CrewPool
Generates optimized carpool rosters for construction crews driving to distant job sites, cutting fuel costs and ensuring on-time arrivals.
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.
Operations manager or project manager at a 10-100 employee construction company in Texas that sends crews to sites 30+ miles from the yard.
Market demand 4.0/10
Narrow the buyer to a segment with an urgent, already-budgeted problem and reference the specific trigger event that creates willingness to pay.
Why this could work in 30 days: one founder can ship the MVP in 4 days; the plan targets a first sale on day 12; 4 sales at $299 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 8.0/10.
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
17 concrete actions. Tick them off as you go — nothing is saved or sent.
Product opportunity and positioning
Target customer
Operations manager or project manager at a 10-100 employee construction company in Texas that sends crews to sites 30+ miles from the yard.
Why now
Firms are bidding work further from the yard to stay busy, fuel is a visible line item, and foremen already coordinate by text, so a forwarded PDF fits the existing habit with no new process.
Problem
Foremen spend hours texting to work out who rides with whom, and crews still drive separate trucks to distant sites, burning fuel, parking money and start time.
Solution
Upload a CSV of names, addresses and vehicle seats plus the site address and start time; CrewPool geocodes, clusters workers into vehicles, routes each vehicle with pick-up times, and returns a PDF plan.
CrewPool sells a document, not a network. A small construction company wins a job 30-60 miles from its yard, and the foreman spends the Friday before start day texting eight workers to work out who rides with whom. Nobody owns that task, so it gets done badly: four trucks leave the yard half-empty, parking is paid four times, and one crew arrives late on the first morning of a new project.
CrewPool takes the roster the company already keeps - names, home addresses or ZIP codes, and who owns a truck with how many seats - plus the job site address and start time, and returns a per-vehicle carpool plan: who rides with whom, the pick-up order, the clock time of each stop and the total route duration. The output is a PDF a foreman can forward on Thursday night and print on Friday morning.
Positioning is deliberately narrow. The buyer is an operations or project manager at a 10-100 person construction firm in Texas, and the trigger is a specific event: a new project starting next week at a distant site. That trigger answers the weakest score in the JEV assessment, market demand at 4.01 normalized. The purchase is not improve efficiency someday; it is we start Monday and I have nine trucks and six seats.
The wedge holds for four reasons a solo founder can test in 30 days. The buyer can approve $299 on a card without procurement. The deliverable is consumed within 48 hours, so feedback is immediate. There is no marketplace cold-start, because the group is already known and already employed by a single company. And there is no regulatory surface: no background checks, no insurance, no payment splitting between workers.
The alternative today is a spreadsheet plus group texts. The alternative software is a generic route planner that optimises one vehicle, or a telematics product sold to fleets that already have GPS. Neither produces a multi-vehicle roster that respects seat capacity and a detour ceiling. That gap - capacity-constrained clustering plus a human-readable plan - is the entire product, and it is small enough for one person to build in four days.
Score context: JEV rates speed to first revenue at 7.97 and solo build feasibility at 7.78, which is where this plan leans. It rates market demand at 4.01, which is why the first ten days are outreach against a named trigger rather than product polish.
Smallest sellable MVP
A paid PDF: upload worker CSV plus site address, pay $299, receive a multi-page carpool plan with per-vehicle routes and pick-up times.
Must-have features
- CSV upload with downloadable template (name, address, has_vehicle, seats)
- Geocoding of home addresses and the job site
- Capacity-constrained clustering into vehicles with a detour ceiling
- Per-vehicle pick-up order and clock times from Directions API
- PDF export with one page per vehicle, static map and schedule
- Stripe Checkout gating PDF generation
- Plan history dashboard for repeat use
Explicitly out of scope
- Mobile app
- Real-time driver tracking
- Driver messaging or SMS
- Payment splitting between workers
- Multi-day or multi-site trips
- Payroll or HR integration
- Safety vetting or background checks
- Marketplace of open seats
Demo flow (what the buyer sees)
- Buyer downloads the CSV template and fills worker rows
- Buyer enters job site address and site start time
- Uploader geocodes addresses and shows any rows that failed
- System clusters workers into vehicles by proximity and seats
- Directions API orders stops per vehicle against the arrival time
- Buyer pays $299 through Stripe Checkout
- PDF plan is emailed and downloadable from the dashboard
Technical implementation plan
Stack
Architecture
A single Next.js app on Vercel holds the dashboard, upload route and a Node-runtime PDF route. Uploads write rows to Supabase, geocoding results are cached by address hash so repeat workers cost nothing, clustering runs in a pure TypeScript module, Directions is called once per vehicle, and the PDF route streams a PDFKit buffer to storage. Stripe Checkout creates a session; the webhook flips the plan to paid and triggers the email with a signed download link. Nothing runs between requests.
Build steps
- Create the Next.js repo, deploy the empty app to Vercel on a purchased domain.
- Provision Supabase, create plans, workers and vehicles tables with row-level ownership.
- Set Google Cloud billing caps and enable Geocoding, Directions and Static Maps keys.
- Build CSV parse and validation with a downloadable template and per-row error display.
- Add geocoding with an address-hash cache in Supabase.
- Implement greedy clustering by nearest-neighbour against seat capacity, then a detour check.
- Call Directions per vehicle with arrival time set, derive stop order and clock times.
- Render PDF with PDFKit, one page per vehicle plus a static map image.
- Gate the PDF route behind a Stripe session check; wire the webhook and Resend email.
- Harden: quota caps, delete-my-data button, refund policy page.
AI components
- A single optional LLM pass that normalises messy pasted addresses into structured rows and flags suspicious entries before geocoding.
- No model in the routing or clustering path: routing is deterministic Directions API output so plans are reproducible and explainable.
Pricing and monetization
One-time Project Plan
Entry offerFor Operations manager planning one distant-site project
- One carpool plan for up to 30 workers
- PDF with per-vehicle routes and pick-up times
- One revision after the roster changes
Monthly Subscription
For Company with ongoing or rolling distant sites
- Unlimited carpool plans for one company
- Plan history and CSV re-use
- Priority email support
Model: One-time $299 per project plan, with a $199 monthly option for firms running distant sites continuously.
First revenue offer: Send a completed free sample plan for one of their upcoming sites, then a $299 Stripe link with the line: pay today and I will run your second site free.
Expected ACV: $299
Upsell path: Any buyer with two or more queued sites gets the $199 monthly pitch; every buyer is asked for one named referral to another project manager.
First-customer acquisition
LinkedIn outreach to operations and project managers at small Texas construction firms, leading with a free sample plan.
Where to find them
- LinkedIn search: Operations Manager or Project Manager, construction, 11-50 employees, Texas
- Firms with yard addresses far from new subdivision work in Houston, DFW, Austin and San Antonio
- Company websites listing current project addresses
- Electrical and mechanical subcontractors in Texas as a second segment test
- Buyers' referrals to project managers at other firms
Exact outreach motions
- Connection note under 300 characters offering a free sample plan for one project.
- Day-9 follow-up to non-accepts with a 40-second Loom of the sample PDF.
- Reply within one hour to any response and ask for a CSV or just worker ZIP codes.
- Send the finished sample plan, then the $299 link with the second-site-free offer.
- Day-16 website contact-form email to ten firms with no LinkedIn presence.
- Referral ask after every paid plan: one named project manager at another firm.
Message template
Hi [Name] - I build a tool that turns your crew list and site address into a carpool plan: who rides with whom, pick-up times, one page per truck. Want a free plan for one of your upcoming sites? No cost, no signup.
30-day execution roadmap
Week 1 · Build & validate · 7 days
Ship an empty app that deploys
- Create the Next.js repo and deploy to Vercel on the crewpool domain.
- Provision Supabase and create plans, workers and vehicles tables.
- Create Stripe products for $299 one-time and $199 monthly.
Deliverable: Live empty app with a working dashboard route.
CSV to clustered vehicles
- Build CSV upload with a downloadable template.
- Geocode addresses through Google Geocoding with a hard daily quota cap.
- Write greedy clustering respecting seat capacity and a 20-minute detour ceiling.
Deliverable: A 12-worker CSV returns clustered vehicles as JSON.
Per-vehicle schedule PDF
- Call Directions per vehicle against the site arrival time and derive pick-up order.
- Render a PDF with one page per vehicle, a static map and stop times.
- Test on synthetic crews of 6, 12 and 30 workers.
Deliverable: A three-page Carpool Plan PDF generated end to end.
Payment before output
- Gate PDF generation behind a Stripe Checkout session.
- Handle the webhook, mark the plan paid and email the download link via Resend.
- Add a clear no-questions refund line to suppress disputes.
Deliverable: A stranger can pay $299 and receive a PDF unattended.
Make v1 sellable
- Reject any plan where a worker adds more than 20 minutes of detour.
- Add the address data notice and a delete-my-data button.
- Cut every feature not required for a first sale.
Deliverable: Production v1 frozen except bug fixes until day 12.
150 Texas prospects
- Filter LinkedIn for operations and project managers at 11-50 employee Texas construction firms.
- Log company, name, city and any stated upcoming project.
- Prioritise firms with yards far from current subdivision work.
Deliverable: Spreadsheet of 150 named prospects with 60 marked high priority.
Demo kit
- Generate one anonymised sample plan on a 45-mile Austin route.
- Write a one-page cover sheet with the fuel and time difference.
- Draft three outreach messages under 300 characters each.
Deliverable: Demo kit: sample PDF, cover sheet, three message variants.
Week 2 · Launch & first sale · 7 days
First 25 contacts
- Send 25 LinkedIn connection notes offering a free sample plan.
- Log each send with a follow-up date in the sheet.
Deliverable: 25 sends logged.
Second 25 contacts
- Send 25 more connection notes.
- Follow up on day-8 non-accepts with a second short note.
Deliverable: 50 total sends logged.
Video proof
- Send 25 more notes plus a 40-second Loom of the sample plan.
- Publish one LinkedIn post showing the redacted sample route.
Deliverable: 75 sends and one public post live.
Convert replies to sample requests
- Reply to every response within one hour.
- Ask for a CSV or just worker ZIP codes plus the site address.
- Log objections verbatim in the sheet.
Deliverable: At least three sample-plan requests in progress.
Close the first $299
- Deliver the sample plan, then send the $299 Stripe link with a one-line summary.
- Offer to run their second site free if they pay today.
- Record the sale only after Stripe shows succeeded.
Deliverable: First $299 collected, or the blocking objection written down.
Deliver fast, fix friction
- Deliver every outstanding sample plan within 12 hours of receiving addresses.
- Fix the single most common CSV upload problem.
Deliverable: All requested plans delivered.
Real quotes only
- Ask each recipient for one sentence on time or fuel saved.
- Publish only quotes actually received, with names only if permitted.
Deliverable: One verified quote or none.
Week 3 · Repeat & systematize · 7 days
Expand the list
- Send 25 notes to DFW and San Antonio firms.
- Send ten website contact-form emails to firms with no LinkedIn presence.
Deliverable: 100 total outreach contacts.
Answer the privacy worry
- Ship a ZIP-only input mode for firms unwilling to share full addresses.
- Add one-click row deletion and a retention note.
Deliverable: Privacy objection has a product answer.
Sites starting within ten days
- Offer three firms a free plan for a site starting within ten days.
- Book a second paid-plan conversation with each.
Deliverable: Three free plans scheduled.
Second and third sales
- Send the $299 link to every sample recipient who has not paid.
- Phone two warm prospects if a number is public.
Deliverable: Two additional close attempts logged.
Manual path timing
- Build three plans in one sitting and time each step.
- Remove any step taking more than five minutes.
Deliverable: Manual plan build time under 20 minutes.
Subscription pitch
- Pitch $199 monthly to any buyer with two or more queued sites.
- Add plan history to the dashboard so repeat use is obvious.
Deliverable: Subscription pitch sent to every past buyer.
Keep what worked
- Rank message variants by reply rate and keep the best.
- Drop the worst-performing segment from the list.
Deliverable: One winning message and one segment selected for week four.
Week 4 · Scale what works · 7 days
130 contacts
- Send 30 notes using the winning variant.
- Follow up with all 40 warm non-responders.
Deliverable: 130+ total contacts logged.
Borrow trust
- Ask every buyer for one named project manager at another firm.
- Write a two-line blurb buyers can paste and forward.
Deliverable: Three referral asks sent.
Subcontractors
- Test the same offer on 15 Texas electrical or mechanical subcontractors.
- Compare reply rate against general contractors.
Deliverable: Segment reply-rate comparison logged.
Visibility
- Post a redacted before-and-after route plan on LinkedIn.
- Comment on ten Texas construction operations posts with the sample offer.
Deliverable: One public post and ten comments live.
Shorten the path
- Add an in-app example dataset so buyers can try before uploading.
- Collapse checkout to a single page.
Deliverable: Demo-to-checkout path in two clicks.
Reconcile to the rules
- Publish the refund policy and confirm Stripe payouts reconcile.
- Verify every sale falls inside the contest window and covers delivery cost.
Deliverable: Reconciliation sheet matching Stripe to the contest rules.
December and January starts
- Send 30 notes focused on sites starting in December and January.
- Re-offer free sample plans to every silent contact.
Deliverable: 160+ total contacts logged.
Days 29-30 · Consolidate · 2 days
Collect before the deadline
- Close every pending $299 link before 13:00 UTC on day 30.
- Deliver all paid plans and refund anything undeliverable.
Deliverable: All paid plans fulfilled.
Freeze and document
- Freeze sales and export the Stripe and TrustMRR ledgers.
- Write the profit reconciliation and the three things that did not work.
Deliverable: Final profit statement and a one-page post-mortem.
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
- Base case: 160 outreach contacts produce 15 free sample plans, of which 8 convert to a $299 one-time plan.
- Conservative: reply and conversion rates run at roughly half the base case, yielding 3 one-time plans.
- Upside: 20 buyers split 16 one-time plans at $299 and 4 monthly subscriptions at $199, giving a blended ARPU of $279.
- Payment fees assume Stripe at 2.9% plus $0.30 per transaction, modelled as a blended 3.5% of collected revenue, 3.7% in the upside where monthly billing adds transactions.
- Ad spend is zero in every scenario because acquisition is organic LinkedIn outreach, consistent with the concept assumption that no paid marketing is needed.
- Operating costs cover Vercel, Supabase, domain and Google Maps API usage at roughly $5-9 per plan generated including free samples.
- All figures are projections from these stated assumptions and are not verified revenue, customers or traction.
Conservative
3 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $31
- Paid marketing
- $0
- Hackathon profit
- $866
- Operating costs
- $45
- Business profit
- $821
Base case
8 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $84
- Paid marketing
- $0
- Hackathon profit
- $2,308
- Operating costs
- $80
- Business profit
- $2,228
Optimistic
20 customers- Refunds
- $299
- Chargebacks
- $0
- Payment fees
- $206
- Paid marketing
- $0
- Hackathon profit
- $5,075
- Operating costs
- $170
- Business profit
- $4,905
Model your own numbers
Runs in your browser · nothing is saved- Collected revenue
- $2,392
- Payment fees (3.9%)
- $93
- Paid marketing
- $0
- Hackathon profit
- $2,299
- Business profit
- $2,219
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 derived from stated assumptions. These are projections, not verified revenue, existing customers or real traction.
Main risks and validation experiments
Risk 1LinkedIn outreach to construction operations managers produces too few replies.
Use two-touch sequencing, add a Loom video, and test a subcontractor segment in parallel.
Send 25 connection notes on day 8 and count accepts and replies by day 11.
Fewer than two replies from 75 sends by day 11 means switch to phone-first outreach and the subcontractor segment.
Risk 2Firms refuse to share worker home addresses for privacy reasons.
Ship ZIP-only input, a deletion button and a written retention note.
Ask five prospects on the first call what minimum location data they would release.
If zero of five will share even ZIP codes, reposition the input as depot-to-site staging and re-test before spending more days.
Risk 3Workers refuse to carpool, so the plan is never used after delivery.
Enforce a 20-minute detour ceiling and let the buyer set minimum vehicle occupancy.
Ask each buyer on the call how many of their crew would actually accept a shared ride.
Two buyers state that fewer than half the crew would carpool; the deliverable has no value and the concept needs a different buyer.
Risk 4Routing produces impractical routes with long detours or wrong pick-up times.
Hard detour rejection, backwards time solving from site start, and a manual stop-order override.
Run three synthetic crews and check every stop against a 20-minute detour limit.
More than 25% of stops exceed the detour ceiling on synthetic data means the clustering logic is not sellable yet.
Risk 5Google Maps API cost grows faster than expected once free sample plans are generated.
Cache geocodes by address hash, cap daily quota, and render one static map per vehicle only.
Measure actual API cost across the first ten generated plans.
Cost above $15 per plan means drop static maps and re-quote before generating more free samples.
Risk 6The $299 decision stalls inside a slow construction procurement habit.
Keep the charge card-sized, offer a free sample first, and ask directly who can approve it.
Ask each respondent: can you approve $299 on a company card today?
Three buyers say it needs approval above their card limit means shift the first offer to a single-site pilot and re-price the entry point.
Hour-by-hour: the first 24 hours
Hours 1-2
- Buy the domain, create the Next.js repo and deploy the empty app to Vercel.
- Create the Supabase project and the plans table.
- Enable Google Maps APIs and set billing caps at $50.
Output: Deployed skeleton with working API keys and a dashboard route.
Hours 3-5
- Build CSV upload with the template download.
- Wire geocoding with an address-hash cache.
Output: Upload returns geocoded coordinates for every row.
Hours 6-8
- Implement greedy clustering with seat capacity.
- Call Directions per vehicle against a fixed arrival time.
Output: A 12-worker synthetic crew produces four vehicle routes.
Hours 9-11
- Render the PDF with PDFKit, one page per vehicle plus static map.
- Check page layout on a phone screen.
Output: First complete Carpool Plan PDF.
Hours 12-14
- Add Stripe Checkout and the webhook that marks a plan paid.
- Send the download link by email through Resend.
Output: Paid flow that issues a PDF without manual steps.
Hours 15-17
- Build a 40-row prospect sheet from LinkedIn filters.
- Send the first ten connection notes.
Output: First ten real outreach sends logged.
Hours 18-20
- Write the offer message and the one-page demo cover sheet.
- Generate the anonymised 45-mile sample plan.
Output: Demo kit ready to attach in any reply.
Hours 21-24
- Run the full flow end to end with a stranger-grade dataset.
- Fix the top three breakages found.
Output: Working v0.5 and a written list of what still blocks a sale.
Full 30-day launch report
Read the full generated report
CrewPool v1 - 30-Day Launch Report
Concept CrewPool v1, JEV score 61/100 (band C). This is the operator manual for one founder from 16 November 2026 13:00 UTC to 16 December 2026 13:00 UTC. Every financial figure below is a projection from stated assumptions, not verified revenue, existing customers or traction.
1. Product opportunity and positioning
CrewPool sells a document, not a network. A small construction company wins a job 30-60 miles from its yard, and the foreman spends the Friday before start day texting eight workers to work out who rides with whom. Nobody owns that task, so it gets done badly: four trucks leave the yard half-empty, parking is paid four times, and one crew arrives late on the first morning of a new project.
CrewPool takes the roster the company already keeps - names, home addresses or ZIP codes, and who owns a truck with how many seats - plus the job site address and start time, and returns a per-vehicle carpool plan: who rides with whom, the pick-up order, the clock time of each stop and total route duration. The output is a PDF a foreman can forward on Thursday night and print on Friday morning.
Positioning is deliberately narrow. The buyer is an operations or project manager at a 10-100 person construction firm in Texas, and the trigger is a specific event: a new project starting next week at a distant site. That trigger answers the weakest score in the JEV assessment, market demand at 4.01 normalised. The purchase is not improve efficiency someday; it is we start Monday and I have nine trucks and six seats.
The wedge holds for four reasons a solo founder can test in 30 days. The buyer can approve $299 on a card without procurement. The deliverable is consumed within 48 hours, so feedback is immediate. There is no marketplace cold-start, because the group is already known and already employed by one company. And there is no regulatory surface: no background checks, no insurance, no payment splitting between workers.
The alternative today is a spreadsheet plus group texts. The alternative software is a generic route planner that optimises one vehicle, or a telematics product sold to fleets that already have GPS. Neither produces a multi-vehicle roster that respects seat capacity and a detour ceiling. That gap is the entire product, and it is small enough for one person to build in four days.
Score context: JEV rates speed to first revenue at 7.97 and solo build feasibility at 7.78, which is where this plan leans. It rates market demand at 4.01, which is why the first ten days are outreach against a named trigger rather than product polish.
2. Smallest sellable MVP
The MVP is a paid PDF. The buyer uploads a CSV of worker rows, enters the site address and start time, pays $299 through Stripe Checkout, and receives a multi-page plan: one page per vehicle with a static map, ordered stops and clock times. Nothing else ships.
Must-haves are the CSV template, geocoding, capacity-constrained clustering, a 20-minute detour ceiling, Directions-backed stop ordering, PDF export, the Stripe gate and a plan history list. Out of scope: mobile app, live tracking, driver messaging, payment splitting, multi-day trips, payroll integration and any safety vetting. Those are excluded because none of them can be sold before the first $299 is collected.
3. Technical implementation plan
Next.js 14 on Vercel, Supabase for plans, workers, vehicles and a geocode cache, Google Maps Geocoding, Directions and Static Maps, PDFKit in a Node-runtime route, Stripe Checkout plus webhook, Resend for delivery.
Build order: deploy an empty app, then CSV parse and validation, then geocoding with an address-hash cache, then greedy nearest-neighbour clustering against seat capacity, then a detour rejection pass, then Directions per vehicle with arrival time set to the site start, then PDFKit rendering, then the Stripe gate and webhook, then quota caps and a delete-my-data control.
The hardest technical risk is pick-up clock times. Directions returns durations for the current moment unless departure time is set correctly, and a plan with wrong pick-up times is worthless to a foreman. Solve it by working backwards from the site start per vehicle and validating against three fixed synthetic crews before any customer sees output. Do not use Puppeteer here: Chromium blows the serverless bundle limit, and PDFKit is enough for pages, tables and images. An LLM is used in exactly one optional place, to normalise messy pasted addresses before geocoding; the routing and clustering paths stay deterministic so plans are reproducible.
4. Pricing and monetization
$299 one-time per project plan, up to 30 workers, including one revision after the roster changes. $199 per month for companies with rolling distant sites and unlimited plans. No free tier: the free sample plan is the trial, delivered by the founder, which keeps the sales conversation warm without a self-serve funnel.
At roughly $5-9 of Google Maps usage per plan, delivery cost is under 4% of the $299 ticket. That is the reason a per-project price works: the work is bounded, the buyer understands it instantly, and it clears a card-approval threshold without procurement.
5. First-customer acquisition
One channel: LinkedIn outreach to operations and project managers at 11-50 employee Texas construction firms. Filter, log 150 names with company, city and any stated upcoming project, and prioritise yards far from current subdivision work. Send 25 connection notes per day from day 8, each offering a free sample plan for one upcoming site.
The message: hi [Name] - I build a tool that turns your crew list and site address into a carpool plan: who rides with whom, pick-up times, one page per truck. Want a free plan for one of your upcoming sites? No cost, no signup.
Reply within one hour. Ask for a CSV, or just worker ZIP codes and the site address. Deliver the sample plan, then send the $299 link with the offer to run their second site free. Every paid buyer gets one referral ask for a named project manager at another firm. Week-one goal: 150 prospects logged and 25 notes sent.
6. 30-day execution roadmap
Days 1-5 build and freeze the product, with the MVP deadline on day 5. Days 6-7 build the prospect list and the demo kit. Days 8-11 run three outreach waves to 75 contacts. Day 12 is the first-sale target: deliver a sample and send the $299 link. Days 13-14 fulfil fast and collect only real quotes. Days 15-21 expand to 100 contacts, ship the ZIP-only privacy answer, land three free samples and pitch the $199 subscription. Days 22-28 scale to 160+ contacts, run referral asks, test subcontractors as a second segment, and reconcile Stripe against the contest rules. Days 29-30 close pending links, fulfil everything and freeze sales at 13:00 UTC on day 30.
7. Revenue and profit scenarios (projections)
Conservative: 3 one-time plans at $299, gross $897, less $31 fees and $45 operating cost. Base: 8 one-time plans, gross $2,392, less $84 fees and $80 operating cost. Upside: 20 buyers blended at $279 (16 one-time at $299 plus 4 subscriptions at $199), gross $5,580, less $299 refund, $206 fees and $170 operating cost.
Hackathon profit, which excludes hosting and API cost, is roughly $866 conservative, $2,308 base and $5,075 upside. Business profit after operating cost is roughly $821, $2,228 and $4,905. These are projections built from stated assumptions about reply rates, sample conversions and blended ARPU. They are not verified revenue and must be replaced by the actual Stripe ledger on day 30.
8. Main risks and validation experiments
Reply rate is the first risk. Test it with 25 notes on day 8; fewer than two replies from 75 sends by day 11 means switching to phone-first outreach and the subcontractor segment. Privacy refusal is the second: ship ZIP-only input and ask five prospects what minimum data they will release; if none will share ZIP codes, reposition the input as depot-to-site staging. Worker refusal is the third: ask on every call how many of the crew would accept a shared ride; two buyers saying fewer than half ends the concept. Route quality and Google Maps cost are the operational risks, measured on the first ten plans with a $15-per-plan ceiling.
9. First 24-hour action plan
Hours 1-2: domain, repo, Vercel deploy, Supabase, Google API keys with a $50 billing cap. Hours 3-5: CSV upload and cached geocoding. Hours 6-8: clustering plus Directions on a synthetic crew. Hours 9-11: PDFKit render with static maps. Hours 12-14: Stripe Checkout, webhook and Resend delivery. Hours 15-17: 40-row prospect sheet and the first ten connection notes. Hours 18-20: offer message, cover sheet and the anonymised 45-mile sample plan. Hours 21-24: run the full flow end to end and fix the top three breakages.
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 →Bulletin Nounou
Bulletin Nounou turns a French childminder's month of attended days into a ready-to-send bulletin de salaire PDF plus a copy-paste Pajemploi figure card, in under three minutes per child.
Standby Desk
A private shift-fill desk for restaurants, hotels and clinics that turns their existing list of past extras into one form, one blast, one accept page, and one mission sheet PDF.
Kitchen Sheet
Kitchen Sheet turns a cloud kitchen's messy WhatsApp lunch orders into a print-ready kitchen prep sheet, one label per order, and a delivery run order in under 60 seconds.