toward her considered strategy: think about the last time you were completely overwhelmed. What did you reach for first? Then, separately: day to day, what do you rely on most? The distance between those two answers is the whole study. An AI moderator ran every session, chasing each participant's specific tension with one or two adaptive probes. No product, tool, category, or source was ever suggested to any participant.
Day to day, this population runs on structure. Phones, alarms, lists, planners: 57% name one of these as their main system. The overwhelmed moment looks nothing like that. In crisis, 55% reached for their body, for music, for another person, or for something to hold. Four reached for nothing at all and rode it out in the dark.
Read the probe answers and the vocabulary flips. The daily words are organized, on track, so I don't forget. The crisis words are calm, ground, reset, safe. The daily tool manages her day. The crisis reach manages her nervous system. Two jobs, and nobody we talked to had one tool doing both well.
It is not a clean swap, though. Of the 52 women whose daily system is structural, 24 stayed with structure in the overwhelmed moment too, and almost all of them stayed on the phone. Which brings us to the complication.
The phone is the most common daily driver (39 of 92), the most common crisis reflex (33 of 92), and for 21 women it is both. But it is not the same phone in both moments. Day to day it holds the alarms and the calendar. In the overwhelmed moment it becomes escape. Asked why she keeps going back to a setup she rated 2 out of 5, one woman answered in three words: "Ease and addiction." Another: "I feel almost guilty for grabbing my phone in those moments."
The phone wins on proximity, not merit. Anything built inside it inherits the scroll as a roommate, in exactly the moment the user is most vulnerable to it.
Here is the number that should reset how anyone thinks about competition in this space. The average setup rating is 3.4 out of 5. Eight women out of 92 gave their system a 5. Seventeen rated it a 1 or a 2, and most of those kept relying on it anyway. Their reasons had nothing to do with it working: simple, always there, comfortable, the only thing I know. "It is easy to use and always on my person. It is comfortable and positive habits are really hard to form."
And 11 women, roughly one in eight, said they have nothing consistent at all. "Nothing really sticks beyond a week."
When she wants something new to try, it comes from social media. Two thirds of the sample (62 of 92) named TikTok, Reddit, or YouTube as their source, then rated their trust in it 3.2 out of 5. They manage the distrust with filters: reading comments to see who tried it first, waiting until an idea shows up repeatedly, asking whether it looks attainable.
The missing character is the clinician. Half this sample is formally diagnosed. Two women out of 92 said new ideas come from a clinician or therapist. Two. The most credentialed source in their lives is nearly absent from the discovery loop, and the channel they trust least is carrying almost all of it. Discovery happens on social, conversion happens through repetition and other people's comments, and there is an open seat for anything that earns clinical-grade trust without the clinic.
At the end, 39 women got the open question: if you could have something that does not exist yet, that would help, what would it do? No examples. No steering. The largest cluster, 11 of 39, asked for the same thing: calm, on demand, in the body, right now. "Help calm my nerves instantly." "Something that could help me in real time, like snap out of my overthinking." "Probably something that like feels like a hug."
One answer is worth printing in full.
Another woman, explaining why headphones work, said the compression on her head makes her thoughts feel "kept in." The pressure itself is doing the job. Now the honest part: eleven people are not a market, and better planners came second at 8 of 39. But we asked an open question, cold, and the biggest group of answers described a calming object for the body. That did not have to happen.
The crisis job is regulation: body, senses, connection. It is a different job than the planner's, and the data says nobody has both. Lead with the overwhelmed moment and let structure come later, if ever.
The phone owns both moments and poisons one of them. A physical, on-body form factor is not a gimmick here; it is the escape hatch participants asked for unprompted. The competition is not other apps in the category. It is the scroll.
This product displaces tools people rate 2 out of 5 and keep for their ease. Nothing to configure, nothing to learn, working in the first minute. If trying the app takes more effort than opening Instagram, Instagram wins.
77 of 92 opted into follow-up research: the seed of a recontact cohort for the deeper study that follows. The cohort: the reflex-gap cases with low efficacy, the 11 with nothing consistent, and the 11 who wished for instant calm.
Discovery on TikTok, conversion in the comments. First exposure will not convert; repetition and visible other-people-tried-it will. Seed real users early and find the vouchers, because the clinician seat is empty and someone will take it.
An earlier auto-generated readout reported 99 sessions and cited figures that did not trace to the transcripts. Those 99 included 7 pilot interviews run on an older, longer script and one partial. This report uses the 92 completed wave sessions on the final protocol only. Every number was recomputed from the raw transcript export, and every quote is traceable to a session ID, verbatim apart from light spelling cleanup. The prior headline of "roughly two thirds" survives the audit. Now it has a denominator.
Sample composition (self-reported): 48 diagnosed by a clinician, 28 identify with ADHD without a formal diagnosis, 16 report traits without identifying. Trust was skipped by design for the 4 who said they do not look for new ideas. Mismatch counts "nothing, I just rode it out" as its own reflex per the study spec; excluding those 4 entirely, the rate is 64% (56 of 88). The reflex-by-driver matrix was read at the group level only; at n=92 individual cells are too thin to name specific swaps. Participant-level data remains with the client.