Score C · 5.96/10Free to read$100K Hackathon · November 16, 2026 → December 16, 2026Nov 15, 2026

Inbox Leak Audit

A one-time, paid audit that scans a shared Gmail support inbox and returns a client-ready PDF naming every customer email that was never answered, how long it sat, and what it likely cost.

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.

Speculative5.96/10
MVP build (solo)
5 days
Launch price
$149
Sales to $1,000 revenue
7 customers
First paying customer
Direct LinkedIn and X outreach to named heads of customer experience and support/VA agency owners, opened with a free five-thread sample audit.
Who pays

Owner or head of customer experience at a 5-50 person e-commerce brand, or owner of a 2-20 person support/VA agency managing client inboxes, on Google Workspace in the US/UK/CA/AU, triggered by a missed email that became a refund, chargeback or public complaint.

Biggest execution risk

Market demand 4.3/10

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 5 days; the plan targets a first sale on day 10; 7 sales at $149 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 8.2/10.

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.

0/10
Entry price
$149
Solo build
5 days
30d base revenue
$1,920
30d hackathon profit
$1,843
Section 1 of 8

Product opportunity and positioning

Target customer

Owner or head of customer experience at a 5-50 person e-commerce brand, or owner of a 2-20 person support/VA agency managing client inboxes, on Google Workspace in the US/UK/CA/AU, triggered by a missed email that became a refund, chargeback or public complaint.

Why now

Support teams are being asked to cut headcount while volume rises; a $149 diagnostic that quantifies dropped demand is cheaper than a helpdesk migration and needs no procurement.

Problem

Unanswered customer emails are invisible in aggregate, so nobody can say how many threads were dropped last quarter, how long they sat, or which ones were refund and quote requests.

Solution

Buyer connects one Gmail inbox read-only and pays $149; the product rebuilds 90 days of threads, flags unresolved ones, classifies them, ranks by revenue at risk, and emails a branded PDF plus CSV appendix in about ten minutes.

Differentiation · Helpdesks sell a new inbox and a migration; freelancers and VAs sell hours and a spreadsheet. This sells one forwardable document with a dollar ranking, in minutes, at a one-time price.

1. Product opportunity and positioning

Shared inboxes hide their own failures. Gmail and every helpdesk show what is open now; none of them produce the artifact an owner actually needs after a bad week: a dated list of every customer email that never got a reply, how long it sat, and which of those were refund or quote requests. That gap is the product. Inbox Leak Audit attaches read-only to the inbox a team already runs, reconstructs 90 days of threads, and returns a branded PDF ranked by revenue at risk.

The buyer is deliberately narrow: the owner or head of CX at a 5-50 person e-commerce brand on Google Workspace in the US, UK, Canada or Australia, or the owner of a 2-20 person support/VA agency running client inboxes. The trigger is a specific event, not a mood: a missed email that turned into a refund, a chargeback or a public complaint, or a Monday backlog nobody can quantify. That buyer does not want a migration, a new workspace or a per-seat subscription. They want evidence, this week, that costs less than one lost order.

Against alternatives: helpdesks (Zendesk, Intercom, Front, Freshworks) sell the inbox itself and require migration and a subscription; analytics add-ons need clean data and weeks; a VA doing this by hand takes a day and produces a spreadsheet nobody forwards. This product produces a client-ready document in about ten minutes for $149, with no integration beyond a read-only OAuth consent.

The JEV score is 59.6 (band C) and it is honest about the weak spots. Market demand is the lowest sub-score at 4.3, which is why the buyer is defined by trigger event rather than firmographics. Monetisation at 4.94 is capped by a one-time price, which is why the plan ships a $449 agency 5-pack and a $49 monthly re-scan. Operational simplicity at 4.74 reflects the friction of Gmail restricted-scope verification, which is why the first 30 days run on a test-user allowlist plus an export-upload fallback instead of waiting on Google review.

Speed to first revenue is the strongest sub-score at 8.18, and the whole plan is built to exploit it: the deliverable exists on day 5, samples go out on day 7, and the first paid audit is targeted for day 10. The competitive defensibility score (6.04) is not defended by code, it is defended by the fact that a one-person operator can hand-build samples faster than an incumbent can ship a report feature.

The output is deliberately a document, not a dashboard. A PDF with a dollar ranking can be forwarded to a boss or a client in one click, survives a screenshot, and needs no onboarding call. That is the whole distribution strategy: the artifact is the marketing.

2. Smallest sellable MVP

The smallest sellable version is one inbox, one payment, one PDF. Everything else is deferred.

Five capabilities must exist on day 5. First, Gmail read-only OAuth with a test-user allowlist and a CSV export upload fallback, so no buyer is blocked by a Google verification queue. Second, 90-day message ingestion and thread reconstruction that flags threads whose last message is inbound with no later outbound reply, and computes hours since that inbound message. Third, an LLM classification pass sorting each unresolved thread into order issue, refund, quote request, complaint or noise, with a confidence value. Fourth, a revenue-at-risk score combining category weight, thread age and recency into one sortable number. Fifth, PDF rendering plus Stripe gating plus email delivery, so the document is released only after payment clears.

Explicitly not built in the first 30 days: a live inbox, reply drafting, SMS/social/voice ingestion, seats and permissions, CRM or Shopify integrations, an analytics dashboard, SOC 2 claims, and any human review step. Each of those either adds a support surface or delays the first sale.

3. Technical implementation plan

Next.js on Vercel, Supabase Postgres for audit, thread and payment records, Gmail API with the read-only scope, an LLM API for classification, HTML-to-PDF rendering, Stripe Checkout, and Resend for delivery. No queue, no worker fleet, no custom ML.

Build order: auth and ingestion, thread reconstruction, classification and scoring, PDF and payment gate, then hardening. Ingestion caps at 500 threads per audit so API cost per report stays under about $0.40. Classification sends only the first and last message of each thread, not full bodies, which keeps prompts small and output consistent.

The hardest technical risk is thread reconstruction, not the LLM. Gmail threadIds collapse subject variants, and a buyer who replied from a personal address or by phone leaves a thread that looks unanswered when it was answered. Mitigation: search outbound messages in a +/- 24 hour window against the same recipient address before marking a thread unresolved, and show a confidence flag on ambiguous rows so the buyer sees the methodology.

4. Pricing and monetization

Three prices. $149 Single Inbox Audit, one-time, 90-day scan, branded PDF, CSV appendix, 30-day shareable link. $449 Agency 5-Pack, one-time, five client inboxes, white-label PDF with the agency logo, per-client CSVs, 24-hour turnaround, which works out to about $89.80 per inbox. $49 per month Monthly Leak Watch for weekly re-scans and a delta PDF. Past buyers get a $79 re-run.

Payment processing is assumed at roughly 4% blended (Stripe 2.9% plus 30 cents, weighted toward the $149 ticket). No paid ads are planned; the acquisition budget is zero and outreach is manual. Delivery cost per audit is inference plus hosting, measured on day 19 and kept under $1.

5. First-customer acquisition

One channel: direct LinkedIn and X outreach, opened with a free five-thread sample. The offer is not a demo call, it is a delivered mini-PDF within 24 hours.

Sequence: build a 60-account list by day 1 and 150 by day 12, sending 20 messages on day 7, 40 on day 8, and 40-60 per wave after that. Every sample ends with a Stripe link. Two follow-ups: 72 hours with the sample's worst finding restated, day 7 with the agency bundle. One build-in-public post per week using an anonymised real finding.

Week-one goal: 60 messages sent, 6 samples delivered, 3 Stripe links out. Base case assumes roughly a 1-in-12 sample-to-payment rate, which is the single number to watch.

6. 30-day execution roadmap

Days 1-5 are build: shell and list, ingestion, thread reconstruction, classification and scoring, then PDF plus payment gate, with a live end-to-end paid test on day 5. Days 6-7 are hardening and the first 20 samples. Days 8-14 are outreach waves with the first paid audit targeted on day 10, an agency 5-pack shipped on day 13, and a funnel review on day 14. Days 15-21 add the recurring tier, referrals, a cost reduction pass and a midpoint pricing decision. Days 22-28 run re-run campaigns, testimonials, a zero-touch purchase path, a lead magnet and an agency affiliate code. Days 29-30 are final-call messages, full fulfilment, Stripe reconciliation and an honest public recap. US Thanksgiving (day 11) is treated as a low-reply day used for product work, not outreach volume.

7. Revenue and profit scenarios

All figures are projections from stated assumptions, not verified revenue.

Conservative: 3 customers at $149 blended ARPU, $447 gross, no refunds, no chargebacks, about $18 in processing fees, no ad spend, $45 operating cost.

Base: 12 customers at $160 blended ARPU (one agency 5-pack pulling the average up), $1,920 gross, $77 processing fees, no ad spend, $120 operating cost.

Upside: 40 customers at $165 blended ARPU, $6,600 gross, one refund at $149, no chargebacks, $264 processing fees, $150 ad spend on two sponsored community posts, $240 operating cost.

The contest ranks hackathon profit, which deducts refunds, chargebacks, processing fees and paid marketing but not hosting or inference. Business profit additionally deducts operating cost. No scenario prices a sale below its delivery cost.

8. Main risks and validation experiments

Gmail restricted-scope verification is the top operational risk; the test-user allowlist and export fallback exist so launch never depends on a Google review. Validation: connect a second real Workspace account on day 3 and confirm ingestion without a verification prompt.

Cold outreach reply rate is the demand risk. Validation: send 20 messages on day 7 and count sample requests; if fewer than two of twenty ask for a sample, the opener is wrong, not the market.

Misclassification risk can produce a wrong headline number and a refund. Validation: hand-label 30 threads from one real inbox and compare against the model before the first paid delivery.

Trust risk: a support owner may refuse inbox access to an unknown founder. Validation: offer the export-upload path in the first message and measure how many choose it.

9. First 24-hour action plan

Hours 1-2: create the repo, the Vercel project and the Supabase schema; publish a landing page with the $149 offer and a live Stripe Checkout link. Output: a URL that can take money.

Hours 3-4: build the 60-account prospect list with name, role, brand, LinkedIn or X URL and public support address. Output: a sheet, not a notion of a list.

Hours 5-7: wire Google OAuth read-only against the founder's own Workspace inbox and confirm 90 days of messages land in Supabase. Output: raw messages stored.

Hours 8-10: write the sample offer message and the two follow-ups; send the first 10 messages. Output: 10 conversations opened.

Hours 11-12: write the day-5 definition of done and the exact PDF layout (cover, ranked table, three fixes, appendix). Output: a one-page spec the build follows.

Section 2 of 8

Smallest sellable MVP

One Gmail inbox connected read-only, $149 paid via Stripe Checkout, and a branded PDF with a ranked list of unanswered threads emailed in under ten minutes.

Must-have features

  • Google OAuth with gmail.readonly scope and a test-user allowlist
  • 90-day ingestion with a CSV/export upload fallback for buyers who refuse OAuth
  • Thread reconstruction: unresolved flag, hours since last inbound, first-response time
  • LLM classification into order issue, refund, quote request, complaint or noise with confidence
  • Revenue-at-risk score and ranked thread table
  • PDF report: cover summary, ranked table, three recommended fixes, CSV appendix
  • Stripe Checkout gating report release
  • Resend email delivery with a 30-day signed report link

Explicitly out of scope

  • No live shared inbox or reply drafting
  • No SMS, social, WhatsApp or voice ingestion
  • No team seats, roles or permissions
  • No CRM, helpdesk or Shopify integrations
  • No ongoing analytics dashboard
  • No SOC 2 or enterprise security review
  • No human review step in delivery

Demo flow (what the buyer sees)

  1. Buyer clicks the $149 Stripe Checkout link on the landing page
  2. Buyer consents to read-only Gmail access, or uploads a mailbox export
  3. Product ingests 90 days of messages and rebuilds threads
  4. Unresolved threads are classified and scored by revenue at risk
  5. Stripe webhook confirms payment and unlocks generation
  6. PDF and CSV are emailed within ten minutes with a 30-day shareable link
Build plan

Technical implementation plan

Stack

Next.js on VercelSupabase PostgresGmail API with gmail.readonly scopeAnthropic or OpenAI API for thread classificationHTML-to-PDF renderingStripe Checkout and webhooksResend transactional emailVercel Cron for weekly re-scans

Architecture

A single Next.js app on Vercel handles OAuth, ingestion and report rendering. Supabase stores raw message metadata, reconstructed threads, classification output and one audit row per payment. The Gmail API is read-only and called with the buyer's token; nothing is written back. Classification is one batched LLM call per audit, sending only the first and last message per unresolved thread. The PDF is rendered server-side, stored in Supabase Storage behind a signed 30-day URL, and emailed via Resend. A Stripe webhook flips the audit from pending_payment to generating; the report never renders before that flip.

Build steps

  1. Create Next.js repo, Vercel project and Supabase schema for audits, threads and messages
  2. Ship landing page with the $149 offer and a live Stripe Checkout link
  3. Implement Google OAuth with read-only scope behind a test-user allowlist
  4. Ingest 90 days of messages and persist raw metadata
  5. Reconstruct threads and compute unresolved flag, age in hours and first-response time
  6. Add LLM classification plus a deterministic revenue-at-risk score
  7. Render the PDF and CSV, store behind a signed URL, deliver via Resend
  8. Wire Stripe webhook to gate generation on successful payment
  9. Add the CSV/export upload fallback path for non-OAuth buyers
  10. Run a full paid test on the founder's own inbox and time it
  11. Harden: failure emails, per-audit status rows, retry on Gmail 429s
  12. Ship Monthly Leak Watch as a Vercel Cron re-scan with delta PDF

AI components

  • LLM classification of each unresolved thread into five categories with a confidence score
  • LLM-generated three recommended fixes on the report cover page
  • Optional LLM summarisation of the worst five threads for the executive summary
Hardest technical risk: Thread reconstruction producing false 'dropped' findings when a thread was actually answered from another address or channel; mitigation is a +/- 24 hour outbound search by recipient address plus a confidence flag on ambiguous rows.
Section 3 of 8

Pricing and monetization

Single Inbox Audit

Entry offer
$149 one-time

For Support lead or owner at a 5-50 person e-commerce brand auditing one shared inbox

  • 90-day thread scan
  • Branded PDF report
  • CSV appendix of unresolved threads
  • 30-day shareable report link

Agency 5-Pack

$449 one-time

For Support or VA agency owner running audits across client inboxes

  • 5 inbox audits
  • White-label PDF with agency logo
  • Per-client CSV appendices
  • 24-hour turnaround

Monthly Leak Watch

$49 / month

For Buyer who wants a recurring re-scan and week-over-week delta

  • Weekly re-scan of one inbox
  • Delta PDF of new dropped threads
  • Email alert when a thread passes 48 hours unresolved

Model: One-time audit as the entry product, a discounted agency 5-pack for repeat volume, and a $49 monthly re-scan for buyers who want the delta.

First revenue offer: Send this: 'I ran a quick read on five threads from your support inbox — three were never answered, the oldest is 11 days old and one is a refund request. Full 90-day audit, ranked by revenue at risk, $149, PDF in your inbox in ten minutes. Want it?'

Expected ACV: $149

Upsell path: Every buyer gets a $79 re-run at day 14, an invitation to $49/month Leak Watch at day 17, and the $449 agency 5-pack if they manage more than one inbox.

Target customers
40
Revenue model
$5,960
Costs
$140
Profit model
$5,820
Section 4 of 8

First-customer acquisition

Primary channel

Direct LinkedIn and X outreach to named heads of customer experience and support/VA agency owners, opened with a free five-thread sample audit.

Where to find them

  • LinkedIn search: 'Head of Customer Experience' at 5-50 person brands with a visible Shopify storefront
  • LinkedIn search: 'Customer Support Manager' at DTC brands running a public support@ address
  • X search: owners posting about support backlog, missed emails or refund complaints
  • Support and VA agency owners on X and LinkedIn with 'we manage client inboxes' in the bio
  • E-commerce operator Slack and Discord communities the founder is already a member of
  • Public support@ addresses harvested from brand contact pages

Exact outreach motions

  • Day 7: 20 LinkedIn DMs offering to audit five threads they paste or forward, no access required
  • Day 8: 40 messages split 25 e-commerce brands and 15 agency owners, each naming one specific finding pattern to expect
  • Day 9: 72-hour follow-up restating the worst finding from their delivered sample
  • Day 12: day-7 follow-up with the $449 agency bundle and a 24-hour turnaround promise
  • Day 13: offer three agencies a paid $449 pilot instead of a free one
  • Weekly: post an anonymised mini-PDF as a build-in-public thread in one operator community

Message template

Hi {first_name} — you run support at {brand}. I pull five threads from a support inbox and show you which ones never got a reply. Last week three of five had been sitting 6+ days, one was a refund request. Want me to run the five for free? No access needed, just paste or forward them.
Week-one goal: 60 outreach messages sent, 6 free five-thread samples delivered, 3 Stripe links issued.
Section 5 of 8

30-day execution roadmap

MVP deadline: day 5First sale target: day 1030 days planned
Week 1 · Build & validate · 7 days
Day 1Setup

Ship the shell and the list

  • Create the Next.js repo, Vercel project and Supabase schema.
  • Publish a landing page with the $149 offer and a live Stripe Checkout link.
  • Build a 60-account prospect sheet with name, role, brand and contact URL.
  • Write the five-thread sample offer and both follow-ups.

Deliverable: Live landing page that takes money, plus 60 named prospects.

Day 2Ingestion

Read-only Gmail into the database

  • Wire Google OAuth with gmail.readonly behind a test-user allowlist.
  • Fetch 90 days of messages with messages.list and messages.get.
  • Persist raw message metadata in Supabase.
  • Add the CSV/export upload fallback path.

Deliverable: One real inbox ingested end to end.

Day 3Threads

Reconstruct threads and find the dropped ones

  • Group messages by threadId and References headers.
  • Compute first inbound to first outbound response time.
  • Flag unresolved threads where the last message is inbound.
  • Test against three fixtures including a 20-message thread.

Deliverable: Unresolved-thread engine with hours-open metric.

Day 4Scoring

Classify and rank by revenue at risk

  • Send first and last message per thread to the LLM for category.
  • Map output to order issue, refund, quote, complaint or noise with confidence.
  • Score revenue at risk from category weight, age and recency.
  • Cap audits at 500 threads to bound API cost.

Deliverable: Ranked, categorised thread list as JSON.

Day 5ShipMVP shipped

PDF, payment gate, delivery

  • Render the report HTML to PDF with a CSV appendix.
  • Gate generation behind the Stripe Checkout webhook.
  • Email the PDF via Resend with a 30-day signed link.
  • Charge your own card and run a full paid audit end to end.

Deliverable: MVP live: pay, connect, PDF arrives in under ten minutes.

Day 6Harden

Fix what the live test exposed

  • Fix failures found in the paid test run.
  • Add a failure email and a per-audit status row.
  • Retry Gmail 429s with backoff.
  • Time three full audits and record runtime.

Deliverable: Reliable run with a measured runtime.

Day 7Samples

First outreach wave

  • Build the one-page mini-PDF sample template.
  • Send 20 LinkedIn messages with the free five-thread offer.
  • Deliver two samples within 24 hours.
  • End every sample with a Stripe link.

Deliverable: 20 messages sent, 2 mini-PDFs delivered.

Week 2 · Launch & first sale · 7 days
Day 8Outreach

Volume and the first replies

  • Send 40 messages, 25 brands and 15 agency owners.
  • Post the anonymised mini-PDF on LinkedIn and X.
  • Reply to every response within two hours.
  • Log objections in the prospect sheet.

Deliverable: 60 total messages, first replies logged.

Day 9Samples

Convert samples into links

  • Deliver four more samples.
  • Lead each email with the single worst dropped thread.
  • Send the 72-hour follow-up to the first 20 prospects.
  • Draft the agency 5-pack one-pager.

Deliverable: 6 samples delivered, 20 follow-ups sent.

Day 10RevenueFirst sale target

First paid audit

  • Send Stripe links to the three warmest prospects.
  • Close the first $149 audit and deliver within 24 hours.
  • Ask for a one-line quote on the report.
  • Fix anything the first buyer flags.

Deliverable: First paid audit delivered and paid.

Day 11Quiet day

US Thanksgiving, low reply volume

  • Send only 15 messages.
  • Ship the $79 re-run discount code for past buyers.
  • Add total revenue-at-risk to the PDF cover.
  • Write follow-up two for non-responders.

Deliverable: Re-run offer live, cover page improved.

Day 12Outreach

Second wave plus proof

  • Send 40 messages including 15 agency owners.
  • Send follow-up two with the agency bundle.
  • Publish build-in-public post two with an anonymised real number.
  • Reply to inbound requests same day.

Deliverable: 100 total messages sent, 2 public posts live.

Day 13Agencies

Make the 5-pack buyable

  • Add the white-label PDF variant with an agency logo field.
  • Build the five-audit bulk flow from one upload.
  • Offer three agencies a paid $449 pilot.
  • Add the 24-hour turnaround badge to the landing page.

Deliverable: Agency 5-Pack purchasable and pitched to 3 agencies.

Day 14Review

Funnel audit and opener rewrite

  • Count messages, replies, samples, links and sales.
  • Kill the weakest message angle and write two new openers.
  • Email every delivered buyer the $79 re-run.
  • Ship the highest-priority PDF fix.

Deliverable: Week-2 funnel numbers and two new openers.

Week 3 · Repeat & systematize · 7 days
Day 15Outreach

New opener at volume

  • Send 50 messages with the strongest opener.
  • Deliver every pending audit same day.
  • Post the mini-PDF in one operator community.
  • Ask two buyers for a peer referral.

Deliverable: 50 new conversations, 2 referral asks made.

Day 16Recurring

Ship Monthly Leak Watch

  • Build the weekly re-scan cron.
  • Generate a delta PDF of threads dropped since the last scan.
  • Add the 48-hour unresolved email alert.
  • Put the $49/month button next to the $149 one.

Deliverable: Monthly Leak Watch live and purchasable.

Day 17Upsell

Convert one-time buyers to recurring

  • Pitch Leak Watch to every past buyer at $49/month.
  • Send 40 agency-owner messages.
  • Deliver all pending audits.
  • Track one-time to recurring conversion in Supabase.

Deliverable: Recurring offers sent to 100% of past buyers.

Day 18Agencies

Seed agency trials

  • Offer five agencies a free sample on one client inbox.
  • Publish an anonymised refund-thread sample.
  • Follow up on every unanswered sample request.
  • Add the top repeated question to the FAQ.

Deliverable: 3 agency trials running.

Day 19Cost

Measure and cut delivery cost

  • Batch-deliver audits and record runtime per audit.
  • Trim classification prompts to two messages per thread.
  • Add CSV download to the hosted report page.
  • Check every delivered report for broken links.

Deliverable: Cost and runtime per audit documented.

Day 20Proof

Publish a concrete finding

  • Post an anonymised case thread with a real dropped-thread count.
  • DM 30 new prospects.
  • Send agency trials a Stripe link.
  • Reconcile Stripe payments against delivered audits.

Deliverable: Case post live, reconciliation clean.

Day 21Decision

Midpoint pricing call

  • Check progress against 12 customers by day 21.
  • If behind, run a 72-hour $99 promo code.
  • If ahead, raise the agency pack to $549.
  • Rewrite the landing headline from the best-converting message.

Deliverable: Pricing decision made and headline updated.

Week 4 · Scale what works · 7 days
Day 22Repeat

Re-run campaign

  • Email every buyer the $79 re-run.
  • Send 50 messages to brands with public support addresses.
  • Deliver a sample to every agency reply.
  • Add a what-to-fix-first page to the PDF.

Deliverable: Re-run campaign sent, 50 messages out.

Day 23Social proof

Collect and publish testimonials

  • Ask three buyers for a public one-line quote.
  • Add the testimonials to the landing page.
  • Post a testimonial thread on X and LinkedIn.
  • DM 20 people who engaged with the posts.

Deliverable: Testimonial block live on the page.

Day 24Self-serve

Zero-touch purchase path

  • Ship pay-connect-PDF with no founder involvement.
  • Add an audit status page with a progress bar.
  • Fall back to export upload when OAuth fails.
  • Test the flow in incognito as a stranger.

Deliverable: Stranger can buy and receive a report unaided.

Day 25Outreach

Best opener, more volume

  • Send 60 messages with the best-performing opener.
  • Follow up every sample older than three days.
  • Test a $249 two-inbox bundle for multi-inbox brands.
  • Deliver all pending audits.

Deliverable: 60 messages sent, bundle offer tested.

Day 26Lead magnet

Free checklist to unlock non-repliers

  • Publish a free dropped-email checklist PDF.
  • Gate it behind an email form.
  • Send the checklist to the entire prospect list.
  • Follow with the $149 offer 24 hours later.

Deliverable: Lead magnet distributed to the full list.

Day 27Partners

Agency affiliate codes

  • Pitch 10 VA and support agencies a 30% affiliate rate.
  • Issue a Stripe promo code per agency.
  • DM agencies on X that manage client inboxes.
  • Track referral codes in Supabase.

Deliverable: Affiliate codes issued to 10 agencies.

Day 28Close

Last-call campaign

  • Offer the $149 audit plus one free December re-run.
  • Email every non-buyer who received a sample.
  • Send 50 last-call messages.
  • Deliver everything within 24 hours of payment.

Deliverable: Last-call campaign live across the full list.

Days 29-30 · Consolidate · 2 days
Day 29Fulfilment

Deliver everything, refund nothing late

  • Send the final 40 messages and follow-ups.
  • Deliver every pending audit before the deadline.
  • Check every Stripe payment has a delivered report.
  • Refund any order that cannot be delivered.

Deliverable: All orders fulfilled, no open tickets.

Day 30Reconcile

Verified number and public recap

  • Reconcile Stripe revenue, refunds and fees before 12:00 UTC.
  • Export the transaction list for TrustMRR verification.
  • Publish the honest 30-day numbers post.
  • Write the post-mortem: what converted, what to keep selling.

Deliverable: Verified profit figure and published recap.

Section 6 of 8

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: blended ARPU sits above the $149 list price because at least one $449 agency 5-pack is sold (Conservative assumes list price only).
  • Base: roughly 1 in 12 delivered samples converts to a paid audit; Upside assumes 1 in 6 (projections, not validated rates).
  • All scenarios: payment processing at about 4% blended on collected revenue, assumed Stripe (2.9% + $0.30 weighted toward the $149 ticket).
  • Conservative: outreach reply rate is low, single-digit sample requests, no paid marketing at all.
  • Upside: two sponsored community placements at $75 each drive the $150 ad spend line.
  • Refunds are assumed at zero in Conservative and Base because the free sample previews the exact artifact; Upside carries one $149 refund for a misclassified headline finding.
  • Chargebacks assumed zero in all scenarios; buyers pay by card via Stripe Checkout with a named deliverable.
  • Operating cost covers Vercel, Supabase, a domain, LLM inference and Resend; measured per audit on day 19 and assumed under $1 of inference per report.
  • No scenario sells below delivery cost; the cheapest offer, the $79 re-run, still clears roughly 20x its marginal inference cost.
  • Hackathon profit excludes hosting and inference; business profit deducts operating_costs_usd.

Conservative

3 customers
$447
collected revenue · avg payment $149
Refunds
$0
Chargebacks
$0
Payment fees
$18
Paid marketing
$0
Hackathon profit
$429
Operating costs
$45
Business profit
$384

Base case

12 customers
$1,920
collected revenue · avg payment $160
Refunds
$0
Chargebacks
$0
Payment fees
$77
Paid marketing
$0
Hackathon profit
$1,843
Operating costs
$120
Business profit
$1,723

Optimistic

40 customers
$6,600
collected revenue · avg payment $165
Refunds
$149
Chargebacks
$0
Payment fees
$264
Paid marketing
$150
Hackathon profit
$6,037
Operating costs
$240
Business profit
$5,797

Model your own numbers

Runs in your browser · nothing is saved
Collected revenue
$1,788
Payment fees (3.9%)
$70
Paid marketing
$0
Hackathon profit
$1,718
Business profit
$1,598

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. No customers, revenue or traction exist at the time of writing.

Section 7 of 8

Main risks and validation experiments

Risk 1Gmail restricted-scope verification blocks public OAuth onboarding and stalls launch.

Mitigation

Run the first 30 days on a test-user allowlist of up to 100 configured users plus a CSV/export upload fallback; start verification paperwork the same week but never depend on it.

Validation experiment

On day 3, connect a second real Google Workspace account that is not the founder's and confirm ingestion completes without a verification prompt.

Kill criteria

If fewer than half of interested buyers will either join the allowlist or upload an export, the OAuth dependency is fatal to self-serve and the concept must move to an upload-only product.

Risk 2Cold LinkedIn and X outreach produces too few sample requests to reach even three paying customers.

Mitigation

Track sample-request rate per 20 messages; rewrite the opener every 20 sends using the highest-performing subject line; add the operator-community post as a second organic surface.

Validation experiment

Send 20 messages on day 7 and count sample requests within 48 hours.

Kill criteria

If fewer than 2 of 20 messages produce a sample request after two opener rewrites, stop outbound and re-source buyers from agency affiliate referrals only.

Risk 3LLM misclassifies refund or quote threads as noise, producing a wrong headline revenue-at-risk number and triggering refunds.

Mitigation

Show category confidence on every row, keep a methodology note in the PDF, and hand-check the top 10 ranked threads before delivery during the first two weeks.

Validation experiment

Hand-label 30 unresolved threads from one real inbox and compare against model output before the first paid delivery.

Kill criteria

If agreement on the top 10 ranked threads falls below 70%, drop the dollar figure from the cover and ship the report as counts and ages only.

Risk 4Buyers treat a one-time audit as a favour, not a product, so there is no repeat revenue.

Mitigation

Ship Monthly Leak Watch on day 16 and pitch it to 100% of past buyers on day 17; offer the $79 re-run on day 14.

Validation experiment

Offer both the $79 re-run and the $49/month tier to the first five buyers and record which one they pick.

Kill criteria

If no buyer accepts either the re-run or the subscription by day 21, the repeat path does not exist and pricing must move to a $299 quarterly retainer instead.

Risk 5A support owner refuses to grant read-only inbox access to an unknown solo founder.

Mitigation

Lead with the free five-thread sample that needs no access at all, and offer the export-upload path in the first message.

Validation experiment

Add an export-upload line to half the outreach and compare sample-request rates against the OAuth-first version.

Kill criteria

If under 20% of sample recipients complete either connection method within 72 hours, the trust cycle is too long for a 30-day window.

Section 8 of 8

Hour-by-hour: the first 24 hours

Hours 1-2

  • Create the Next.js repo, Vercel project and Supabase schema for audits, threads and messages.
  • Publish a landing page with the $149 offer, the PDF sample image and a live Stripe Checkout link.

Output: A public URL that can accept a card payment today.

Hours 3-4

  • Build a 60-row prospect sheet: name, role, brand, LinkedIn or X URL, public support address.
  • Write the five-thread sample offer and the two follow-up messages.

Output: 60 named prospects and three ready-to-send messages.

Hours 5-7

  • Wire Google OAuth with gmail.readonly behind a test-user allowlist.
  • Fetch 90 days of messages from the founder's own Workspace inbox and confirm rows land in Supabase.

Output: Raw messages stored and queryable.

Hours 8-10

  • Send the first 10 LinkedIn messages with the free five-thread offer.
  • Log every send with a timestamp in the sheet so follow-ups are automatic on day 3.

Output: 10 conversations opened, follow-up queue seeded.

Hours 11-12

  • Write the day-5 definition of done.
  • Sketch the exact PDF layout: cover summary, ranked table with age and category, three fixes, CSV appendix.

Output: A one-page build spec the next four days follow.

Complete report

Full 30-day launch report

Read the full generated report

Inbox Leak Audit — 30-day launch report

1. Product opportunity and positioning

Shared inboxes hide their own failures. Gmail and every helpdesk show what is open now; none of them produce the artifact an owner actually needs after a bad week: a dated list of every customer email that never got a reply, how long it sat, and which of those were refund or quote requests. That gap is the product. Inbox Leak Audit attaches read-only to the inbox a team already runs, reconstructs 90 days of threads, and returns a branded PDF ranked by revenue at risk.

The buyer is deliberately narrow: the owner or head of CX at a 5-50 person e-commerce brand on Google Workspace in the US, UK, Canada or Australia, or the owner of a 2-20 person support/VA agency running client inboxes. The trigger is an event, not a mood: a missed email that became a refund, a chargeback or a public complaint, or a Monday backlog nobody can quantify. That buyer does not want a migration, a new workspace or a per-seat subscription. They want evidence this week that costs less than one lost order.

Against the alternatives: helpdesks such as Zendesk, Intercom, Front and Freshworks sell the inbox itself and require migration plus a subscription; analytics add-ons need clean data and weeks; a VA doing this by hand takes a day and produces a spreadsheet nobody forwards. This product produces a client-ready document in about ten minutes for $149, with no integration beyond a read-only OAuth consent.

The JEV score is 59.6, band C, and it is honest about the weak spots. Market demand is the lowest sub-score at 4.3, which is why the buyer is defined by trigger event rather than firmographics. Monetisation at 4.94 is capped by a one-time price, which is why the plan ships a $449 agency 5-pack and a $49 monthly re-scan. Operational simplicity at 4.74 reflects Gmail restricted-scope friction, which is why the first 30 days run on a test-user allowlist plus an export-upload fallback rather than waiting on a Google review.

Speed to first revenue is the strongest sub-score at 8.18, and the plan exploits it: the deliverable exists on day 5, samples go out on day 7, the first paid audit is targeted for day 10. Defensibility (6.04) is not code, it is the fact that one operator can hand-build samples faster than an incumbent can ship a report feature.

The output is a document, not a dashboard. A PDF with a dollar ranking survives a screenshot, forwards to a boss or a client in one click, and needs no onboarding call. The artifact is the marketing.

2. Smallest sellable MVP

One inbox, one payment, one PDF. Gmail read-only OAuth behind a test-user allowlist with a CSV upload fallback; 90-day ingestion and thread reconstruction that flags threads whose last message is inbound; LLM classification into order issue, refund, quote, complaint or noise with a confidence value; a revenue-at-risk score from category weight, thread age and recency; and PDF rendering gated behind a Stripe webhook, delivered by email with a 30-day signed link.

Not built in the first 30 days: a live inbox, reply drafting, SMS/social/voice ingestion, seats and permissions, CRM or Shopify integrations, an analytics dashboard, SOC 2 claims, and any human review step. Each one adds a support surface or delays the first sale.

3. Technical implementation plan

Next.js on Vercel, Supabase Postgres, Gmail API with the read-only scope, an LLM for classification, HTML-to-PDF rendering, Stripe Checkout and Resend. No queue, no worker fleet, no custom model. Audits cap at 500 threads; classification sends only the first and last message per thread, which keeps inference under about $0.40 per report.

The hardest risk is thread reconstruction, not the LLM. Gmail threadIds collapse subject variants, and a buyer who replied from a personal address leaves a thread that looks unanswered. Mitigation: search outbound messages within a +/- 24 hour window against the same recipient address before marking a thread unresolved, and print a confidence flag on ambiguous rows so the methodology is visible.

4. Pricing and monetization

$149 Single Inbox Audit, one-time: 90-day scan, branded PDF, CSV appendix, 30-day link. $449 Agency 5-Pack: five client inboxes, white-label PDF, per-client CSVs, 24-hour turnaround, about $89.80 per inbox. $49 per month Leak Watch: weekly re-scan, delta PDF, 48-hour unresolved alert. Past buyers get a $79 re-run.

Processing is assumed at roughly 4% blended on Stripe. Paid marketing is budgeted at zero in Conservative and Base; the Upside case spends $150 on two sponsored community placements. Delivery cost per audit is measured on day 19 and must stay under $1.

5. First-customer acquisition

One channel: direct LinkedIn and X outreach opened with a free five-thread sample. The offer is not a demo call, it is a delivered mini-PDF within 24 hours ending in a Stripe link.

60 accounts are listed on day 1, 150 by day 12. 20 messages go out on day 7, 40 on day 8, then 40-60 per wave. Two follow-ups: 72 hours with the sample's worst finding restated, day 7 with the agency bundle. One anonymised build-in-public post per week carries the same call to action.

Week-one goal: 60 messages, 6 samples delivered, 3 Stripe links out. The number to watch is sample-to-payment conversion; Base assumes roughly 1 in 12, Upside 1 in 6.

6. 30-day execution roadmap

Days 1-5 build: shell and list, ingestion, thread reconstruction, classification and scoring, then PDF plus payment gate, with a live end-to-end paid test on day 5. Day 6 hardens and times the run. Day 7 sends the first 20 messages and two samples.

Days 8-14 are outreach waves: 40 messages on day 8, samples on day 9, first paid audit targeted for day 10, the agency 5-pack shipped and pitched on day 13, funnel review on day 14. Day 11 is US Thanksgiving and is treated as a low-reply day used for product work rather than outreach volume.

Days 15-21 add Monthly Leak Watch on day 16, pitch it to every past buyer on day 17, seed three agency trials, cut inference cost on day 19, publish a real anonymised count on day 20, and make a pricing decision on day 21: $99 promo if behind 12 customers, $549 agency pack if ahead.

Days 22-28 run the $79 re-run campaign, collect three testimonials, ship a zero-touch purchase path, distribute a free dropped-email checklist, and issue affiliate codes to ten agencies at 30%. Days 29-30 are final-call messages, complete fulfilment, Stripe reconciliation and an honest public recap. Long arcs across all 30 days: the prospect list grows every wave, and every delivered audit produces one more piece of shareable evidence.

7. Revenue and profit scenarios

All figures are projections from stated assumptions, not verified revenue, and no customers exist at the time of writing.

Conservative: 3 customers at $149 blended ARPU, $447 gross, zero refunds, zero chargebacks, about $18 in processing fees, zero ad spend, $45 operating cost.

Base: 12 customers at $160 blended ARPU, one agency 5-pack pulling the average up, $1,920 gross, $77 processing fees, zero ad spend, $120 operating cost.

Upside: 40 customers at $165 blended ARPU, $6,600 gross, one $149 refund, $264 processing fees, $150 ad spend, $240 operating cost.

The contest ranks hackathon profit, which deducts refunds, chargebacks, processing fees and paid marketing but not hosting or inference. Business profit also deducts operating cost. No scenario prices a sale below its delivery cost; even the $79 re-run clears roughly twenty times its marginal inference cost.

8. Main risks and validation experiments

Gmail restricted-scope verification is the top operational risk. The allowlist and export fallback exist so launch never depends on a Google review. Test on day 3 with a second real Workspace account.

Outreach reply rate is the demand risk. Send 20 messages on day 7 and count sample requests; fewer than two of twenty after two opener rewrites means the message is wrong, and the fallback is agency affiliate referrals.

Misclassification can produce a wrong headline number and a refund. Hand-label 30 threads and compare before the first paid delivery; if agreement on the top ten rows drops below 70%, remove the dollar figure and ship counts and ages.

Trust risk: a support owner may refuse inbox access. Measure how many sample recipients choose the export path versus OAuth within 72 hours; below 20% completion means the trust cycle is too long for 30 days.

Repeatability risk: a one-time audit may be treated as a favour. Offer the $79 re-run and the $49 tier to the first five buyers; if neither is accepted by day 21, switch the offer to a $299 quarterly retainer.

9. First 24-hour action plan

Hours 1-2: repo, Vercel project, Supabase schema, and a landing page with a live Stripe Checkout link. Output: a URL that can take a card today.

Hours 3-4: 60-row prospect sheet with name, role, brand, contact URL and public support address, plus the sample offer and two follow-ups written. Output: a list, not an intention.

Hours 5-7: Google OAuth with the read-only scope against the founder's own Workspace inbox, 90 days of messages stored in Supabase. Output: real data in the database.

Hours 8-10: first 10 LinkedIn messages sent, each logged with a timestamp so the day-3 follow-up is automatic. Output: 10 open conversations.

Hours 11-12: a one-page build spec for day 5 and the exact PDF layout — cover summary, ranked table with age and category, three fixes, CSV appendix. Output: the document the next four days follow.

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 →