Positioning, how we make money, and how we get there — assuming validation passes.
For the autoimmune patient who lives at the mercy of unpredictable flares, Flarewise is the AI companion that learns your personal triggers and warns you before a flare hits — so you can act early instead of just reacting.
They commiserate about the past. Flarewise forecasts your future.
They only record (Bearable, Flaredown). Flarewise predicts — the thing none of them do.
Seen twice a year. Flarewise is the daily layer between visits — and produces the report you bring in.
"Why one disease?" — you're right to push. The core prediction engine is cross-condition: the same wearable signals (HRV, resting HR, skin temperature) precede flares across IBD, RA and lupus, and the model is n-of-1 anyway — it learns each person, not a disease. So the marginal cost of a second condition at launch is mostly content + a trigger list + a community, not a new engine. That changes the answer:
Support the top autoimmune conditions out of the gate — IBD + RA first, lupus & Hashimoto's close behind. Wider top-of-funnel, day-0 value for more people, and a hedge if one condition's signal proves harder to reproduce.
Lead with IBD — deepest evidence (IBD Forecast, Gastroenterology 2025), youngest/most content-native community (~3M US, intensely online), and the "which foods trigger me?" hook that drives logging adherence. RA is the fast-follow (RA Forecast, Nature Sci Rep 2025 — cleanest passive-wearable evidence).
This resolves the tension: you get the breadth that fixes acquisition and day-0 value, without the dilution of generic "autoimmune" content — because your content engine produces condition-specific content cheaply across all of them (that's the edge). Umbrella path: IBD + RA → lupus → thyroid → inflammatory skin.
Persona — Sofia, 29, Crohn's. A flare = her usual 2 normal stools/day become 6–8 urgent, sometimes bloody; cramping, crushing fatigue, sometimes fever — days lost or an ER trip. The product's job: warn her before it hits, with almost no effort. The wearable does the heavy lifting passively; she actively logs only ~5 taps + a photo.
# bowel movements · urgency · blood (y/n) · Bristol scale (tap a picture 1–7) · pain 0–10 · energy 0–10 · stress slider · "what'd you eat?" → snap a photo, AI tags the foods (or pick frequent-foods) · meds taken.
sleep · resting HR · HRV · skin temperature (Oura / Apple Watch / Fitbit) · menstrual cycle · weather. Zero effort — the signals the research shows precede flares by weeks.
The payoff she pays for: "⚠️ Flare risk elevated for the next 3–5 days — your resting HR is up, HRV is down, you slept poorly 2 nights, and you had dairy + alcohol yesterday. Ease off, prioritize sleep — here's the note for your GI." And over time: "Your top-2 personal triggers are dairy and low sleep." — the trigger fingerprint no forum or generic tracker gives her.
The flaw you spotted: if net value only arrives after weeks of journaling, most users churn in week 1 and never reach the payoff — and the prediction model never gets the data. The fix is to make Flarewise a complete autoimmune companion that's useful on day 0, with the personal forecast as the flagship that deepens over the first month. Value is a ladder, not a cliff:
The moment she onboards (condition, meds, known triggers), she gets an evidence-grounded AI companion — "is this safe with my meds?", "what do I eat during a flare?" — plus a personalized playbook and the population-likely triggers to watch. Real utility from minute one, before a single log.
Logging itself pays off immediately: a clean, shareable doctor-visit summary she can bring to her next appointment after even a week — solving the "how have you been?" blank most patients dread. The habit that feeds the model is valuable on its own.
The passive wearable is the cold-start shortcut: a physiological baseline forms in ~7 days (the research window), so early-warning can start within the first week — long before the full personal model matures. Prediction isn't a month-away promise; a useful version arrives fast.
As it learns her unique trigger fingerprint, the forecast sharpens and the "your top-2 triggers are dairy + low sleep" insight lands. This is the paid, compounding, un-buyable feature — but it sits on top of a product that already earned its place.
Tracking is free (the hook + the data collector); prediction is paid. ~$9–15/mo (or $79–99/yr) for AI flare forecasts, personalized trigger insights, early-warning alerts, and the doctor-visit report. Retention is structural: the prediction gets better the longer you log → leaving means losing your model.
Monetize the moments of need — route to telehealth (rheumatology/GI/functional med), relevant products, and diagnostics. Affiliate/referral revenue on genuinely helpful hand-offs. No clinical license — you route, you don't treat.
The aggregated, anonymized, structured trigger→symptom→flare→treatment-response dataset is gold to the booming immunology pipeline: RWE data licensing to pharma/CROs (6–7-figure deals) and trial recruitment — your engaged patient panel is a channel autoimmune trials pay a lot to reach (hundreds to thousands of $ per enrolled patient). This is the second, larger revenue engine — and it only exists because Layer 1 built the moat.
Your concern, answered: the growth engine is not manually posting in Facebook groups (unscalable) or waiting on organic search indexing (too slow). It's the thing you're already best at: a faceless short-form video machine. Chronic-illness / autoimmune is one of the most emotionally-engaged, share-heavy audiences on Reels, TikTok and Shorts — "signs your flare is coming", "foods I quit for my Crohn's", "what my watch caught before my flare". That content travels fast and free, it's your exact faceless-studio playbook, and it fans out across conditions at near-zero marginal cost. Three layers, fast → durable:
The primary engine. Daily condition-specific Reels/TikTok/Shorts from the studio machine. Viral-capable, scalable, free, and multi-condition. This is the growth flywheel — not a side channel.
A shareable "What's your flare risk?" quiz as top-of-funnel (instant value, built to share), amplified by a small paid-social test to condition audiences — the fastest, most measurable way to buy the Gate-1 demand read instead of waiting for it. try the live quiz →
Programmatic condition/symptom/trigger articles that compound in Google and AI answers. Slow to start, but the near-zero-CAC base under everything — the long-term moat, not the launch plan.
Video engine (IBD-led, RA next) → companion + free tracker with day-0 value → build the logging habit + the dataset. (Runs through/with validation.)
Ship the AI prediction, turn on the paid tier + care-routing. Recurring revenue; deepen IBD + RA before the next condition.
Productize the dataset — data licensing + trial-recruitment deals. The unicorn layer.