Score B · 6.90/10Free to read$100K Hackathon · November 16, 2026 → December 16, 2026Nov 15, 2026

Skeleton Sheet

Skeleton Sheet turns three competitors' best-performing Facebook and Instagram posts into a white-labeled 30-day content pack — 30 caption drafts, the structural blueprint behind each one, and a posting calendar — for freelance social media managers, delivered in under 12 hours.

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.

Promising6.90/10
MVP build (solo)
5 days
Launch price
$149
Sales to $1,000 revenue
7 customers
First paying customer
Direct DMs to freelance social media managers sourced from two named Facebook communities for social media managers and LinkedIn search for 'social media manager' plus 'freelance'.
Who pays

Solo or two-person social media management freelancers in the US, UK, Canada and Australia managing 4–12 local-business clients (dentists, med spas, gyms, restaurants, home services) who just signed a client or are entering a monthly retainer cycle and must deliver a month of posts this week.

Biggest execution risk

Operational simplicity 5.3/10

Why this could work in 30 days: one founder can ship the MVP in 5 days; the plan targets a first sale on day 10; 7 sales at $149 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 9.0/10.

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

14 concrete actions. Tick them off as you go — nothing is saved or sent.

0/14
Entry price
$149
Solo build
5 days
30d base revenue
$2,980
30d hackathon profit
$2,727
Section 1 of 8

Product opportunity and positioning

Target customer

Solo or two-person social media management freelancers in the US, UK, Canada and Australia managing 4–12 local-business clients (dentists, med spas, gyms, restaurants, home services) who just signed a client or are entering a monthly retainer cycle and must deliver a month of posts this week.

Why now

Freelancers already pay out of pocket per client, LLM and managed-scraping costs have collapsed to under $2 per pack, and the artifact is exactly the deliverable they are contractually on the hook for — so no integration, ad-account access or platform API approval stands between the founder and a first payment.

Problem

Every new client and every monthly cycle the freelancer must produce 20–30 platform-native posts, which today means 2–4 hours of scrolling competitor pages, screenshotting top posts and guessing at structure, followed by revision loops when the client says the captions sound generic.

Solution

The buyer submits 2–3 competitor page URLs plus a six-field client brief; a managed collector pulls the top public posts by engagement ratio, an LLM chain labels each post's skeleton (hook type, format, emotional lever, CTA pattern, length band), regenerates those skeletons in the client's voice, and renders a white-labeled PDF plus CSV calendar of 30 posts with hashtag sets and image briefs, emailed within 12 hours.

Differentiation · Generic AI writers produce plausible captions with no evidence they work in that specific local niche; Skeleton Sheet derives every caption from a structural skeleton observed in real competitor posts for that exact niche, and returns a client-ready pack instead of a prompt output or an analytics dashboard.

Skeleton Sheet — Opportunity and Positioning

The buyer and the moment of pain

Skeleton Sheet sells to one person: a freelance social media manager with four to twelve local-business clients on retainer. The buying moment is not abstract — it is the week a new dentist, med spa, gym, restaurant or home-services client signs, or the week the monthly retainer cycle restarts and a month of posts is due. At that moment the freelancer has a contractual obligation and no content. Their current method is to scroll three competitor pages, screenshot what looked like it worked, and guess at why, then draft 20–30 captions by hand. That is two to four hours per client of unpaid research, followed by a revision loop when the client reads the drafts and says they sound generic and off-brand.

This is a cost-of-goods decision, not a new budget line. The freelancer is already being paid for the month of posts; $149 spent to produce it in twelve hours instead of two days is arithmetic, not a strategy debate. There is no procurement, no team seat, no ad-account access, no OAuth grant and no platform API approval between the pitch and the payment.

What is actually being sold

One artifact: a white-labeled 30-post pack. That means thirty caption drafts in the client's voice, a structural blueprint page labelling each post's hook type, format, emotional lever, CTA pattern and length band, a hashtag set per post, a one-line image brief per post, and a CSV calendar mapped to real dates in the month ahead. Delivered as a PDF and CSV link in the buyer's inbox within twelve hours, no account required.

Why this is not PostRocket again

The original PostRocket sold measurement — when to post, how a page performed — and died the moment Facebook shipped a free native dashboard that did the same thing more reliably. Rebuilding a dashboard would be re-entering a race that has already been lost, against Meta, Buffer and Sprout Social simultaneously. Skeleton Sheet deliberately refuses measurement. It sells the finished deliverable: content the freelancer can ship to a client the same day. Native insights tooling cannot produce that, and neither can a generic AI writer, because the value sits in the evidence chain — the skeletons come from posts that demonstrably earned engagement for that exact niche.

Honest defensibility

The moat is thin and should be described as thin: niche-specific evidence plus turnaround plus a white-label format a freelancer can hand to a client with their own logo. A competitor could copy the surface. What they cannot instantly copy is a library of per-niche skeleton patterns and a twelve-hour delivery promise backed by a pipeline that costs under $2 a pack to run. The first thirty days are not about building a moat; they are about proving that forty freelancers will pay $149 for an artifact they already owe someone else.

Positioning the founder should repeat verbatim

"Your three closest competitors already proved what works in your client's niche. I turn that into thirty posts, the blueprint behind each one, and a calendar — in twelve hours, white-labeled, for $149."

Section 2 of 8

Smallest sellable MVP

A paid intake form (competitor URLs + six-field brief) that triggers a pipeline producing a white-labeled 30-post PDF and CSV calendar, emailed within 12 hours.

Must-have features

  • URL intake form accepting 2–3 competitor pages plus name, offer, tone, audience, banned topics and call-to-action
  • Managed scraper returning the top 15 public posts per page ranked by engagement ratio
  • Skeleton extractor labelling each post with hook type, format, emotional lever, CTA pattern and length band
  • Regeneration engine rewriting each skeleton into the client's offer and voice for 30 captions
  • Hashtag set and one-line image brief per post
  • PDF renderer with a blueprint appendix plus a CSV calendar mapped to the month
  • Stripe Checkout with webhook-triggered generation and Resend delivery email
  • Admin page to inspect, re-run and refund a pack
  • Paste-in post fallback for competitor pages the collector cannot fetch

Explicitly out of scope

  • Social platform OAuth, auto-posting or scheduling
  • Analytics dashboard, historical tracking or benchmarking
  • Team seats, agency multi-client workspace or client approval portal
  • Image generation — art direction is text only
  • Mobile app or white-label client portal
  • Subscription billing
  • Calls, onboarding sessions or done-for-you services

Demo flow (what the buyer sees)

  1. Prospect opens the intake page and enters three competitor page URLs.
  2. Prospect fills six brief fields: client name, offer, tone, audience, banned topics, CTA.
  3. Stripe Checkout collects $149 and the webhook writes a job row.
  4. Collector pulls the top 15 public posts per page by engagement ratio and caches them.
  5. Extractor returns a labelled JSON skeleton per post.
  6. Regeneration produces 30 captions, hashtags and image briefs in the client's voice.
  7. Renderer builds the white-labeled PDF with blueprint appendix and a CSV calendar.
  8. Resend emails the download link, target under 12 hours from payment.
  9. Founder reviews the job in admin and re-runs or refunds if quality fails.
Build plan

Technical implementation plan

Stack

Next.js on Vercel for intake, admin and the PDF routeSupabase for job records, cached scrape payloads and pack storageApify or ScrapingBee managed scraper for public post dataOpenAI or Anthropic API for skeleton extraction and caption regenerationStripe Checkout plus webhook for payment and fulfilmentResend for delivery emailReact-PDF for pack renderingCloudflare Turnstile on the intake form to stop bot submissions

Architecture

A single Next.js app on Vercel holds the intake form, an admin page and a PDF render route. Payment or an admin click creates a row in Supabase. A background route calls the managed scraper for each competitor URL, caches the raw posts so a page is never fetched twice for the same client, and normalises them into a top-15-by-engagement-ratio list. Two chained LLM calls follow: one extracts labelled skeletons as strict JSON, one regenerates 30 captions against the six brief fields, with hashtags and image briefs generated in the same pass. The renderer builds the PDF and CSV, stores both in Supabase Storage and Resend emails the signed download link. Failed jobs retry twice, then surface in admin with a one-click re-run using the cached skeletons so no re-scraping is needed.

Build steps

  1. Scaffold Next.js app on Vercel and create Supabase tables for jobs, posts and packs.
  2. Build the intake form with six brief fields and 2–3 competitor URL inputs.
  3. Integrate the managed scraper, cache responses, rank posts by engagement ratio.
  4. Ship the paste-in fallback for pages that fail collection.
  5. Write the skeleton extractor prompt chain and validate JSON output on 20 posts.
  6. Write the regeneration chain producing 30 captions, hashtags and image briefs.
  7. Generate the CSV calendar mapped to real dates.
  8. Render the white-labeled PDF with a blueprint appendix.
  9. Wire Stripe Checkout and the fulfilment webhook.
  10. Add Resend delivery email with the download link.
  11. Build the admin page for inspection, re-run and refund.
  12. Run one end-to-end paid test transaction and log cost per pack.

AI components

  • Skeleton extractor: one LLM call per collected post returning hook type, format, emotional lever, CTA pattern and length band as strict JSON
  • Regeneration engine: one LLM call per post writing the client-voice caption, hashtag set and image brief
  • Quality gate: a lightweight pass that flags captions breaching banned topics or falling outside the length band for regeneration
Hardest technical risk: The managed scraper: local-business Facebook and Instagram pages vary in structure, rate-limit unpredictably and can break without notice, which is why the paste-in fallback and the scrape cache ship in the first two days rather than later.
Section 3 of 8

Pricing and monetization

Single Client Pack

Entry offer
$149 one-time

For Freelance social media manager delivering one client's month of posts

  • 30 caption drafts in the client's voice
  • Structural blueprint page per post
  • CSV content calendar mapped to the month
  • One-line image brief per post
  • Hashtag set per post
  • White-labeled PDF plus CSV download
  • 12-hour turnaround

Three-Client Bundle

$349 one-time

For Freelancer onboarding two or three new clients at once

  • Three client packs, 90 posts total
  • Priority 12-hour turnaround
  • One free regeneration per pack
  • White-labeled PDF plus CSV for each client

Model: One-time per-client packs, no subscription inside the first 30 days.

First revenue offer: Build a free 30-post pack for one of the prospect's existing clients from public competitor pages, send it unsolicited with the blueprint page attached, and price the next one at $149 with 12-hour delivery.

Expected ACV: $149

Upsell path: Bundle the next two clients at $349, then sell a paid monthly refresh at $149 per client per cycle, with every buyer asked for one intro to another freelancer.

Target customers
40
Revenue model
$5,960
Costs
$80
Profit model
$5,880
Section 4 of 8

First-customer acquisition

Primary channel

Direct DMs to freelance social media managers sourced from two named Facebook communities for social media managers and LinkedIn search for 'social media manager' plus 'freelance'.

Where to find them

  • Two Facebook communities for freelance social media managers — members posting about a new client signed or a content calendar due
  • LinkedIn search for 'social media manager' filtered to freelance and solo, US/UK/CA/AU
  • LinkedIn posts in the last 30 days announcing a new client win
  • Comment threads under content-calendar and 'client wants more engagement' posts in those same two groups
  • Buyers' own client lists published on their portfolio sites (used only to pick a sample target)

Exact outreach motions

  • Send 15 DMs per day with a free sample pack built for one real client of that prospect.
  • Build three free sample packs per day for the warmest replies, capped at 20 minutes each.
  • Follow up once, two days later, on every non-replier with a different sample.
  • Post the blueprint page and a before/after inside both groups twice a week with a direct checkout link.
  • Ask every buyer for one intro to another freelancer inside the delivery email.

Message template

Hey [Name] — I reverse-engineered [Client]'s three closest competitors' best-performing posts into a 30-day pack: 30 captions in [Client]'s voice, the structural blueprint behind each, hashtags, image briefs and a calendar. It's attached, yours free, no catch. If it's useful I'll do your next client for $149 in 12 hours. Want the other two done?
Week-one goal: 40 DMs sent, 3 free sample packs delivered, 10 replies, 1 paid $149 pack collected by day 10.
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 1Build

Foundation

  • Scaffold the Next.js app on Vercel with the six-field intake form.
  • Create Supabase tables for jobs, cached posts and packs.
  • Open Stripe and connect the TrustMRR payout account.

Deliverable: Deployed intake form with an empty job pipeline.

Day 2Build

Collector

  • Wire the managed scraper to fetch the top 15 public posts per page.
  • Test on three real local-business competitor pages.
  • Ship the paste-in fallback for pages that fail.

Deliverable: Ranked post collector plus a working fallback path.

Day 3Build

Skeleton extractor

  • Write the extractor chain for hook, format, lever, CTA and length band.
  • Run it on 20 collected posts and tighten the JSON schema.
  • Persist skeletons per job.

Deliverable: Labelled skeletons for 20 real posts.

Day 4Build

Regeneration and calendar

  • Generate 30 captions from skeletons against the six brief fields.
  • Emit hashtag sets and one-line image briefs per post.
  • Map posts to real dates and export the CSV calendar.

Deliverable: Full 30-post text pack and CSV calendar.

Day 5BuildMVP shipped

Monetise

  • Render the white-labeled PDF with a blueprint appendix.
  • Wire Stripe webhook to generation and Resend delivery.
  • Build admin for inspect, re-run and refund.
  • Complete one paid end-to-end test transaction.

Deliverable: Live, chargeable product at $149.

Day 6Sell

First samples

  • Build a free pack for one existing client of each of three prospects.
  • QA every caption for niche fit and brand tone.
  • Fix the three worst output defects found.

Deliverable: Three free sample packs ready to send.

Day 7Sell

Prospect list

  • Build a 40-name list from the two Facebook groups and LinkedIn search.
  • Filter to freelancers who posted about a new client or calendar in 30 days.
  • Finalise the DM copy and sample attachment.

Deliverable: 40 qualified prospects and final DM copy.

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

Outreach wave one

  • Send 15 DMs with the free sample attached.
  • Post the sample pack in group one.
  • Log every reply in a tracking sheet.

Deliverable: 15 DMs sent and one group post live.

Day 9Sell

Outreach wave two

  • Send 15 more DMs.
  • Build free packs for the three warmest replies.
  • Follow up once with day-8 non-repliers.

Deliverable: 30 DMs total and three new samples delivered.

Day 10SellFirst sale target

First sale

  • Send 10 more DMs.
  • Offer the $149 pack to every warm reply with 12-hour turnaround.
  • Handle objections in-thread, no calls.

Deliverable: First paid $149 pack collected.

Day 11Deliver

Fulfilment

  • Run each paid job through the pipeline and inspect output.
  • Deliver PDF plus CSV within 12 hours of payment.
  • Log each defect with its dollar impact.

Deliverable: All paid packs delivered on time.

Day 12Sell

Referral ask

  • Ask every buyer for one intro to another freelancer.
  • Add the referral line to the delivery email.
  • Ship the remaining 10 DMs.

Deliverable: Referral requests sent to every buyer.

Day 13Market

Group proof

  • Publish the blueprint page and a before/after in both groups.
  • Answer every comment the same hour.
  • Link directly to the $149 checkout.

Deliverable: Two group posts live with a checkout link.

Day 14Review

Funnel math

  • Count DMs, replies, samples, paid packs and refunds.
  • Identify the worst-converting step.
  • Rewrite the DM or the sample for that step.

Deliverable: Funnel table and one queued fix.

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

Volume

  • Send 20 DMs using the rewritten message.
  • Deliver three free samples to warm replies.
  • Ship all paid packs same day.

Deliverable: 20 DMs, three samples, zero late deliveries.

Day 16Build

Output quality

  • Add a banned-topics filter and tone check to regeneration.
  • Auto-regenerate any caption outside its length band.
  • Re-test on 10 posts.

Deliverable: Lower defect rate on the next ten packs.

Day 17Sell

Volume

  • Send 20 DMs.
  • Follow up with every day-15 non-replier.
  • Repost the sample in group two.

Deliverable: 40 touches and one group post.

Day 18Monetise

Bundle

  • Add the $349 Three-Client Bundle to checkout.
  • Offer the bundle to any buyer with two or more clients.
  • Update the delivery email with the bundle link.

Deliverable: Bundle live and offered to all buyers.

Day 19Sell

Proof

  • Ask two buyers for a short quote with written permission.
  • Add one approved quote to the landing page.
  • Send 20 DMs.

Deliverable: Landing page with one approved quote.

Day 20Deliver

Operations

  • Write the 15-minute manual runbook for one pack.
  • Cache every scrape so no page is fetched twice per client.
  • Set a $2 per-pack cost alarm.

Deliverable: Runbook plus cached scrape layer.

Day 21Review

Pipeline health

  • Measure scraper failure rate per niche.
  • Measure average generation time and API spend per pack.
  • Fix the top failure mode.

Deliverable: Failure rate and cost per pack recorded.

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

Peak outreach

  • Send 25 DMs.
  • Build three free samples for the highest-fit replies.
  • Chase every open quote.

Deliverable: 25 DMs, three samples, all quotes chased.

Day 23Sell

Referral loop

  • Re-offer the bundle and referral ask to every past buyer.
  • Reward any converting intro with a free regeneration.
  • Send 20 new DMs.

Deliverable: Referral offers sent and 20 new DMs.

Day 24Market

Case artifact

  • Build one anonymised before/after sample page.
  • Post it in both groups.
  • Pin it on the founder's profile.

Deliverable: Anonymised case page live.

Day 25Build

Reliability

  • Add one-click admin re-run using stored skeletons.
  • Add a Stripe dispute webhook that flags the job.
  • Test both paths end to end.

Deliverable: Re-run and dispute flagging working.

Day 26Sell

Volume

  • Send 25 DMs.
  • Follow up with day-22 non-repliers.
  • Close every open bundle quote.

Deliverable: 25 DMs sent and follow-ups completed.

Day 27Sell

Late-month push

  • Lead every message with 12-hour turnaround.
  • Send 20 DMs.
  • Deliver every outstanding pack.

Deliverable: No backlog and 20 DMs sent.

Day 28Review

Economics

  • Recompute revenue, refunds, fees and API spend.
  • Set the day-29 and day-30 target number.
  • Cut any task with no revenue path.

Deliverable: Final two-day plan with a target number.

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

Last call

  • DM every non-replier once with a final free-sample offer.
  • Post the last group post of the window.
  • Confirm receipt on every delivered pack.

Deliverable: All lists worked to the end.

Day 30Close

Reconcile

  • Match every sale against the TrustMRR-connected Stripe account.
  • Issue any outstanding regeneration or refund.
  • Write the post-mortem with conversion rates per step.

Deliverable: Reconciled ledger and written post-mortem.

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

  • Conservative and Base: a single $149 pack is the only thing sold; bundle revenue is excluded so any bundle sale is upside.
  • Upside: blended ARPU rises to $155 because roughly one in four buyers takes the $349 three-client bundle.
  • Payment fees assume Stripe at 2.9% + $0.30 plus a small international card mix, landing near 3.5% of collected revenue (3.9% in Upside where bundle volume raises average order size).
  • Refunds: Conservative assumes one refunded pack, Base one, Upside about 5% of gross.
  • Chargebacks appear only in Upside, at roughly 2.5% of gross, because digital goods disputes usually land after delivery and outside the first 30 days.
  • Ad spend is $0 in Conservative and Base; Upside assumes $300 of paid spend on sponsored newsletter or community placements, not ads at scale.
  • Operating costs cover Vercel, Supabase, the managed scraper, LLM inference and email: about $2 per pack delivered plus roughly $30 of monthly fixed tooling.
  • All figures are projections from stated assumptions, not verified revenue, customers or traction.
  • Every sale must cover its delivery cost: $149 against roughly $2.40 of inference and scraping is comfortably above cost.

Conservative

6 customers
$894
collected revenue · avg payment $149
Refunds
$149
Chargebacks
$0
Payment fees
$31
Paid marketing
$0
Hackathon profit
$714
Operating costs
$45
Business profit
$669

Base case

20 customers
$2,980
collected revenue · avg payment $149
Refunds
$149
Chargebacks
$0
Payment fees
$104
Paid marketing
$0
Hackathon profit
$2,727
Operating costs
$90
Business profit
$2,637

Optimistic

40 customers
$6,200
collected revenue · avg payment $155
Refunds
$310
Chargebacks
$155
Payment fees
$248
Paid marketing
$300
Hackathon profit
$5,187
Operating costs
$200
Business profit
$4,987

Model your own numbers

Runs in your browser · nothing is saved
Collected revenue
$2,980
Payment fees (3.9%)
$116
Paid marketing
$0
Hackathon profit
$2,864
Business profit
$2,774

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, sales or traction exist yet; every number here is a projection.

Section 7 of 8

Main risks and validation experiments

Risk 1Public Facebook and Instagram post collection breaks or is blocked, removing the evidence base.

Mitigation

Paste-in post fallback ships on day 2 and the scrape cache prevents repeat fetches.

Validation experiment

Run the collector against 20 competitor pages across five niches on day 2 and record the success rate per niche.

Kill criteria

If fewer than 60% of target pages collect successfully across five niches and paste-in takes more than 10 minutes per prospect, reposition to 'you send the posts, we return the pack' or stop.

Risk 2Buyers judge the AI captions as generic and request refunds inside the same groups where they were acquired.

Mitigation

Blueprint appendix makes the evidence visible per post, plus one free regeneration on every pack.

Validation experiment

Send five free sample packs on days 6–7 and ask each recipient to rate 'would you ship this to a client unedited' 1–5.

Kill criteria

If average ship-ready rating is below 3 out of 5 across five samples, change the regeneration prompt chain before charging anyone.

Risk 3Cold DM conversion falls far below the plan and 40 DMs per day produces no paid packs.

Mitigation

Daily funnel logging from day 8 so the worst step is rewritten within 48 hours.

Validation experiment

Send 40 DMs over days 8–9 and measure reply rate and sample-request rate separately.

Kill criteria

If 100 DMs produce fewer than 8 replies and zero paid packs by day 14, switch the free-sample target from 'their existing client' to 'their own personal brand' or stop.

Risk 4Freelancers decide to build their own prompt stack instead of paying $149.

Mitigation

Sell the finished white-labeled artifact and the 12-hour promise, not the prompt.

Validation experiment

Ask every non-buyer in the thread one question: what would make this an instant yes.

Kill criteria

If more than half of non-buyers cite 'I can do this myself' rather than price or trust, raise the deliverable to full calendar plus image briefs or stop.

Risk 5Chargebacks and disputes arrive after delivery and are not controllable.

Mitigation

Clear no-refund-after-delivery terms on checkout, delivery confirmation email, and a Stripe dispute webhook flagging the job.

Validation experiment

Track refund and dispute rate per ten packs sold across the 30 days.

Kill criteria

If combined refunds and chargebacks exceed 15% of gross revenue, pause paid outreach and fix quality before selling again.

Section 8 of 8

Hour-by-hour: the first 24 hours

Hours 1-2

  • Create the Vercel and Supabase projects.
  • Scaffold the Next.js app with the intake route.

Output: Deployed skeleton app with a reachable URL.

Hours 3-5

  • Build the intake form with 2–3 competitor URL fields and the six brief fields.
  • Add Cloudflare Turnstile.

Output: Working intake form that writes a job row to Supabase.

Hours 6-9

  • Wire the managed scraper and fetch the top 15 posts per page by engagement ratio.
  • Cache raw payloads in Supabase.

Output: Collector returning ranked real posts for three test pages.

Hours 10-13

  • Write and test the skeleton extractor chain on ten collected posts.
  • Lock the JSON label schema.

Output: Ten labelled skeletons with a stable schema.

Hours 14-17

  • Write the regeneration chain for 30 captions, hashtags and image briefs.
  • Render a first draft PDF with the blueprint appendix.

Output: One complete 30-post pack PDF from a real local-business niche.

Hours 18-21

  • Create Stripe Checkout for the $149 Single Client Pack and connect the TrustMRR payout account.
  • Add the fulfilment webhook and Resend delivery email.

Output: A chargeable link that delivers a pack by email.

Hours 22-24

  • Shortlist five prospects and pick one existing client of the best-fit prospect.
  • Generate the first free sample pack.

Output: One free sample pack ready to send on day 2.

Complete report

Full 30-day launch report

Read the full generated report

Skeleton Sheet — 30-Day Launch Report

1. Product opportunity and positioning

Skeleton Sheet sells one artifact: a white-labeled 30-post content pack built from what already earns engagement for a client's three closest competitors. The buyer is a freelance social media manager running four to twelve local-business clients — dentists, med spas, gyms, restaurants, home services — in the US, UK, Canada or Australia. They are on retainer. Every new client and every monthly cycle hits the same wall: 20–30 platform-native posts due this week, two to four hours of scrolling and screenshotting and guessing per client, then a revision loop because the drafts sound generic.

This is not an analytics product. PostRocket's original trap was selling measurement that a free native dashboard commoditised; measurement is a lost race against Meta, Buffer and Sprout Social. Skeleton Sheet skips measurement entirely and ships the finished deliverable: 30 captions in the client's voice, a labelled structural blueprint per post (hook type, format, emotional lever, CTA pattern, length band), hashtag sets, a one-line image brief, and a CSV calendar mapped to the month.

The wedge holds because $149 is a cost of goods paid out of an existing client retainer, not a new budget line. No OAuth, no ad-account access, no platform API approval, no procurement, no team seat. The buyer approves instantly because the output is the thing they already owe a client this week. Defensibility is thin and should be described as thin: niche-specific evidence plus a twelve-hour turnaround plus a white-label format. The first month is not about a moat; it is about proving forty freelancers will pay for an artifact they already owe someone else.

2. Smallest sellable MVP

Eight pieces and nothing more: an intake form taking 2–3 competitor URLs plus six brief fields; a managed collector returning the top 15 public posts per page ranked by engagement ratio; a skeleton extractor labelling hook type, format, emotional lever, CTA pattern and length band; a regeneration engine rewriting each skeleton into the client's voice for 30 captions; hashtag sets and one-line image briefs; a renderer producing a white-labeled PDF with blueprint appendix and a CSV calendar; Stripe Checkout with a webhook that triggers generation and emails the link; and a private admin page to inspect, re-run and refund.

Explicitly out of scope: auto-posting, scheduling, analytics, team seats, client approval portals, image generation, mobile app, subscriptions. Each of those adds a week and zero first-month revenue. The paste-in fallback for failed collections is built on day 2, not day 20, so the pipeline stays sellable even if the collector degrades.

3. Technical implementation plan

Next.js on Vercel serves the intake form, admin page and PDF route; Supabase holds jobs, cached posts and pack files; Apify or ScrapingBee handles collection; OpenAI or Anthropic runs two chained calls (extract skeletons, regenerate captions); Stripe Checkout plus webhook handles payment and fulfilment; Resend delivers; React-PDF renders.

The flow is linear: payment or admin click writes a job row, the collector fetches and caches top posts per competitor, the extractor returns strict-JSON skeletons, regeneration produces 30 captions with hashtags and image briefs, the renderer builds PDF and CSV, files land in storage, Resend emails a signed link. Failures retry twice, then surface in admin with one-click re-run using cached skeletons so nothing is re-scraped. The hardest risk is the collector — page structures vary and rate limits appear without warning — which is why the fallback and the cache ship in the first 48 hours. Target delivery cost is under $2 of inference plus roughly $0.40 of scraping per pack.

4. Pricing and monetization

$149 for a Single Client Pack, one-time. $349 for a Three-Client Bundle with priority turnaround and one free regeneration per pack. No subscription in the first 30 days: freelancer demand is lumpy and a one-time charge removes the cancel-before-value objection.

The first-revenue offer is a free 30-post pack built for one of the prospect's existing clients, sent unsolicited with the blueprint page attached, followed by "$149 for your next client, delivered in 12 hours." The free pack costs under $2 to make and is the entire sales pitch. Upsell runs bundle, then a paid monthly refresh at $149 per client per cycle.

5. First-customer acquisition

One channel: direct DMs to freelance social media managers found in two named Facebook communities for social media managers and via LinkedIn search for "social media manager" plus "freelance", filtered to people who posted about a new client or a content calendar in the last 30 days. Target 40 prospects.

Daily loop: 15 DMs each morning, three free sample packs each afternoon for the warmest replies, one follow-up two days later. The group post — blueprint page plus before/after — lands in the same two communities twice a week with a direct checkout link. Week-one goal: 40 DMs, 3 samples delivered, 10 replies, 1 paid pack by day 10.

6. 30-day execution roadmap

Days 1–5 are build only: scaffold, collector with fallback, extractor, regeneration plus CSV, then PDF, Stripe, Resend and admin, ending with one paid end-to-end test. Nothing is sold before day 5.

Days 6–7 produce three free sample packs and a 40-name prospect list. Days 8–14 are the sales sprint: 15 DMs daily, first paid pack targeted for day 10, same-day fulfilment, referral asks on day 12, two group posts on day 13, and a funnel review on day 14 that names the worst-converting step and rewrites it.

Days 15–21 raise volume to 20 DMs a day, fix output quality with a banned-topics filter and length-band checks, launch the $349 bundle on day 18, collect one approved quote, and record scraper failure rate and cost per pack.

Days 22–28 push to 25 DMs a day, re-open every past buyer for bundle and referral, publish one anonymised before/after case page, and add one-click admin re-run plus dispute flagging. Days 29–30 are for last-call DMs, final deliveries and reconciliation against the TrustMRR-connected Stripe account. No new features after day 25.

7. Revenue and profit scenarios

These are projections built from stated assumptions, not verified revenue or traction.

Conservative — 6 customers at $149 is $894 gross, one refund, about $31 in processing fees, no ad spend. Hackathon profit lands near $714, and roughly $45 of hosting and inference leaves business profit near $669.

Base — 20 customers at $149 is $2,980 gross, one refund, about $104 in fees, no ad spend. Hackathon profit lands near $2,727 and business profit near $2,637 with about $90 of operating cost.

Upside — 40 customers at a $155 blended ARPU (bundle mix) is $6,200 gross, with $310 refunds, $155 chargebacks, about $248 in fees and $300 of paid placements. Hackathon profit lands near $5,187 and business profit near $4,987 after $200 of operating cost.

Hackathon profit is gross minus refunds, chargebacks, payment fees and paid marketing. Business profit subtracts operating costs as well. Hosting, tooling and inference are excluded from the contest figure but included in the business figure.

8. Main risks and validation experiments

Collector failure is the top risk. Test 20 pages across five niches on day 2; if fewer than 60% collect successfully and paste-in takes over ten minutes, reposition to a paste-the-posts product or stop.

Output quality is second. Send five free packs on days 6–7 and ask each recipient to rate "would you ship this unedited" from 1 to 5. Below 3 average, rewrite the regeneration chain before charging anyone.

Channel thinness is third. If 100 DMs produce fewer than 8 replies and no paid pack by day 14, change the free-sample target from their client to their own personal brand, or stop.

Self-build substitution is fourth: if more than half of non-buyers say "I could prompt this myself", raise the deliverable rather than cutting price. Refunds and chargebacks are fifth; above 15% of gross, pause outreach and fix quality.

9. First 24-hour action plan

Hours 1–2: create Vercel and Supabase projects and scaffold the app. Hours 3–5: intake form with six brief fields and Turnstile. Hours 6–9: collector on three test pages with caching. Hours 10–13: skeleton extractor validated on ten posts. Hours 14–17: regeneration chain and first draft PDF from a real niche. Hours 18–21: Stripe Checkout, TrustMRR-connected payout account, webhook and Resend email. Hours 22–24: shortlist five prospects and generate the first free sample pack. Day 2 starts with that sample in an inbox.

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 →