Forward or snap any document — policies, warranties, IDs, lab reports, school forms — and Drawer names it, files it, remembers every date in it, and answers questions about it. The junk drawer, upgraded with a brain.
Insurance arrives on WhatsApp. Warranties hide in email. Property papers sit in a physical drawer. Lab reports live in a hospital app you deleted. Nothing is findable at the moment it's needed.
A lapsed car policy discovered after the accident. A passport with 5 months' validity found the week before a trip. A dead warranty on a repair the company would have covered. Each miss costs hundreds to thousands — and they're all just date failures.
In most households, exactly one person knows where everything is. That's a single point of failure families discover at the worst possible time — an emergency, a hospitalization, a death.
Cloud drives hold files and understand nothing. Scanner apps capture and forget. The job to be done isn't "store my documents" — it's "never let a document surprise me."
Paperless-ngx, the self-hosted document archive, is one of the most popular OSS projects in home computing. Ordinary people run Docker, Redis, and OCR workers at home to organize household paperwork. Demand demonstrated through suffering is the strongest consumer signal there is.
Hosted "Paperless for families" services are appearing; App Store early movers converge on AI scanner-organizers with auto-tagging and document Q&A — but none has distribution, a reminder engine, or chat-native ingestion.
Consumer document-conversion intents ("pdf to excel"-class searches) show huge, stable, year-round volume — including in India — proving ordinary people wrangle documents constantly, not occasionally.
Before building: Google Trends 5-year curves on "document organizer app" and reminder-wedge terms, app-store keyword volumes (discovery happens in the store), and 25 family interviews. If organizer searches are flat rather than growing, the GTM shifts from harvesting search to creating the moment of need — same product, different playbook.
Encrypted at rest under a per-household key hierarchy rooted in hardware security modules; documents decrypted only momentarily to be read; never used for training; no ads, ever — with a documented path to client-held keys. For a product holding IDs, deeds, and medical records, privacy isn't a feature list — it's the entire brand promise, and the moat Big Tech's ad-funded models can't credibly copy.
A document is processed once, then cached forever; reminders are cron jobs. Marginal cost per family trends toward zero after month one — consumer-subscription economics without consumer-AI burn.
The upside case isn't a bigger percentage — it's category creation: if "family document brain" becomes a default app the way password managers did, the ceiling is every household with a phone. Password managers were also "just a utility" until 1Password passed $250M ARR.
| Captures | Understands | Reminds | Family-shared | |
|---|---|---|---|---|
| Drawer | Camera + WhatsApp + email | Yes | Yes | Yes |
| Google Drive / iCloud | Manual upload | No | No | Partial |
| Scanner apps (CamScanner-class) | Camera only | Barely | No | No |
| Paperless-ngx (self-host) | Watch folders | Yes, with setup | DIY | DIY |
| App Store early movers | Camera | Partial | No | No |
| DigiLocker (India, gov) | Gov-issued docs only | No | No | No |
DigiLocker is a complement, not a competitor — it proves Indians will keep documents in an app, while covering only government-issued documents with zero intelligence. Drawer is where everything else lives, and what none of these do is open the app for you: a reminder engine is the difference between a filing cabinet and a family member.
Full capture, search, and Q&A. Enough to feel the magic and trust the product with more.
Unlimited documents, all reminders, 5 members, shared vaults, export anytime.
Visa pack, tax-season pack, home-loan pack — checklist, quality checks, one ordered PDF.
Organic-first CAC (content + invites) targeted under $3. Blended ARPU ~$30/yr, consumer churn assumed 4%/mo at launch falling as the vault deepens — switching cost grows with every document filed. Target LTV:CAC > 5 by month 12; Moments packs lift ARPU in exactly the months users are most engaged.
The reminder feed is intent at its purest: "insurance lapses in 7 days" is the highest-intent renewal lead in existence. Referral partnerships on renewals (insurance, AMC contracts, passport services) monetize without ads and without selling data — disclosed, opt-in, aligned with the user's own deadline.
"Document organizer," "document scanner," and "warranty tracker" app-store searches are steady, high-intent, and dominated by dated utilities with weak ratings. Ranking there converts existing demand while the content engine creates new demand — the two-channel mix a consumer utility needs.
Validation gate. Trends + ASO keyword data + 25 family interviews. Kill criteria defined up front: if neither search demand nor interview urgency materializes, stop here at near-zero cost.
GATE · 15/25 interviews rate the reminder feed "would use weekly"
MVP: capture → understand → remind on mobile, WhatsApp ingestion, waitlist launch with the 3-second-warranty demo video.
GATE · 1,000 waitlist → 300 beta families · D30 retention ≥ 25%
Habit + revenue: family vaults, regional languages, first Moments pack (visa), paid tier on.
GATE · 10K MAU · 500 paying families · invite rate ≥ 1.5 members/household
Scale loops: email auto-ingest, renewal referral partnerships, Germany launch, ASO push.
GATE · 100K installs · 5K paying · reminder-driven weekly open rate ≥ 40%
The risk. Mitigation: reminders (protection from loss) as the paid core rather than storage; Moments packs monetize high-stakes episodes where paying is painless; India price at UPI-autopay impulse level. The week-4 interview gate tests this before a line of code.
The unfair channel depends on Meta's Business API terms and per-conversation pricing at consumer scale. Mitigation: verify in week 1 with a working prototype and cost model; camera + email are fully independent fallbacks.
A wrong expiry date is worse than none. Mitigation: confidence thresholds, user-confirmed extractions (one tap), and never firing a reminder on an unconfirmed date.
Google or Apple could bundle this. They've had a decade of scanner apps and haven't — because it needs obsessive focus and a trust posture their ad models undercut. Mitigation: move fast on the family graph; switching costs compound with every filed document.
One leak ends a product built on trust. Mitigation: envelope encryption by architecture not policy, minimal server-side plaintext, external audit before scale, and a security posture treated as the #1 engineering budget line.
If the Trends gate shows organizer searches large-but-flat, the harvest channel shrinks. Mitigation already in the plan: the content engine creates the moment of need instead of waiting for it.
Every household runs on an invisible calendar of obligations — renewals, filings, expiries, payments — currently stored in one person's head. Drawer digitizes that calendar as a side effect of filing paper. Own the household's date graph and you become the app a family checks the way they check the weather: the operating system for life admin. Password managers proved a "boring" trust utility can quietly become default infrastructure in a billion pockets.
Two quarters, one small team: weeks 0–4 validation gate (Trends, ASO, 25 interviews) → MVP → 300-family beta hitting D30 ≥ 25%. Bootstrappable to the beta gate; the seed conversation starts the day the retention number is real. What's needed now: design-partner families, the WhatsApp API cost verification, and the discipline to honor the kill criteria.