GhostFrame Proof
A phone-based capture app that makes move-out cleaners shoot pixel-aligned before/after photo pairs and auto-generates a tamper-sealed, client-facing proof report for a single job, so deposit disputes end with the cleaner getting paid.
One product, one founder, thirty days. Everything below is built to be executed inside the contest window — the official rules still decide what counts.
Decide in 20 seconds
The facts that decide whether this is worth thirty days of yours.
Owner-operators of 1-8 person move-out and deep-clean companies in the US, UK and Australia who invoice landlords and property managers $250-$900 per job and get paid after a walkthrough; the owner answers their own phone and buys software the same day.
Operational simplicity 5.0/10
Why this could work in 30 days: one founder can ship the MVP in 5 days; the plan targets a first sale on day 12; 21 sales at $49 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 8.5/10.
Score and projections come from the generated report and are not guarantees. The $100K Hackathon is run by Ship or Die — read the official rules before you build.
Your first 24 hours
10 concrete actions. Tick them off as you go — nothing is saved or sent.
Product opportunity and positioning
Target customer
Owner-operators of 1-8 person move-out and deep-clean companies in the US, UK and Australia who invoice landlords and property managers $250-$900 per job and get paid after a walkthrough; the owner answers their own phone and buys software the same day.
Why now
Every cleaner already shoots photos on a phone, and mobile browsers now handle live camera compositing, Web Crypto hashing and installable PWAs, so the proof layer needs no native app, no hardware and no per-job human review.
Problem
When a landlord withholds a deposit or claims a room wasn't cleaned, the cleaner has 40 unordered camera-roll photos, no per-room timestamps and no way to prove the after-shot came from the same position as the before-shot, so they eat a $150-$400 re-clean or lose the property-management account.
Solution
The cleaner picks a room template, shoots a before photo, then the app overlays that shot at 40% opacity on the live camera so the after shot lines up; the app generates a branded PDF with paired photos per room, capture timestamps, GPS, a SHA-256 seal per image and a manifest hash, delivered as an expiring link with a one-tap client acknowledge.
1. Product opportunity and positioning
Move-out and deep-clean jobs are invoiced at $250-$900 and paid after a walkthrough, so the cleaner's cash depends on someone else's memory of a room. When a landlord withholds a deposit or claims a room "wasn't cleaned", the cleaner's evidence is a camera roll: forty unordered photos, no per-room timestamps, no way to show the after shot came from the same position as the before shot. Losing that argument costs a $150-$400 re-clean or a property-management account. GhostFrame Proof sells the artifact that ends the argument, not a cleaning tool.
The buyer is deliberately narrow: the owner-operator of a 1-8 person move-out cleaning company in the US, UK or Australia. That person answers their own phone, buys software the same day and runs no procurement. JEV scored this concept 66.4/100 overall but 8.47/10 on speed to first revenue and 7.41/10 on reachable customers, which says the binding constraint is founder outreach throughput, not demand discovery.
Positioning line: "Landlord kept the deposit? Send them this." The comparison is never another SaaS product; it is a WhatsApp thread of JPEGs. GhostFrame wins that comparison on one visible property. The after shot is framed against a ghost of the before shot, and every PDF page carries a SHA-256 image seal, capture timestamp and GPS with a manifest hash in the footer. A landlord cannot claim the photos are a different room, a different angle or a different day without also claiming the hash was forged.
Why now: cleaners already shoot photos on phones, and the browser can now composite a live camera feed, hash on upload and ship an installable PWA. That removes the native app, the hardware and the per-job human review that would otherwise make a $49 subscription uneconomic.
The honest weakness is credibility. A self-generated, self-hashed report is not an independent escrow of truth, and a landlord or deposit scheme may still reject it. The 30-day plan treats that as a hypothesis to test with real landlords in week three, not as a solved fact.
Smallest sellable MVP
A mobile web flow that captures one ghost-aligned before/after pair per room for a single job and delivers a branded, hash-sealed PDF behind an expiring link with a client acknowledge button.
Must-have features
- 12-item editable room shot list (kitchen surfaces, oven interior, inside cupboards, bathroom, floors, skirting, windows, bins)
- Ghost overlay: before shot rendered at ~40% opacity over the live camera with a lock and retake control
- Server-side PDF generator with two-up before/after pages, room labels, timestamps, GPS, logo and phone number
- SHA-256 hash per original image plus a manifest hash printed in the report footer
- Expiring share link with a client acknowledge button recording name, timestamp and report ID
- Stripe Checkout, magic-link accounts and a $49/month gate on report generation
- 30-day original retention with a one-click dispute ZIP (originals plus manifest JSON)
Explicitly out of scope
- No AI cleanliness or damage scoring
- No scheduling, invoicing, quoting, CRM or staff GPS tracking
- No native iOS or Android apps; PWA only
- No long-term team photo library or digital asset management
- No cleaner marketplace or directory
- No property-management or field-service integrations
Demo flow (what the buyer sees)
- Founder opens the app and creates a job against a move-out template
- Takes a before shot of the oven interior on a phone
- Walks to a new position and shoots the after shot against the 40% ghost overlay
- Uploads, server hashes both originals and writes the manifest row
- PDF renders with paired pages, timestamps, GPS and the footer seal
- Cleaner opens the expiring link, sends it to the landlord, landlord taps acknowledge
Technical implementation plan
Stack
Architecture
A Next.js PWA on Vercel serves the capture page and the report viewer. Capture happens entirely in the browser: the video stream is drawn to a canvas with the previous frame composited at 0.4 alpha, then re-drawn at native resolution on capture, downscaled to 1600px on the long edge and uploaded to Supabase Storage. A route handler hashes each original, stores the digest and writes a manifest row. PDF generation runs on demand in a serverless function, reads the manifest, lays out two-up pages with pdf-lib and returns a signed URL. Subscription state lives in Supabase and is set by a Stripe webhook; a scheduled function purges originals at day 30.
Build steps
- Scaffold Next.js, deploy to Vercel, attach the domain
- Create Supabase tables: jobs, images, manifests, subscriptions
- Build the room template editor seeded with twelve move-out items
- Implement camera capture with geolocation and client-side downscale
- Composite the before shot at 0.4 alpha over the live video with lock and retake
- Hash originals server-side on upload and write the manifest row
- Generate the PDF with pdf-lib: two-up pages, labels, timestamps, logo, footer hash
- Issue expiring signed share links with the acknowledge action
- Add magic-link auth and gate report generation behind an active Stripe subscription
- Add the dispute ZIP endpoint and the day-30 retention purge
AI components
- None in the MVP: the overlay is an alpha-composited canvas and hashing is deterministic SHA-256.
- Post-launch candidate only: a vision check that flags rooms with a before shot but no matching after shot.
Pricing and monetization
Solo
Entry offerFor Owner-operator running 1-5 cleaners and 20-60 jobs a month
- Unlimited sealed before/after reports
- Ghost-overlay capture flow
- Custom logo and phone branding
- Expiring client sign-off link
- 30-day original retention and dispute ZIP export
Founding Crew
For Cleaners who want to lock in before a price change and get their template built for them
- Six months of Solo
- Done-with-you room template matched to their typical move-out
- Dispute ZIP export
Model: Flat $49/month Solo subscription with unlimited jobs and seats, plus a $199 one-time Founding Crew prepay for cleaners who want their room template built for them.
First revenue offer: Send me two before photos from your next move-out job and I'll build you a sealed before/after report free in ten minutes. Forward it to the landlord and see if they push back; if it holds, it's $49/month.
Expected ACV: $560
Upsell path: Founding Crew at $199 converts a hesitant $49 trial into six months prepaid, and each buyer's referral link is worth one free month so one customer produces the next.
First-customer acquisition
Cold email and Instagram DM to move-out cleaning companies found on Google Maps in ten metros.
Where to find them
- Google Maps searches such as 'move out cleaning Austin' across ten US, UK and Australian metros
- Yelp business listings for deep-clean and end-of-tenancy cleaners
- Instagram accounts posting before/after cleaning reels
- Facebook groups for cleaning business owners and end-of-tenancy cleaners
- Thumbtack and Angi cleaner profiles with visible contact details
Exact outreach motions
- Build a 300-row list with owner name, email, phone and Instagram handle, flagging the 60 with recent posts
- Send 60 cold emails with subject 'your move-out photos, sealed' and a sample report link
- DM 25 Instagram accounts with a 20-second screen recording of overlay capture turning into a sealed PDF
- Ask three warm prospects for two real before photos and rebuild their report free
- Follow up on day 2, 5 and 9 with one specific high-dispute room type: oven interior, inside cupboards, carpet edges
- After a report lands, ask for the property manager's email so the report also sells the cleaner's PM contract
Message template
Subject: your move-out photos, sealed
Hi {name} - I saw {company} does move-out cleans in {metro}. When a landlord says a room wasn't cleaned, a camera roll doesn't win that argument. Send me two before photos from your next job and I'll build you a sealed before/after report free in ten minutes: paired shots, per-room timestamps, a hash seal on every page. Forward it to the landlord and see if they push back. Worth a try?
- {founder}, GhostFrame Proof30-day execution roadmap
Week 1 · Build & validate · 7 days
Scaffold the app and start the prospect list
- Init the Next.js repo, deploy to Vercel, attach the domain.
- Create Supabase tables for jobs, images, manifests and subscriptions.
- Create the Stripe $49 monthly price in test mode.
- Open the prospect sheet and log the first 40 cleaners.
Deliverable: Deployed skeleton, live domain, 40-row prospect sheet.
Room template and camera flow
- Build the twelve-item room checklist as an editable job template.
- Add getUserMedia capture with geolocation.
- Downscale shots to 1600px client-side before upload.
Deliverable: Camera capture page storing shots per room.
Positional alignment
- Render the before shot at 40% opacity over the live video element.
- Add lock and retake controls with a centre alignment guide.
- Test on two Android phones and one iPhone.
Deliverable: Working ghost overlay on three devices.
Sealed PDF generation
- Lay out two-up before/after pages per room with pdf-lib.
- Embed logo, phone number, room label, timestamp and GPS.
- Render a six-room sample from test photos.
Deliverable: Sample six-room PDF report.
Hashing and share links
- Hash each original on upload and store the digest.
- Compute and print the manifest hash in the footer.
- Issue expiring signed links with a client acknowledge button.
Deliverable: Sealed report link recording client sign-off.
Payments, auth, production
- Add magic-link auth and gate report generation behind an active subscription.
- Wire Stripe Checkout and the Billing Portal.
- Deploy to production and run one live test purchase.
Deliverable: GhostFrame Proof live and sellable with a test transaction.
Build the outreach list
- Run ten metro searches and capture owner name, email and Instagram.
- Deduplicate and flag the 60 best-fit companies.
- Write the three-email sequence.
Deliverable: 300-row prospect list with sequence drafts.
Week 2 · Launch & first sale · 7 days
Demo video and landing page
- Record a 20-second screen capture of overlay to sealed PDF.
- Publish a one-page site with the video, the price and a sample report.
- Build three sample reports from stock before/after photos.
Deliverable: Landing page, demo video, three sample reports.
First 85 touches
- Send 60 cold emails with the sample report link.
- Send 25 Instagram DMs to flagged accounts.
- Log every reply and objection in the sheet.
Deliverable: 85 touches sent, replies logged.
Do the work manually
- Ask three warm prospects for two photos from their next job.
- Build each free sealed report by hand.
- Record a 90-second walkthrough per report.
Deliverable: Three free reports delivered with walkthroughs.
Convert warm replies
- Send day-2 follow-ups with a specific oven-interior sample.
- Offer Founding Crew at $199 to anyone who replied.
- Book call slots for tomorrow.
Deliverable: Three calls booked.
First paying customer
- Run each call as a live rebuild of their own report.
- Send the Stripe link during the call, not after.
- Ask each buyer for one referral name.
Deliverable: First paying customer and one referral name.
Batch two
- Send 60 more emails with a new subject line.
- Post the demo clip in two cleaner Facebook groups.
- Reply to every open thread within the hour.
Deliverable: 60 new emails, two group posts, live threads handled.
Fix what breaks
- Fix the top support issue from week one.
- Publish a deposit-dispute template page.
- Cut every message angle that produced no replies.
Deliverable: Patched product, dispute page, one winning angle.
Week 3 · Repeat & systematize · 7 days
Batch three at volume
- Send 80 emails to the remaining list.
- Send 30 DMs to accounts that posted this week.
- Add a fourth touch line naming a metro.
Deliverable: 110 touches sent.
Turn buyers into a channel
- Ship referral links worth one free month.
- Ask every buyer to drop the report link in two groups.
- Email each buyer's property manager with the cleaner cc'd.
Deliverable: Referral links live, property-manager emails sent.
Property-manager demand pull
- Email 40 property managers asking what proof they want from cleaners.
- Offer to send a sample sealed report.
- Capture the exact wording they use for a 'room not cleaned' claim.
Deliverable: 40 property-manager conversations, objection wording captured.
Dispute ZIP export
- Package originals plus manifest JSON into one download.
- Add the export behind auth on the share page.
- Test the ZIP opens on Windows and macOS.
Deliverable: One-click dispute ZIP export.
Short-form proof
- Record three 30-second clips: oven interior, inside cupboards, carpet edges.
- Post one per day on X and TikTok.
- Embed each clip on the landing page as a room-specific example.
Deliverable: Three clips published and embedded.
Batch four
- Send 80 emails to UK and Australian cleaners.
- Follow up day-5 on batch one.
- Log conversion rate by metro.
Deliverable: 80 international emails, metro conversion table.
Halfway audit
- Compute paying customers, churn and refunds to date.
- Kill the worst metro and the worst message angle.
- Rewrite the offer line using the strongest objection heard.
Deliverable: Halfway audit with two decisions made.
Week 4 · Scale what works · 7 days
Batch five
- Send 80 emails led by the strongest objection.
- DM 30 more Instagram accounts.
- Call three non-responders who list a phone number.
Deliverable: 110 touches plus three phone calls.
Make month two automatic
- Email each customer their month-one job count and a report link.
- Ask two customers for a written line usable on the site.
- Confirm Stripe renewals run without intervention.
Deliverable: Usage recap sent, two quotes collected.
iOS Safari reliability
- Harden camera permission and orientation handling on iOS.
- Add an upload fallback when getUserMedia fails.
- Test on two borrowed iPhones.
Deliverable: iOS capture path with upload fallback.
Batch six
- Send 80 emails to franchise owner-operators.
- Post a before/after carousel in three groups.
- Reply to every unanswered thread.
Deliverable: 80 emails, three group posts, inbox cleared.
Founding Crew push
- Offer $199 Founding Crew to every warm non-converter with a 48-hour deadline.
- Build their room template the same day.
- Send the Stripe link right after the template lands.
Deliverable: Founding Crew offers out, templates built.
Proof without fake logos
- Write one anonymised walkthrough titled 'a $340 deposit dispute, closed'.
- Publish it as a page linked from the landing page.
- Keep it factual: describe the workflow, name no customer.
Deliverable: One anonymised walkthrough page.
Batch seven
- Send 80 final-batch emails.
- Re-contact every prospect who opened but never replied.
- Ask buyers to name one competitor worth contacting.
Deliverable: 80 emails, re-contact list cleared.
Days 29-30 · Consolidate · 2 days
Convert everything outstanding
- Call every warm lead who has not paid.
- Send Stripe links with a one-line reminder.
- Fix any checkout failure blocking a payment.
Deliverable: Every pending payment collected or closed.
Final numbers and submission
- Reconcile Stripe payouts against TrustMRR records.
- Document refunds, fees and ad spend for the profit line.
- Submit the entry before 13:00 UTC.
Deliverable: Submitted entry with reconciled numbers.
Revenue and profit scenarios
How the numbers are defined
Hackathon profit — Collected revenue − refunds − chargebacks − payment processing fees − paid marketing spend.
Business profit — Hackathon profit − operating costs (hosting, domains, tooling, AI/API delivery costs).
The contest ranks entrants on hackathon profit only. Hosting, tooling and AI/API delivery costs are not deducted by the organisers, but each sale still has to cover them — so we show both. Definitions verified on 2026-11-15 against the official rules.
Stated assumptions
- Conservative: 3 paying cleaners by day 30 at a flat $49; roughly 1% conversion on about 300 touches.
- Base: 15 paying cleaners at a flat $49; roughly 4% conversion on about 400 touches.
- Upside: 30 paying cleaners at a blended $74 first payment (25 Solo at $49 plus 5 Founding Crew at $199 prepaid).
- Payment fees assume Stripe at about 4.5% blended: 2.9% + $0.30 domestic, higher on UK and Australian cards.
- ad_spend_usd is 0 in every scenario because the channel is cold email and DM, not paid marketing.
- Operating costs cover Vercel, Supabase storage and compute, Resend and the domain, about $1.20-$3.80 per customer per month at these volumes.
- Refunds and chargebacks assume none in conservative, one refunded Solo month in base, one refunded Founding Crew plus one chargeback in upside.
- All three scenarios are projections built from stated assumptions, not verified revenue.
Conservative
3 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $7
- Paid marketing
- $0
- Hackathon profit
- $140
- Operating costs
- $18
- Business profit
- $122
Base case
15 customers- Refunds
- $49
- Chargebacks
- $0
- Payment fees
- $33
- Paid marketing
- $0
- Hackathon profit
- $653
- Operating costs
- $60
- Business profit
- $593
Optimistic
30 customers- Refunds
- $199
- Chargebacks
- $49
- Payment fees
- $100
- Paid marketing
- $0
- Hackathon profit
- $1,872
- Operating costs
- $115
- Business profit
- $1,757
Model your own numbers
Runs in your browser · nothing is saved- Collected revenue
- $735
- Payment fees (3.9%)
- $29
- Paid marketing
- $0
- Hackathon profit
- $706
- Business profit
- $646
Hypothetical arithmetic, not a forecast. The fee rate is a flat 3.9% assumption, refunds and chargebacks are left at zero here, and none of this is tracked data.
Scenario models from stated assumptions, not verified revenue.
Main risks and validation experiments
Risk 1Cold email and DM conversion lands under 3%, making even the base case unreachable inside 30 days.
Track reply rate per metro and per message angle from day 9 and cut the losing variants by day 14; add phone calls at day 22.
Send 60 emails and 25 DMs on day 9 with two subject lines and measure replies within 72 hours.
Fewer than 3 replies from 85 touches by day 12 means the message, not the volume, is wrong; rewrite the offer before sending anything else.
Risk 2Landlords and deposit schemes reject a self-generated, self-hashed report as evidence, breaking the core proof claim.
Position it as a dispute-conversation artifact rather than a legal certificate, and add the dispute ZIP so a scheme receives originals plus the manifest.
Email 40 property managers on day 17 and ask what evidence they accept for a 'room not cleaned' claim.
If fewer than 5 of 40 property managers say paired dated photos would change their decision, drop the landlord-facing framing and sell it as internal job QA.
Risk 3Ghost overlay drifts in dark, cramped or reflective rooms, producing pairs worse than plain photos.
Add an explicit retake control and a per-room note field so the cleaner can flag an unusable pair before generating the report.
Shoot five rooms in a real vacant property on day 3 and have one cleaner score alignment pass or fail per room.
If more than one room in five fails alignment with a mid-range Android phone, ship a fixed-tripod guide mode instead of freehand overlay.
Risk 4iOS Safari camera and geolocation restrictions break capture for a meaningful share of buyers.
Add an upload fallback path on day 24 so a failed camera still produces a sealed report from the photo library.
Test capture on two borrowed iPhones on day 3 and again after hardening on day 24.
If the fallback is needed on more than 30% of captures, prioritise a minimal native wrapper over further PWA polish.
Risk 5Field-service photo tools already own contractor photo documentation and are the default answer.
Do not compete on job management; compete on the sealed per-room pair and the expiring client sign-off link those tools do not produce.
Ask the first five prospects what they use today and read back their exact words before pitching.
If prospects say their current tool already produces landlord-accepted sealed reports, stop and re-cut the wedge.
Hour-by-hour: the first 24 hours
Hours 1-2
- Create the Next.js repo and deploy the empty app to Vercel.
- Attach the domain and confirm HTTPS.
Output: Live URL serving a placeholder page.
Hours 3-5
- Create the Supabase project and write the jobs, images and manifests tables.
- Seed the twelve-item move-out room template.
Output: Working database schema with a seeded template.
Hours 6-8
- Build the camera capture page with getUserMedia and geolocation.
- Downscale captured frames to 1600px client-side.
Output: Capture page that saves shots locally against a job.
Hours 9-11
- Composite the previous shot at 0.4 alpha over the live video.
- Test alignment on the founder's own phone in a real kitchen and bathroom.
Output: Ghost overlay capture working end to end on one device.
Hours 12-14
- Open the prospect sheet and run three Google Maps searches.
- Log 40 cleaners with owner name, email and Instagram handle.
Output: 40-row prospect list ready for day 9 outreach.
Full 30-day launch report
Read the full generated report
GhostFrame Proof: 30-Day Solo Launch Report
1. Product opportunity and positioning
Move-out and deep-clean jobs are invoiced at $250-$900 and paid after a walkthrough, so the cleaner's cash depends on someone else's memory of a room. When a landlord withholds a deposit or claims a room "wasn't cleaned", the cleaner's evidence is a camera roll: forty unordered photos, no per-room timestamps, no way to show the after shot came from the same position as the before shot. Losing that argument costs a $150-$400 re-clean or a property-management account. GhostFrame Proof sells the artifact that ends the argument, not a cleaning tool.
The buyer is deliberately narrow: the owner-operator of a 1-8 person move-out cleaning company in the US, UK or Australia. That person answers their own phone, buys software the same day and runs no procurement. JEV scored this concept 66.4/100 overall but 8.47/10 on speed to first revenue and 7.41/10 on reachable customers, which says the binding constraint is founder outreach throughput, not demand discovery.
Positioning line: "Landlord kept the deposit? Send them this." The comparison is never another SaaS product; it is a WhatsApp thread of JPEGs. GhostFrame wins on one visible property: the after shot is framed against a translucent ghost of the before shot, and every PDF page carries a SHA-256 seal, capture timestamp and GPS with a manifest hash in the footer. A landlord cannot credibly claim a different room, angle or day without also claiming the hash was forged.
Why now is a browser story. Cleaners already shoot photos on phones, and the browser can now composite a live camera feed, hash on upload and ship an installable PWA. That removes the native app, the hardware and the per-job human review that would kill a $49 subscription.
The honest weakness is credibility. A self-generated, self-hashed report is not an independent escrow of truth, and a deposit scheme may reject it. The plan treats that as a week-three experiment with real property managers, not a solved fact.
2. Smallest sellable MVP
The sellable unit is one job. Capture a before/after pair for up to twelve room items with the ghost overlay, then emit one branded, hash-sealed PDF behind an expiring link with a client acknowledge button. Everything else waits: no AI cleanliness scoring, no scheduling, no invoicing, no native app, no team library, no marketplace.
The closing demo fits in ten minutes. The founder takes a prospect's own two before photos, rebuilds them as a sealed report and emails the link. If that link does not make the cleaner say "send that to my landlord", the product is wrong, not the pitch.
3. Technical implementation plan
Next.js on Vercel serves the PWA and the report viewer. Capture stays in the browser: the video stream is drawn to a canvas with the previous frame at 0.4 alpha, re-drawn at native resolution on capture, downscaled to 1600px and uploaded to Supabase Storage. A route handler hashes each original with SHA-256, stores the digest and writes the manifest row. PDF generation runs on demand with pdf-lib and returns a signed URL. Supabase holds subscription state, set by a Stripe webhook. Resend delivers reports and sign-off confirmations. A scheduled function purges originals at day 30.
The hardest technical risk is overlay stability on iOS Safari: if the preview crops differently from the captured frame, the pair visibly misaligns and the core claim collapses. Build the retake control on day 3, not day 20.
4. Pricing and monetization
$49/month Solo, unlimited jobs and seats, branded reports, expiring sign-off links, 30-day retention and dispute ZIP export. $199 Founding Crew buys six prepaid months plus a done-with-you room template; it exists to convert hesitation into cash inside the contest window. There is no free tier, because free trains cleaners to treat this as a utility and kills the urgency a 30-day launch depends on. The free offer is one manually built report, which is both the demo and the sales motion.
5. First-customer acquisition
One channel: cold email and Instagram DM to cleaners listed on Google Maps in ten metros. Build 300 rows with owner name, email and handle. Lead with the landlord-deposit angle, offer to build one free sealed report from their own photos, and follow up on day 2, 5 and 9 with a single high-dispute room type: oven interior, inside cupboards, carpet edges. Convert on the report, not on a pitch. Then ask for the property manager's email, because that address turns one cleaner into a recurring referral into PM contracts.
6. Detailed 30-day execution roadmap
Days 1-6 are build: scaffold, capture, ghost overlay, PDF, hash seal, Stripe and production launch by day 6. Day 7 builds the list. Days 8-12 are the first sales sprint: demo video, 85 touches, three free reports, three calls, first paying customer by day 12. Days 13-21 run batches two to four, ship referral links, open the property-manager segment and add the dispute ZIP. Days 22-28 run batches five to seven, harden iOS, push Founding Crew prepays and publish one anonymised walkthrough. Days 29-30 are collection and submission, with every Stripe payout reconciled before 13:00 UTC.
7. Revenue and profit scenarios
Three projections, all labelled as projections. Conservative: 3 customers at $49, $147 gross, $7 processing fees, $18 operating. Base: 15 at $49, $735 gross, $49 refunded, $33 fees, $60 operating. Upside: 30 at a blended $74 first payment, $2,220 gross, $199 refunded, $49 chargeback, $100 fees, $115 operating.
Contest ranking uses hackathon profit: gross minus refunds, chargebacks, processing fees and paid marketing, which is zero here because the channel is organic. Hosting, domains and AI delivery costs are excluded from that figure and only reduce business profit. What a sale must still clear is its own delivery cost, roughly $1.20-$3.80 per customer per month against a $49 price, so no sale is sold below cost. These are hypotheses from stated assumptions, not verified revenue.
8. Main risks and validation experiments
The first risk is conversion. Under 3% reply rate makes even the base case unreachable, so day 9 sends two subject lines and day 12 measures replies before more volume goes out. The second is credibility: day 17 emails 40 property managers asking what evidence they accept for a "room not cleaned" claim; fewer than 5 accepting dated paired photos means repositioning as internal job QA. The third is overlay drift: five real rooms shot on day 3, and if more than one in five fails with a mid-range Android, ship a tripod guide mode. The fourth is iOS Safari, handled with an upload fallback on day 24. The fifth is the incumbent field-service photo tool, answered by competing only on the sealed pair and the expiring sign-off link.
9. First 24-hour action plan
Hours 1-2 deploy the skeleton and domain. Hours 3-5 create Supabase tables and seed the twelve-item template. Hours 6-8 build camera capture with geolocation and client-side downscale. Hours 9-11 composite the ghost overlay and test it in a real kitchen and bathroom. Hours 12-14 run three Google Maps searches and log 40 cleaners with contact details. By hour 14 there is a working overlay on one device and a prospect list that makes day 9 possible.
This brief was generated from public data about established startups, then narrowed and scored for a solo 30-day launch. Revenue and profit figures are scenario models, not verified results, guarantees or evidence of existing customers. Validate demand yourself before you spend a month on it — and check the contest rules for what counts toward your score.
Related ideas
All ideas →DepthCert
Upload a short stereo video and get a calibration and depth-accuracy certificate for your robot's vision rig in minutes.
Flowproof Audit
Paste in an n8n or Make workflow export and get a scored reliability report plus drop-in patch nodes that stop silent failures, duplicate runs and untyped AI outputs before a client notices.
Missed Call Work Order
Turns every missed call at a one-to-three-truck plumbing, HVAC or electrical shop into a structured job ticket texted to the owner and a booking text to the caller within 60 seconds.