Score C · 6.24/10Free to read$100K Hackathon · November 16, 2026 → December 16, 2026Nov 16, 2026

LotSheet Resin Reports

LotSheet turns a recycler's raw lab results and a buyer's spec sheet into a branded, pass/fail resin lot datasheet PDF and a ready-to-send reply email in minutes.

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.

Speculative6.24/10
MVP build (solo)
5 days
Launch price
$199
Sales to $1,000 revenue
6 customers
First paying customer
Cold direct outreach to named quality and sales managers at small plastics recyclers and compounders, offering a free 24-hour datasheet conversion.
Who pays

Quality manager or sales manager at a small plastics recycler or compounder (5-200 staff) selling regrind, flake or recycled resin pellets, English-speaking US, UK and EU first.

Biggest execution risk

Market demand 4.4/10

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 11; 6 sales at $199 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 8.4/10.

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

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

0/15
Entry price
$199
Solo build
5 days
30d base revenue
$1,490
30d hackathon profit
$1,442
Section 1 of 8

Product opportunity and positioning

Target customer

Quality manager or sales manager at a small plastics recycler or compounder (5-200 staff) selling regrind, flake or recycled resin pellets, English-speaking US, UK and EU first.

Why now

The contest window rewards a deliverable a buyer has already asked for; the datasheet is that deliverable, and it can be produced and paid for with a card inside one week.

Problem

Every buyer sends a different spec sheet, and someone retypes the lot's lab values into a Word template by hand while checking each against the buyer's limits, burning hours and letting borderline or off-spec values slip through.

Solution

Upload the lot's lab results and the buyer's spec sheet; LotSheet extracts rows, normalises units, applies the buyer's limits and guard band, and renders a branded one-page datasheet PDF with pass/near-miss/fail per property, plus a non-conformance summary and a copyable reply email.

Differentiation · A Word template still requires the manual comparison and unit conversion that causes the errors; LotSheet does the tolerance math and re-issues the same lot against a second buyer's spec in one click.

1. Product opportunity and positioning

A small plastics recycler or compounder sells regrind, flake or recycled pellets into a market where every converter and brand buyer maintains its own incoming spec sheet. Before a load is accepted, the buyer asks for a technical datasheet proving the lot meets those limits. The recycler's evidence is a lab report — often a PDF from an external lab — plus a spreadsheet. Someone on the quality or sales side retypes MFI, density, tensile yield, elongation, Izod, IV, moisture, ash, colour Lab* and contamination values into a Word template, then eyeballs each against the buyer's operator and limit.

That is the bottleneck. It is not scientific, it is clerical, and it recurs every time a buyer asks. It takes hours per buyer, it produces transcription errors, and the person doing it is usually the most expensive non-technical person in a small company. When a borderline value ships unnoticed, the cost is far larger than the labour: a rejected load, a claim, a buyer who stops calling.

LotSheet Resin Reports is positioned as the document engine for that moment. The input is numeric evidence the customer already owns. The output is a file the buyer has explicitly requested, so the seller has a deadline and a reason to pay this week. There is nothing to integrate, no data to license, no certification to obtain, and no lab work involved. The product formats and compares; it never invents a property value.

The wedge is deliberately narrow: one lot, one buyer spec, one branded PDF, one reply email. It expands naturally without new acquisition, because the same saved lot can be re-rendered against a second buyer's spec sheet in one click, and every re-render is a paid report.

Competitive alternative today is Excel, Word and a careful employee. LotSheet does not replace the lab, the LIMS or the ERP, and it makes no accreditation or food-contact claim. It replaces the spreadsheet comparison and the manual layout. That framing keeps the sale small, below procurement thresholds, and answerable by a single quality manager with a company card.

The commercial hypothesis to test in 30 days is stated plainly: small recyclers currently rebuild buyer datasheets by hand, and at least one in five who receive a free converted datasheet will pay $199 to keep the workflow. If free conversions arrive but no one pays, the wedge is wrong and the pricing or output must change.

2. Smallest sellable MVP

One screen takes a lot's lab results (CSV paste or a manual property grid), one screen takes the buyer's spec sheet (PDF upload with a mandatory confirm-and-edit table), and one renderer produces a branded one-page PDF plus a non-conformance summary and reply-email text. A Stripe Checkout link sells a 5-report pack and a magic-link account holds the report credits and the saved lot library.

The MVP is sellable before PDF extraction is perfect, because the confirm screen lets the founder or the customer correct every parsed row. Scanned and unusual spec sheets fall back to the manual grid rather than producing a silent guess.

3. Technical implementation plan

Next.js on Vercel, Supabase for Postgres, auth and file storage, Stripe Checkout for packs and the monthly plan, a hosted multimodal LLM for PDF table extraction, pdf-lib or React-PDF for the datasheet, and Resend for emailed copies. The extraction step outputs structured JSON rows only, and nothing renders until a human confirms the row table. The tolerance engine is pure deterministic TypeScript: unit normalisation, min/max/range/plus-minus operators, a configurable guard band, then a pass, near-miss or fail label. Every render is written to storage with a shareable link so a re-issue against another spec is a copy-and-render, not a rebuild.

The hardest technical risk is messy input: scanned spec sheets, mixed units and property names that do not match the fixed set. The mitigation is the confirm screen plus a manual grid fallback, and a rule that unparsed rows stay unparsed rather than being inferred.

4. Pricing and monetization

Three offers: Starter 5 Lot Reports at $199 one-time, Monthly Desk at $99 per month for up to 40 datasheets with the saved lot library, and a $99 Concierge First Datasheet where the founder builds the first report with the customer and sets the spec up for reuse. The first revenue offer is the concierge or the pack, whichever the prospect reaches for after receiving a free converted datasheet. Upsell path is pack, then monthly desk, then re-issues as buyer count grows.

5. First-customer acquisition

One channel: cold direct messages to named quality and sales managers at small recyclers and compounders, sourced from public trade directories and LinkedIn. The offer is concrete and free to accept: send one lab report and the buyer's spec sheet, receive a formatted pass/fail datasheet within 24 hours, no signup. Delivery happens through the product, and the paid ask is made against that exact file.

6. 30-day execution roadmap

Days 1-5 build and deploy the MVP. Days 6-7 build the 150-name list and send the first 25 messages. Days 8-14 deliver concierge conversions, push the first paid pack, and post anonymised before/after datasheets in plastics-recycling groups. Days 15-21 ship the saved lot library and one-click re-issue, run two more outreach batches, and test the monthly plan with a real subscriber. Days 22-28 work referrals, a second directory-based outreach pass, a parsing reliability pass, and a monthly-upgrade push. Days 29-30 close the outreach sprint and reconcile Stripe against TrustMRR. Daily cadence throughout is fixed: 25 new outbound messages, all inbound conversions delivered inside 24 hours, one product improvement shipped.

7. Revenue and profit scenarios

The three scenarios below are projections from stated assumptions, not verified revenue. Conservative: 3 customers at a blended $133 average (mix of $99 concierge and $199 packs), $399 gross, $12 payment fees, $22 operating cost. Base: 10 customers at $149 blended, $1,490 gross, $48 payment fees, $55 operating cost. Upside: 22 customers at $163 blended, $3,586 gross, one $99 refund, $111 payment fees, $150 paid marketing, $110 operating cost. Payment fees assume Stripe at 2.9% plus $0.30. Paid marketing is zero in the two lower cases because the channel is organic outreach.

8. Main risks and validation experiments

First, free concierge work consuming all available hours: capped at three deliveries per day. Second, parsing errors producing a wrong pass/fail: the confirm screen is mandatory and unparsed rows stay unparsed. Third, holiday-season reply rates below 10%: measured daily, with volume increased rather than message rewritten. Fourth, prospects calling it 'just a Word template': the paid ask is always made against their own delivered file. Kill criteria for each are in the structured plan.

9. First 24-hour action plan

Hours 1-2: scaffold the repo, Supabase schema for lots, specs and reports, and deploy an empty page. Hours 3-5: build the manual property grid and the fixed property set. Hours 6-8: build the tolerance engine and unit normalisation with test fixtures. Hours 9-10: wire Stripe Checkout for the $199 pack. Hours 11-12: start the 150-name prospect list and draft the outreach message. Everything produced in those blocks is a working artifact, not a plan.

Section 2 of 8

Smallest sellable MVP

A deployed app where a customer pastes or uploads a lot's lab results, confirms the buyer's spec rows, pays $199 by card and downloads a branded pass/fail datasheet PDF plus a reply-email draft.

Must-have features

  • Manual property grid plus CSV paste mapped onto a fixed property set (MFI, density, tensile yield, elongation, Izod, IV, moisture, ash, colour L*a*b*, contamination)
  • Buyer spec-sheet PDF parser with a mandatory confirm-and-edit row table
  • Deterministic tolerance engine: unit normalisation, min/max/range/plus-minus operators, configurable guard band
  • Pass, near-miss and fail label per property
  • Branded one-page datasheet PDF with logo, lot ID, issue date and method column
  • Non-conformance summary plus a copyable reply email
  • Saved lot library with one-click re-issue against a second buyer's spec
  • Stripe Checkout for the $199 5-report pack and magic-link accounts

Explicitly out of scope

  • No LIMS, instrument or ERP integration
  • No certification, accreditation or food-contact claims
  • No supplier marketplace or buyer directory
  • No AI-generated or predicted property values
  • No multi-role accounts, SSO or enterprise audit trail
  • No mobile app in the first 30 days
  • No ERP export formats beyond PDF and CSV

Demo flow (what the buyer sees)

  1. Customer uploads or pastes the lot's lab result rows
  2. Customer uploads the buyer's spec sheet PDF
  3. Product proposes property/limit rows and the customer confirms or edits them
  4. Tolerance engine normalises units and labels each property pass, near-miss or fail
  5. Customer uploads a logo and confirms lot ID and issue date
  6. Product renders the branded one-page datasheet PDF and non-conformance summary
  7. Customer copies the reply-email draft and sends the PDF to the buyer
  8. Lot is saved; the same lot is re-issued against a second buyer's spec in one click
Build plan

Technical implementation plan

Stack

Next.js on VercelSupabase Postgres, auth and file storageStripe CheckoutHosted multimodal LLM API for PDF table extractionpdf-lib or React-PDF for datasheet renderingResend for emailed PDFsTailwind CSS

Architecture

A Next.js app on Vercel talks to Supabase for auth, lot records, spec records, report records and file storage. Two ingestion paths write into the same normalised row table: a deterministic CSV/manual grid path and an LLM extraction path that returns structured JSON only. Nothing renders until the user confirms the row table. A pure TypeScript tolerance module converts units, applies the buyer's operators and guard band, and emits pass/near-miss/fail per property. The renderer produces the PDF, stores it in Supabase storage, and issues a shareable link. Stripe Checkout credits report counts to the account; every render decrements one credit.

Build steps

  1. Scaffold Next.js repo on Vercel with Supabase project and env vars
  2. Define schema: accounts, lots, lot_property_values, buyer_specs, spec_rows, reports, credits
  3. Build the manual property grid and CSV paste into the fixed property set
  4. Write the unit normalisation and tolerance engine with fixture tests for min/max/range/plus-minus
  5. Build the spec-sheet upload, LLM extraction call and mandatory confirm-and-edit table
  6. Build the datasheet PDF renderer with logo, lot ID, issue date, method column and status colours
  7. Build the non-conformance summary and reply-email text generator
  8. Add magic-link auth, credit ledger and Stripe Checkout for the $199 pack
  9. Add saved lot library and one-click re-issue against a second spec
  10. Deploy, run three real lab reports end to end, fix rendering and parsing defects

AI components

  • Multimodal LLM for PDF/scan table extraction into structured property rows
  • Optional LLM drafting of the reply email body from the non-conformance list, always user-editable
Hardest technical risk: Scanned, photographed or unusual buyer spec sheets that the extraction model reads poorly, producing a wrong limit that flows into a wrong pass/fail label. Mitigated by the mandatory confirm screen, a manual grid fallback, and a hard rule that unparsed rows stay blank rather than being inferred.
Section 3 of 8

Pricing and monetization

Concierge First Datasheet

Entry offer
$99 one-time

For Quality or sales manager who wants the first datasheet built for them while they watch

  • One done-with-you datasheet build
  • Their buyer's spec sheet set up for reuse
  • Shareable PDF link

Starter - 5 Lot Reports

$199 one-time

For Small recycler with occasional buyer spec requests

  • 5 branded lot datasheets
  • Spec-sheet parser and tolerance check
  • Non-conformance summary and reply-email drafts
  • Saved lot library

Monthly Desk

$99 / month

For Recycler or compounder issuing datasheets weekly

  • Up to 40 lot datasheets per month
  • Saved lot library with one-click re-issue
  • Shareable PDF links

Model: One-time report packs for occasional buyers, a low monthly plan for weekly issuers, and a done-with-you first datasheet, all paid by card through Stripe Checkout.

First revenue offer: Send me one lab report and the spec sheet your buyer uses and I will return a formatted pass/fail datasheet PDF within 24 hours, free, no signup. If it saves you the afternoon, the 5-report pack is $199.

Expected ACV: $199

Upsell path: Concierge $99 converts to the $199 pack, pack buyers with weekly volume move to the $99 Monthly Desk, and every additional buyer spec is a paid re-render on an already-saved lot.

Target customers
25
Revenue model
$4,975
Costs
$200
Profit model
$4,775
Section 4 of 8

First-customer acquisition

Primary channel

Cold direct outreach to named quality and sales managers at small plastics recyclers and compounders, offering a free 24-hour datasheet conversion.

Where to find them

  • Public member directories of national and state recycling associations
  • National and regional plastics recycler and compounder trade directories
  • LinkedIn search for quality manager, QA manager and sales manager titles at firms with 5-200 staff in plastics recycling
  • Company websites of regrind, flake and recycled pellet sellers that publish a contact or sales mailbox
  • Exhibitor lists from regional plastics and recycling trade shows
  • Posts and comment threads in plastics-recycling LinkedIn groups and trade forums

Exact outreach motions

  • Direct message to the named quality manager: offer a free 24-hour conversion of one existing lab report against their buyer's spec sheet
  • Email to the sales manager with the same offer, framed around answering a buyer's spec request the same day
  • Follow-up after three days asking for the next lot they expect to ship
  • Ask each delivered prospect for one referral to a compounder or buyer contact they work with
  • Post an anonymised before/after datasheet pair in two or three plastics-recycling LinkedIn groups as a free resource
  • Re-contact prospects whose free datasheet was delivered but not paid, using their own file as the proof

Message template

Hi [Name] - quick one. When a buyer sends their spec sheet, how long does your team spend rebuilding the lot datasheet by hand? Send me one lab report and the spec sheet that buyer uses and I will send back a formatted pass/fail datasheet PDF within 24 hours, free, no signup. If it is useful the 5-report pack is $199. Worth a try?
Week-one goal: 150 named prospects in the list, 50 messages sent, at least 5 replies and 3 free datasheet conversions delivered inside 24 hours.
Section 5 of 8

30-day execution roadmap

MVP deadline: day 5First sale target: day 1130 days planned
Week 1 · Build & validate · 7 days
Day 1Scaffold

Deployable skeleton and schema

  • Create Next.js repo on Vercel and Supabase project
  • Define accounts, lots, specs, reports and credits tables
  • Deploy an empty authenticated page
  • Start the prospect spreadsheet with the first 30 names

Deliverable: Live empty app and a schema in Supabase

Day 2Intake

Lab result entry

  • Build the fixed property set and manual grid
  • Add CSV paste mapped to the grid
  • Persist lot records to Supabase
  • Enter two real-style test lots as fixtures

Deliverable: Working lot intake screen with saved lots

Day 3Tolerance engine

Deterministic comparison

  • Write unit normalisation for each property
  • Implement min/max/range/plus-minus operators and guard band
  • Emit pass, near-miss, fail per property
  • Cover each operator with fixture tests

Deliverable: Tested tolerance engine module

Day 4Spec parsing

Buyer spec sheet intake

  • Add spec sheet PDF upload and storage
  • Call extraction model for structured rows
  • Build the confirm-and-edit row table
  • Fallback to manual grid when rows do not parse

Deliverable: Spec parser with mandatory confirmation step

Day 5Output and payMVP shipped

PDF, email and checkout

  • Render branded one-page datasheet PDF
  • Generate non-conformance summary and reply email
  • Wire Stripe Checkout for the $199 pack and credit ledger
  • Run one lot end to end on production

Deliverable: MVP live: paid path produces a real datasheet PDF

Day 6List building

150 named prospects

  • Pull recycler and compounder names from public directories
  • Find quality or sales manager by name on LinkedIn
  • Record company, name, title, channel and lot type
  • Clean the list to 150 rows with contact routes

Deliverable: 150-row prospect list with named contacts

Day 7First outbound

25 messages sent

  • Send 25 direct messages with the free 24-hour offer
  • Log every send in the tracker
  • Queue replies for same-day delivery
  • Draft the anonymised before/after asset from test lots

Deliverable: 25 sends logged, first replies in hand

Week 2 · Launch & first sale · 7 days
Day 8Concierge delivery

Convert replies to datasheets

  • Deliver every received lab report and spec as a datasheet within 24 hours
  • Correct parsing errors manually and note each defect
  • Send the PDF with the reply-email draft
  • Ask one question: what would you have done without this

Deliverable: First 3 free datasheets delivered

Day 9Outbound batch two

Volume and follow-up

  • Send 25 new messages
  • Follow up the day-7 batch after three days
  • Fix the top parsing defect found on day 8
  • Log reply rate by segment

Deliverable: 50 cumulative sends and a defect fix shipped

Day 10Parsing reliability

Real spec sheets

  • Collect every spec sheet variant received so far
  • Harden extraction prompts and row mapping for those variants
  • Add clearer empty-row warnings on the confirm screen
  • Re-render two delivered datasheets with fixes

Deliverable: Parsing pass rate measured on real customer spec sheets

Day 11First paid askFirst sale target

Turn delivered files into revenue

  • Send the $199 pack offer against each delivered datasheet
  • Offer the $99 concierge to anyone hesitating on the pack
  • Verify Stripe payment routes into the TrustMRR-connected account
  • Log every objection verbatim

Deliverable: First paid order or a documented objection list

Day 12Delivery and referrals

Fulfil and ask

  • Fulfil paid orders and confirm credit balance shows correctly
  • Ask every buyer and free recipient for one referral
  • Send 25 new outbound messages
  • Publish the anonymised before/after datasheet pair

Deliverable: Paid customers served and referral asks sent

Day 13Public proof

Groups and forums

  • Post the before/after asset in two plastics-recycling LinkedIn groups
  • Reply to every comment with the free-conversion offer
  • Send 25 new outbound messages
  • Track inbound from group posts separately

Deliverable: Two public posts live with inbound tracked

Day 14Week two review

Numbers and next batch

  • Compute reply rate, conversion rate and revenue to date
  • Cut the weakest message angle and keep the best
  • Refresh the list with 50 new named prospects
  • Plan week three outreach batches

Deliverable: Week-two metrics sheet and revised message

Week 3 · Repeat & systematize · 7 days
Day 15Re-issue feature

Saved lot library

  • Ship saved lot library with one-click re-issue against a second spec
  • Add shareable read-only PDF links
  • Add report history per lot
  • Re-render an existing customer lot against a new buyer spec as proof

Deliverable: Re-issue live and demonstrated to one customer

Day 16Outbound batch three

50 prospects

  • Send 25 new messages
  • Follow up all day-9 and day-12 non-replies
  • Offer the re-issue demo to every past recipient
  • Log replies with lot type and firm size

Deliverable: 75 cumulative sends and follow-ups cleared

Day 17Reply quality

Email and non-conformance output

  • Improve the reply-email draft to name the flagged properties explicitly
  • Add a copy button and plain-text variant
  • Test on three real non-conformance cases
  • Ship and note defects

Deliverable: Reply-email draft usable without editing

Day 18Monthly plan test

Recurring revenue

  • Enable the $99 Monthly Desk checkout
  • Offer it to any prospect who mentions weekly datasheets
  • Confirm subscription credits and limits work
  • Log who chose monthly versus pack

Deliverable: At least one monthly subscription attempt tested end to end

Day 19Outbound batch four

Volume plus delivery

  • Send 25 new messages
  • Deliver all inbound conversions within 24 hours
  • Fix the top confirm-screen complaint
  • Ask two delivered prospects for a referral

Deliverable: 100 cumulative sends and all deliveries on time

Day 20Conversion push

Paid against proof

  • Re-offer the $199 pack to every delivered-but-unpaid prospect
  • Send the pack link with their own datasheet attached
  • Offer the $99 concierge as a lower-friction entry
  • Record final objection themes

Deliverable: Conversion attempt logged for every delivered prospect

Day 21Week three review

Pricing and channel

  • Review revenue, refunds and payment fees
  • Test a 3-report $129 pack against the $199 pack
  • Decide whether to keep or drop the monthly plan
  • Refresh the prospect list with 50 new names

Deliverable: Pricing decision documented with data

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

Warm introductions

  • Ask every past customer and delivered prospect for one named referral
  • Send referral messages the same day
  • Offer a free extra report for each referral that converts
  • Log referral source in the tracker

Deliverable: Referral requests sent with named targets

Day 23Second surface

Directory email

  • Email 20 firms whose directory listing shows recycled resin sales
  • Test a subject line focused on the buyer's spec request
  • Send 25 standard direct messages as control
  • Compare reply rates between the two

Deliverable: Second-surface reply rate measured against control

Day 24Outbound batch five

Objection handling

  • Send 25 new messages plus all follow-ups
  • Rewrite the message to address the top objection
  • Deliver all inbound conversions same day
  • Update the FAQ used in replies

Deliverable: Revised message shipped and sends logged

Day 25Reliability pass

Never a wrong number

  • Audit every rendered report for silent fallbacks
  • Block rendering when a property has no confirmed limit
  • Add a visible formats-supplied-data-only footnote
  • Re-test on two scanned spec sheets

Deliverable: Safe-failure behaviour verified on scanned inputs

Day 26Proof asset

Anonymised case files

  • Assemble an anonymised three-datasheet sample pack
  • Attach it to all follow-up emails
  • Post it once in a trade forum
  • Track downloads or replies it produces

Deliverable: Sample pack live and attached to outreach

Day 27Upgrade push

Pack to monthly

  • Offer Monthly Desk to every pack buyer who used three or more reports
  • Confirm subscription billing behaves correctly
  • Ask each buyer how many buyers request datasheets
  • Log expansion potential per account

Deliverable: Upgrade offers sent with usage evidence

Day 28Reconciliation

Money and records

  • Reconcile Stripe payments against TrustMRR records
  • Confirm no pre-launch payments exist
  • Check refunds and chargebacks are zero or documented
  • Compute hackathon profit to date

Deliverable: Reconciliation sheet matching Stripe to TrustMRR

Days 29-30 · Consolidate · 2 days
Day 29Final sprint

Last outbound push

  • Send 30 final messages to the untouched list segment
  • Follow up every unpaid delivered prospect once
  • Deliver all inbound conversions within hours
  • Post the before/after asset one last time

Deliverable: Full list contacted at least once

Day 30Close

Final state

  • Record final verified revenue, refunds and fees
  • Write the one-page result summary with next 30-day plan
  • Hand over the top three objections and the fix for each
  • Freeze the build and archive the prospect tracker

Deliverable: Final revenue record and next-month plan

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: 3 paying customers at a blended $133 ARPU, mostly $99 concierge conversions.
  • Base: 10 paying customers at a blended $149 ARPU, mix of $99 concierge and $199 packs.
  • Upside: 22 paying customers at a blended $163 ARPU, with at least two Monthly Desk subscriptions at $99.
  • All scenarios: roughly one in five free conversions pays inside the window; reply rate 8-12% on cold direct messages.
  • Payment processing assumed at Stripe 2.9% plus $0.30 per transaction, effective 3-4% on this mix.
  • Paid marketing is $0 in conservative and base because the channel is organic outreach; upside assumes $150 of sponsored posts or paid group placements.
  • Operating cost covers Vercel, Supabase, the extraction API, PDF rendering and email, budgeted at under $1 per report plus fixed monthly tooling.
  • Upside includes one $99 refund and no chargebacks.
  • All figures are projections from stated assumptions, not verified revenue or existing customers.

Conservative

3 customers
$399
collected revenue · avg payment $133
Refunds
$0
Chargebacks
$0
Payment fees
$12
Paid marketing
$0
Hackathon profit
$387
Operating costs
$22
Business profit
$365

Base case

10 customers
$1,490
collected revenue · avg payment $149
Refunds
$0
Chargebacks
$0
Payment fees
$48
Paid marketing
$0
Hackathon profit
$1,442
Operating costs
$55
Business profit
$1,387

Optimistic

22 customers
$3,586
collected revenue · avg payment $163
Refunds
$99
Chargebacks
$0
Payment fees
$111
Paid marketing
$150
Hackathon profit
$3,226
Operating costs
$110
Business profit
$3,116

Model your own numbers

Runs in your browser · nothing is saved
Collected revenue
$1,990
Payment fees (3.9%)
$78
Paid marketing
$0
Hackathon profit
$1,912
Business profit
$1,857

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. Projections only; no customers, revenue or traction are claimed.

Section 7 of 8

Main risks and validation experiments

Risk 1Free concierge deliveries absorb every working hour and leave no time to convert to paid volume.

Mitigation

Cap at three concierge deliveries per day and use the product, not manual Word work, to produce each one.

Validation experiment

Track minutes per delivered datasheet for the first five conversions and compare against the daily outreach quota.

Kill criteria

If average delivery exceeds 45 minutes after ten conversions and outreach drops below 15 messages a day, stop free conversions and move to paid-only intake.

Risk 2Parsing errors on scanned or unusual spec sheets produce a wrong pass/fail label and destroy trust.

Mitigation

Mandatory confirm-and-edit screen, manual grid fallback, no rendering of unconfirmed limits, and an explicit formats-supplied-data-only footnote.

Validation experiment

Feed five real spec sheets including at least two scans through extraction and count rows needing correction before rendering.

Kill criteria

If more than 30% of rows need manual correction across ten real spec sheets, move to manual spec entry only and reposition around layout plus tolerance math.

Risk 3Holiday-season cold reply rates fall far below the assumed 10%, starving the funnel.

Mitigation

Increase daily send volume, add directory email as a second surface, and prioritise prospects with visible active buyer activity.

Validation experiment

Compare reply rate on 50 messages sent in week one against 50 sent in week three with a revised angle.

Kill criteria

If 150 messages produce fewer than five replies by day 14, the channel or segment is wrong and outreach must move to warm introductions through trade forums.

Risk 4Prospects see the output as just a Word template and refuse $199.

Mitigation

Always make the paid ask against the prospect's own delivered datasheet, emphasising the tolerance math, unit conversion and one-click re-issue.

Validation experiment

Offer the $199 pack to five delivered prospects and record the exact objection wording for each refusal.

Kill criteria

If ten delivered prospects all refuse and every objection is price or perceived value, drop to a $49 single-report price or change the buyer segment.

Risk 5A buyer treats the datasheet as a certification and pushes liability onto the founder.

Mitigation

Standing footer stating the document formats customer-supplied data only, plus no accreditation or food-contact language anywhere in the product.

Validation experiment

Ask three prospects directly whether their buyers would accept a supplier-generated datasheet without certification.

Kill criteria

If more than one prospect says their buyer requires accredited third-party documentation, narrow the segment before spending further effort.

Section 8 of 8

Hour-by-hour: the first 24 hours

Hours 1-2

  • Create Next.js repo on Vercel and Supabase project
  • Define accounts, lots, specs, reports and credits tables
  • Deploy an empty authenticated page

Output: Live deployed app and schema

Hours 3-5

  • Build the fixed property set constant
  • Build the manual property grid screen
  • Add CSV paste mapped into the grid

Output: Working lab-result intake screen

Hours 6-8

  • Write unit normalisation for MFI, density, tensile, elongation, Izod, IV, moisture, ash and colour
  • Implement min/max/range/plus-minus operators and guard band
  • Emit pass, near-miss or fail and cover each operator with a fixture test

Output: Tested tolerance engine module

Hours 9-10

  • Wire Stripe Checkout for the $199 5-report pack
  • Add the credit ledger decrement on render
  • Test one checkout in Stripe test mode

Output: Working paid path in test mode

Hours 11-12

  • Build the 150-row prospect spreadsheet structure
  • Add the first 25 named recycler or compounder prospects
  • Draft and save the outreach message template

Output: Prospect tracker with 25 named contacts and a final message

Complete report

Full 30-day launch report

Read the full generated report

1. Product opportunity and positioning

A small plastics recycler or compounder sells regrind, flake or recycled pellets into a market where every converter and brand buyer keeps its own incoming spec sheet. Before a load is accepted, the buyer asks for a technical datasheet proving the lot meets those limits. The recycler's evidence is a lab report, often a PDF from an external lab, plus a spreadsheet. Someone on the quality or sales side retypes MFI, density, tensile yield, elongation, Izod, IV, moisture, ash, colour Lab* and contamination values into a Word template and then eyeballs each against the buyer's operator and limit.

That is the bottleneck. It is clerical, not scientific, and it recurs every time a buyer asks. It takes hours per buyer, produces transcription errors, and the person doing it is usually the most expensive non-technical person in a small company. When a borderline value ships unnoticed, the cost is a rejected load and a buyer who stops calling.

LotSheet Resin Reports is positioned at exactly that moment. Input is numeric evidence the customer already owns. Output is a file the buyer explicitly requested, so the seller has a deadline and a reason to pay this week. Nothing needs integrating, no data is licensed, no certification is needed and no lab work is involved. The product formats and compares; it never invents a property value.

The wedge is deliberately narrow: one lot, one buyer spec, one branded PDF, one reply email. It expands without new acquisition because the same saved lot re-renders against a second buyer's spec in one click, and every re-render is a paid report.

The alternative today is Excel, Word and a careful employee. LotSheet does not replace the lab, the LIMS or the ERP, and it makes no accreditation or food-contact claim. That framing keeps the sale small, below procurement thresholds, and answerable by one quality manager with a company card.

The commercial hypothesis to test in 30 days: small recyclers currently rebuild buyer datasheets by hand, and at least one in five who receive a free converted datasheet will pay $199 to keep the workflow. If free conversions arrive but nobody pays, the wedge or the price is wrong.

2. Smallest sellable MVP

Three screens and a payment link. Screen one takes the lot's lab results by CSV paste or manual property grid. Screen two takes the buyer's spec sheet PDF and shows a mandatory confirm-and-edit row table. Screen three renders a branded one-page datasheet PDF plus a non-conformance summary and reply-email text. Stripe Checkout sells a 5-report pack, and a magic-link account holds credits and the saved lot library.

The MVP is sellable before extraction is perfect because the confirm screen lets a human correct every parsed row. Scanned or unusual spec sheets fall back to the manual grid rather than producing a silent guess.

3. Technical implementation plan

Next.js on Vercel, Supabase for Postgres, auth and storage, Stripe Checkout for packs and the monthly plan, a hosted multimodal LLM for PDF table extraction, pdf-lib or React-PDF for rendering, Resend for emailed copies. Extraction returns structured JSON rows only; nothing renders until the row table is confirmed. The tolerance engine is pure deterministic TypeScript: unit normalisation, min/max/range/plus-minus operators, configurable guard band, then pass, near-miss or fail. Every render is stored with a shareable link so re-issuing against another spec is copy-and-render.

The hardest risk is messy input: scans, mixed units, property names outside the fixed set. Mitigation is the confirm screen, the manual grid fallback, and a hard rule that unparsed rows stay unparsed.

4. Pricing and monetization

Starter 5 Lot Reports at $199 one-time, Monthly Desk at $99 per month for up to 40 datasheets, and a $99 Concierge First Datasheet where the founder builds the first report with the customer and sets the spec up for reuse. The first ask goes against the file the prospect already received. Upsell path is concierge to pack, pack to monthly, and re-issues as the customer's buyer count grows.

5. First-customer acquisition

One channel: cold direct messages to named quality and sales managers at small recyclers and compounders sourced from public trade directories and LinkedIn. The offer is free to accept: send one lab report and the buyer's spec sheet, receive a formatted pass/fail datasheet within 24 hours, no signup. Delivery runs through the product so the paid ask is made against that exact file.

6. 30-day execution roadmap

Days 1-5 build and deploy. Day 1 scaffolds the repo and schema; day 2 ships lab-result intake; day 3 ships the tolerance engine with fixture tests; day 4 ships spec parsing with the confirm screen; day 5 ships PDF rendering, the reply email and Stripe Checkout, and the MVP goes live. Days 6-7 build the 150-name list and send the first 25 messages. Days 8-14 deliver concierge conversions inside 24 hours, make the first paid ask on day 11, publish anonymised before/after datasheets in plastics-recycling groups, and review metrics on day 14. Days 15-21 ship the saved lot library and one-click re-issue, run two more outreach batches, test the monthly plan with a real subscriber, and review pricing on day 21. Days 22-28 work referrals, add a directory-email surface as a second test, run a reliability pass that blocks rendering on unconfirmed limits, push pack-to-monthly upgrades, and reconcile Stripe against TrustMRR on day 28. Days 29-30 close the final outreach sprint and record final verified revenue.

Daily cadence never changes: 25 new outbound messages, every inbound conversion delivered within 24 hours, one product improvement shipped.

7. Revenue and profit scenarios

Projections only, from stated assumptions, not verified revenue. Conservative: 3 customers at a blended $133 average, $399 gross, $12 payment fees, $22 operating cost. Base: 10 customers at $149 blended, $1,490 gross, $48 payment fees, $55 operating cost. Upside: 22 customers at $163 blended, $3,586 gross, one $99 refund, $111 payment fees, $150 paid marketing, $110 operating cost. Payment fees assume Stripe at 2.9% plus $0.30. Paid marketing is zero in the two lower cases because the channel is organic. Operating cost covers Vercel, Supabase, extraction API, PDF rendering and email at under roughly $1 per report plus fixed tooling.

8. Main risks and validation experiments

Four risks carry kill criteria. Free concierge work eating all hours: capped at three per day, measured in minutes per delivery, and abandoned if delivery exceeds 45 minutes after ten conversions. Parsing errors producing a wrong label: confirm screen plus manual fallback, tested on ten real spec sheets, with manual-only entry if correction exceeds 30% of rows. Holiday reply rates: measured week one against week three on 50-message batches, with channel change if 150 messages yield fewer than five replies. Commodity perception, that it is just a Word template: the paid ask is always against the prospect's own delivered file, and a $49 single-report price is the fallback if ten delivered prospects all refuse. A fifth risk, liability framing, is handled by a standing formats-supplied-data-only footnote and no certification language anywhere.

9. First 24-hour action plan

Hours 1-2: scaffold the repo, Supabase schema and deployed page. Hours 3-5: property grid and CSV paste. Hours 6-8: tolerance engine with fixture tests. Hours 9-10: Stripe Checkout for the $199 pack in test mode. Hours 11-12: prospect tracker structure with the first 25 named contacts and the final message. Each block ends with an artifact, not a plan.

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 →