MockSmith
MockSmith turns a tutor's syllabus and past papers into a print-ready, blueprint-weighted mock exam with answer key and OMR sheet 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.
Owner-operator of a small test-prep coaching center or independent tutor with 1-10 tutors and 20-200 students who already sells mock tests for a high-stakes exam.
Operational simplicity 4.8/10
Remove regulated data, human review steps and custom infrastructure; keep the product self-serve and stateless on managed services.
Why this could work in 30 days: one founder can ship the MVP in 5 days; the plan targets a first sale on day 12; 11 sales at $99 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 8.7/10.
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
11 concrete actions. Tick them off as you go — nothing is saved or sent.
Product opportunity and positioning
Target customer
Owner-operator of a small test-prep coaching center or independent tutor with 1-10 tutors and 20-200 students who already sells mock tests for a high-stakes exam.
Why now
Constrained generation plus cheap OCR and PDF tooling makes a solo-built, self-serve artifact generator viable; the buyer already sells mocks, so the output is monetizable on day one.
Problem
Every mock exam is hand-assembled from scattered PDFs: matching official section weights, balancing difficulty, formatting answer keys and OMR sheets. Generic AI dumps produce unusable, unbalanced sets.
Solution
The tutor enters or uploads an exam blueprint plus past papers; MockSmith locks the section counts, difficulty mix and syllabus tags, generates a 60-question mock, and exports a branded print-ready PDF pack with OMR sheet, answer key and per-question rationales.
1. Product opportunity and positioning
MockSmith sells a finished artifact, not a tool subscription. The buyer is the owner-operator of a small test-prep coaching center or an independent tutor with 1-10 tutors and 20-200 students, already selling mock tests for a high-stakes exam. They are triggered by a new batch intake or the start of mock-test season, when they must produce five to twenty mocks in a few weeks.
The painful task is production, not teaching. A center owner opens four PDFs, an old OMR sheet, a Word template and the official syllabus, then spends two to four evenings assembling one 60-question paper: counting questions per section to match official weights, guessing at difficulty balance, writing an answer key, formatting an OMR bubble sheet, and checking for duplicates. Generic AI tools write plausible questions but no blueprint compliance, so the tutor still does the assembly work and cannot trust the set.
MockSmith's wedge is the Blueprint Lock. The tutor supplies the exam blueprint (section counts, question types, target difficulty mix) plus one to three past papers or their own notes. The parser converts that into a machine-checkable constraint set. The generator produces a 60-question mock that is rejected and regenerated until every section count, difficulty target and syllabus tag is satisfied. The output is a print-ready PDF pack: branded cover, instructions page, question paper, OMR sheet, answer key, and a rationale appendix tagging each question by syllabus node and difficulty.
The center downloads it, prints it, and sells or assigns it the same week. That is why the price is defensible: the buyer's alternative is either evenings of labor or an expensive question bank licence that still requires assembly.
Positioning is deliberately narrow. MockSmith is not a student-facing tutor, not a live test delivery platform, not an autograder, and not a marketplace. Those are explicitly out of scope. The product wins on one claim: blueprint compliance plus print-ready packaging in minutes.
Why this is sellable now rather than in a year: constrained LLM generation with retry loops is cheap enough that a pack costs the founder a few dollars, OCR and PDF assembly are commodity managed services, and a solo operator can ship the whole pipeline in under a week on serverless infrastructure. There is no integration, no procurement cycle and no regulatory approval between the founder and the first payment.
The JEV score of 65.7 (band B) reflects exactly this shape: strong speed to first revenue and reachable customers, weaker operational simplicity because generation quality and OCR variance create real support surface. The thirty-day plan below is built to convert the speed advantage into collected revenue before the quality risk compounds.
2. Smallest sellable MVP
The MVP is a single page where a tutor enters a center name, picks an exam, fills in a blueprint form (or uploads a blueprint PDF), uploads one to three past papers, pays, and receives an email with a print-ready PDF pack.
Must-have: the blueprint form and parser; OCR plus topic extraction from uploaded papers; a constrained generator with retry-until-compliant logic and duplicate detection; the PDF pack (cover, instructions, questions, OMR, answer key, rationale appendix); Stripe Checkout with one-time credits; and email delivery of download links.
Out of scope for thirty days: student accounts, live test delivery, auto-grading, LMS integrations, a question marketplace, human subject-matter review, team permissions, and any student-facing AI tutor.
3. Technical implementation plan
Next.js on Vercel front end and API routes. Supabase for auth, file storage and project records. PDF.js for text extraction, a managed OCR API for scanned papers, OpenAI for structured generation, React-PDF for assembly, Stripe for payment, Resend for delivery.
The build order matters. Blueprint schema first, because everything downstream is validated against it. Then the generator loop, then the PDF template, then payment. Payment is wired on day five, not day one, but the Stripe account and TrustMRR connection are set up on day one so no revenue is missed.
The hardest technical risk is question quality under constraint: the model may satisfy counts while producing duplicate, trivial or factually wrong items. Mitigation is a validation gate plus a cheap human check on the first ten packs.
4. Pricing and monetization
Three tiers: Single Mock at $99 one-time, Exam Season Pack at $499 one-time for eight mocks, and Coaching Center Monthly at $199 per month for ten mocks. The first sale should be a $99 Single Mock because it requires no committee and can be approved from an operating budget. The $499 season pack is the real revenue target and gets pitched after one delivered pack.
5. First-customer acquisition
One channel only: personalized sample mocks sent by Instagram DM and WhatsApp to owner-operators of small coaching centers in a single metro, sourced from Google Maps and Instagram. Build a 300-row list, filter to centers with 1-10 tutors and visible mock-test activity, generate a free 10-question branded sample per prospect, and send a short DM offering the full 60-question pack.
6-9. Roadmap, scenarios, risks, first day
The thirty-day roadmap runs build (days 1-5), quality and list building (6-8), outreach and first close (9-14), season-pack push (15-21), and closing volume plus reconciliation (22-30). Financial scenarios are projections only, not verified traction. The three risks that decide the outcome are cold-outreach response rate, question quality causing refunds, and Stripe review on a new account taking many $499 payments; each has a cheap validation experiment and an explicit kill criterion below. The first twenty-four hours produce a deployed skeleton, live payment rails, and the first fifty prospects, so that no build day is ever disconnected from the sales pipeline.
Smallest sellable MVP
A paid one-page flow: center name + exam, blueprint form or upload, past-paper upload, Stripe Checkout, and an emailed print-ready PDF pack with OMR sheet and answer key.
Must-have features
- Blueprint Lock form: section names, question counts, question types, difficulty mix
- Blueprint PDF parser that extracts section weights into a constraint schema
- Past-paper upload with OCR and syllabus-topic extraction
- Constrained generator that retries until section counts, difficulty targets and topic tags pass validation
- Duplicate and near-duplicate question detector
- Print-ready PDF pack: branded cover, instructions, question paper, OMR sheet, answer key, rationale appendix
- Stripe Checkout with one-time pack credits
- Resend email delivery of download links
- Admin-free generation queue with retry and failure email
Explicitly out of scope
- Student accounts or live test delivery
- Auto-grading or OMR scanning
- LMS and school system integrations
- Question bank marketplace
- Human subject-matter review service
- Team permissions or enterprise procurement features
- Mobile app
- Adaptive learning
Demo flow (what the buyer sees)
- Tutor enters center name and selects the exam
- Tutor fills the blueprint form or uploads a blueprint PDF
- Tutor uploads one to three past papers or notes
- System parses the blueprint and extracts syllabus topics
- Generator produces a 60-question set and validates against the blueprint
- Failed items are regenerated until constraints pass
- System assembles the branded PDF pack including OMR sheet and answer key
- Tutor pays $99 via Stripe Checkout
- Email arrives with download links to the pack
Technical implementation plan
Stack
Architecture
A single Next.js app serves the form, the dashboard and the API. Uploads land in Supabase Storage; a parse job converts blueprint and past papers into a JSON constraint object. A generation job calls OpenAI with that constraint object, validates the response, regenerates only failing slots, deduplicates, then hands the accepted question set to React-PDF, which writes the pack to storage. Stripe webhook flips the project from draft to paid, enqueues generation, and on completion Resend emails the signed download links.
Build steps
- Deploy Next.js skeleton to Vercel and create Supabase project
- Define the blueprint JSON schema and build the blueprint form
- Add PDF.js text extraction and OCR fallback for uploads
- Write the blueprint parser prompt returning the schema, with validation
- Build the constrained generator with per-slot retry and duplicate detection
- Build the React-PDF pack template (cover, instructions, questions, OMR, key, rationales)
- Add Stripe Checkout for three tiers and webhook handling
- Add generation queue, retry handling and failure notification
- Add Resend email with signed download links
- Run three real exam blueprints end to end and fix the top failures
- Ship live, connect TrustMRR, run one real test purchase and refund it
AI components
- Blueprint parser LLM call that converts uploaded blueprint text into a constraint schema
- Syllabus-topic extraction from past papers
- Constrained 60-question generator with per-section retries
- Rationale text per question
- Duplicate and near-duplicate detection embedding
Pricing and monetization
Single Mock
Entry offerFor Independent tutor testing the tool for one exam
- 1 blueprint-weighted 60-question mock
- Answer key and OMR sheet
- Per-question syllabus and difficulty tags
Exam Season Pack
For Small coaching center preparing a batch for one exam
- 8 blueprint-weighted mocks
- Branded covers and OMR sheets
- Answer keys and rationale appendices
- 2 revision rounds on the same blueprint
Coaching Center Monthly
For Center producing mocks every month across multiple exams
- 10 mocks per month
- Up to 3 exam blueprints
- Priority email support
Model: One-time pack credits with an optional monthly plan for repeat producers.
First revenue offer: $99 Single Mock: send a free 10-question branded sample, then offer the full 60-question print-ready pack with answer key and OMR sheet, delivered within 24 hours, regenerated free if any section weight is off.
Expected ACV: $199
Upsell path: Single Mock buyers are pitched the $499 Exam Season Pack after their first delivered pack; repeat producers move to the $199 monthly plan, and each paid buyer earns $100 credit toward their next pack for every referred center that pays.
First-customer acquisition
Personalized sample mocks sent by Instagram DM and WhatsApp to owner-operators of small coaching centers in one metro area.
Where to find them
- Google Maps listings for test-prep coaching centers in one metro, filtered to 1-10 tutor operations
- Instagram accounts of those centers showing mock-test or batch announcements
- WhatsApp business numbers published on center websites and Maps listings
- Local exam-prep Facebook and Telegram groups where center owners post batch openings
- Referrals from each paid buyer (two names per buyer)
Exact outreach motions
- Send a DM with a 10-question sample mock built from that center's name and their public exam syllabus
- Attach a one-page PDF preview of the branded cover, OMR sheet and answer key
- Offer the $99 Single Mock with 24-hour delivery and free regeneration on weight mismatch
- Follow up after 48 hours with a 45-second Loom showing the pack being assembled
- Ask every buyer for two other center owners and offer $100 credit per paid referral
Message template
Hi [Name] - I built a 10-question [Exam] sample using [Center Name]'s own branding so you can see the format. It takes your syllabus weights and turns them into a print-ready 60-question mock with answer key and OMR sheet in under a day. Want the full pack for $99 this week? I'll regenerate free if any section weight is off.
30-day execution roadmap
Week 1 · Build & validate · 6 days
Payment rails and deploy target exist before anything else
- Buy domain and deploy Next.js skeleton to Vercel
- Create Supabase project with auth, storage and tables
- Open Stripe account and connect TrustMRR in test mode
Deliverable: Live skeleton URL and configured payment account
Turn a blueprint into a machine-checkable schema
- Define blueprint JSON schema and build the form
- Add PDF.js extraction with OCR fallback
- Write parser prompt and validate output against schema
Deliverable: Blueprint JSON produced from an uploaded past paper
Questions that must pass constraints or regenerate
- Build per-slot constrained generation
- Add duplicate and near-duplicate detection
- Add retry and failure logging
Deliverable: 60-question set with section, difficulty and topic tags
The artifact the buyer actually pays for
- Build React-PDF template: cover, instructions, questions
- Add OMR sheet and answer key pages
- Add rationale appendix and center logo/name variables
Deliverable: Downloadable print-ready PDF pack
Money in, pack out, no manual step
- Wire Stripe Checkout for three tiers
- Add webhook that flips project to paid and enqueues generation
- Send Resend email with signed download links
Deliverable: End-to-end paid purchase works live
Prove the output is usable on real blueprints
- Run three real exam blueprints through the pipeline
- Fix the top three failure modes found
- Measure API cost per pack
Deliverable: Three sample packs plus a cost-per-pack figure
Week 2 · Launch & first sale · 7 days
300 named prospects in one metro
- Pull coaching centers from Google Maps in one metro
- Filter to 1-10 tutor centers with mock-test activity
- Map each to an Instagram handle or WhatsApp number
Deliverable: 300-row prospect sheet with exam and contact
Personalized proof before any ask
- Generate 20 ten-question branded samples for top prospects
- Export each as a short PDF preview
- Draft the DM variants
Deliverable: 20 personalized sample packs ready to send
Start the outreach clock
- Send 30 DMs with sample attached
- Log every send and reply in the sheet
- Set 48-hour follow-up reminders
Deliverable: 30 sends logged with reply column live
Reach statistical signal on response rate
- Send 40 more DMs to a fresh segment
- Reply to all inbound within 2 hours
- Send first 48-hour follow-ups
Deliverable: 70 total touches and first real conversations
Convert interest into a payment link
- Record 45-second Loom of pack assembly
- Send Stripe links to warm leads
- Offer $99 Single Mock to price-sensitive leads
Deliverable: Five warm leads with payment links sent
Collect the first real payment
- Follow up every warm lead individually
- Offer free watermarked sample then full $99 pack
- Deliver the first paid pack within 24 hours
Deliverable: First paid order and delivered pack
Turn one buyer into distribution
- Ask buyer for two referral names
- Capture a one-line quote and permission to use it
- Fix any complaint raised in delivery
Deliverable: Two referral introductions and one quote
Week 3 · Repeat & systematize · 8 days
Reduce refund surface
- Patch the top three complaint categories
- Add the revision round to the pack spec
- Re-run one earlier blueprint to confirm improvement
Deliverable: Generator v1.1 shipped
Raise average order value
- Pitch $499 Exam Season Pack to every prior buyer
- Send 40 new DMs
- Send 20 second-follow-ups
Deliverable: First $499 conversation opened
Prove the channel repeats outside one metro
- Build a 200-prospect list in a second metro
- Generate 15 branded samples
- Update DM copy with any winning phrase
Deliverable: 200 new prospects and 15 samples
Push touch count past 200
- Send 60 DMs across both lists
- Reply and qualify every response same day
- Book calls only for $499 interest
Deliverable: Eight warm leads in pipeline
Keep delivery under 24 hours
- Generate and deliver three paid packs
- Log generation time and API cost per pack
- Escalate any blueprint the parser mishandles
Deliverable: Three packs delivered with cost logged
Make referrals mechanical
- Ask all buyers for two names with $100 credit offer
- Send credit codes to referrers
- Follow up unredeemed credit
Deliverable: Five referral asks sent
One credible artifact, no invented metrics
- Publish an anonymized walkthrough of a blueprint-to-pack run
- Post in two exam-prep owner communities
- Invite comments to a free sample
Deliverable: One public walkthrough post live
Cut what does not convert
- Calculate reply rate and close rate per message variant
- Kill the weakest variant and double the best
- Recount collected revenue net of fees
Deliverable: Midpoint conversion report and one decision
Week 4 · Scale what works · 7 days
Hit the remaining prospects hard
- Send 60 DMs to untouched prospects
- WhatsApp follow-up to all non-responders
- Send Stripe links to new warm leads
Deliverable: 60 new touches and new links out
Keep delivery cost under $5 per pack
- Cache blueprint parsing per exam
- Cap retry loops at a fixed budget
- Measure cost per pack after changes
Deliverable: Cost per pack reduced and documented
Convert repeat producers
- Pitch $199 monthly to any buyer with two or more packs
- Add monthly plan to the pricing page
- Send three targeted offers
Deliverable: First monthly plan offer sent
Protect profit from refunds
- Deliver all outstanding packs
- Process any refund immediately and log the cause
- Fix the recurring refund cause
Deliverable: Zero undelivered orders
Highest-yield sends only
- Send 80 messages to referral leads and non-responder follow-ups
- Offer the free 10-question sample again
- Prioritize centers with visible mock-test calendars
Deliverable: 80 warm touches logged
Convert evidence into close rate
- Request a testimonial from every buyer
- Ask each buyer for two referral names
- Add approved quotes to the landing page
Deliverable: Two approved quotes live on the site
One exam, one metro, ten packs
- Batch-generate ten samples for a single exam
- Send a coordinated DM wave to that segment
- Attach a deadline offer
Deliverable: Ten samples sent in one segment
Days 29-30 · Consolidate · 2 days
Convert pipeline before the deadline
- Offer the $499 season pack with two revision rounds ending December 16
- Follow up every unpaid link
- Take payment only through the TrustMRR-connected account
Deliverable: All open links chased and resolved
Submit verified profit
- Reconcile Stripe payouts, refunds, chargebacks and fees
- Confirm every delivered order is above delivery cost
- Submit verified profit and write a one-page post-mortem
Deliverable: Reconciled profit figure submitted
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, mostly $99 Single Mocks, blended ARPU $149; Base: 10 customers with a $199 blended mix; Upside: 25 customers with season packs lifting blended ARPU to $259.
- All figures are projections from stated assumptions, not verified revenue or existing customers.
- Payment processing assumed at roughly 4.5% of collected revenue (Stripe card pricing plus international card surcharge).
- Operating costs cover Vercel, Supabase, domain, OCR API and OpenAI usage at roughly $5 of inference per pack plus fixed tooling.
- Organic outreach (DMs, WhatsApp, referrals) costs $0; ad spend is $0 in Conservative and Base, $200 in Upside for sponsored posts.
- Refunds and chargebacks kept at or below 5% of revenue and modeled explicitly per scenario.
- Hosting, tooling and AI delivery costs are not deducted from hackathon profit but are deducted from business profit.
- No pre-launch customers or payments; all revenue occurs inside the contest window.
Conservative
3 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $20
- Paid marketing
- $0
- Hackathon profit
- $427
- Operating costs
- $25
- Business profit
- $402
Base case
10 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $90
- Paid marketing
- $0
- Hackathon profit
- $1,900
- Operating costs
- $90
- Business profit
- $1,810
Optimistic
25 customers- Refunds
- $199
- Chargebacks
- $0
- Payment fees
- $324
- Paid marketing
- $200
- Hackathon profit
- $5,752
- Operating costs
- $220
- Business profit
- $5,532
Model your own numbers
Runs in your browser · nothing is saved- Collected revenue
- $990
- Payment fees (3.9%)
- $39
- Paid marketing
- $0
- Hackathon profit
- $951
- Business profit
- $861
Hypothetical arithmetic, not a forecast. The fee rate is a flat 3.9% assumption, refunds and chargebacks are left at zero here, and none of this is tracked data.
Scenario models from stated assumptions, not verified revenue.
Main risks and validation experiments
Risk 1Cold DM response rate far below 10%, missing every revenue scenario.
Send in waves of 30 and rewrite the message after each wave based on replies; lead with the personalized sample, not the price.
Send 30 DMs on day 9 and count replies within 72 hours.
Fewer than 2 replies from 30 personalized sends means the message or list is wrong; change the segment before spending more days.
Risk 2Generated questions are duplicated, trivial or factually wrong, causing refunds and lost referrals.
Slot-level validation, duplicate detection, and a human spot-check of the first ten packs before delivery.
Run three real exam blueprints on days 6 and 7 and count rejected or regenerated slots.
If more than 20% of questions fail validation or a buyer refunds for quality twice, stop outreach and fix the generator.
Risk 3Stripe review or hold on a new account receiving several $499 payments quickly.
Set up the account and complete verification on day 1; keep refund rate low; respond to Stripe requests within hours.
Run one real $99 test purchase and refund it on day 5 to confirm payouts work.
If payouts are held beyond 7 days, switch to the alternate processor and re-verify TrustMRR connectivity.
Risk 4Uploaded past papers carry copyright restrictions the buyer cannot grant.
Terms require the buyer to confirm rights; generate from blueprints and topic lists where possible rather than verbatim copying.
Ask the first five buyers directly what papers they hold rights to.
If most prospects cannot legally supply papers, pivot the input to blueprint plus tutor notes only.
Risk 5Tutors prefer free ChatGPT or an existing question bank.
Sell the packaged artifact and blueprint compliance, not the questions; show the OMR sheet and answer key in the first message.
Ask every non-buyer one question: what stopped you?
If the top objection from ten prospects is price versus free tools, move the entry offer to $49 for a 30-question pack.
Hour-by-hour: the first 24 hours
Hours 1-2
- Buy domain and deploy Next.js skeleton to Vercel
- Create Supabase project with storage bucket and project table
Output: Live skeleton URL and database ready
Hours 3-5
- Open Stripe account, complete verification
- Connect the payment account to TrustMRR
- Create three products: $99, $499, $199 monthly
Output: Payment rails configured and verified
Hours 6-8
- Write the blueprint JSON schema
- Draft the parser prompt against one real past paper
Output: Blueprint schema file and a working parser draft
Hours 9-11
- Pull 50 coaching centers from Google Maps in one metro
- Record center name, exam focus, Instagram handle or WhatsApp number
Output: First 50 qualified prospects in a sheet
Hours 12-14
- Generate one 10-question branded sample for the strongest prospect
- Draft the DM and the Loom script
Output: One sample pack PDF and the first outreach message ready to send on day 9
Full 30-day launch report
Read the full generated report
MockSmith - 30-Day Launch Report
Product: MockSmith (concept version 1). JEV score 65.7, band B.
1. Product opportunity and positioning
MockSmith sells a finished artifact. The buyer is the owner-operator of a small test-prep coaching center or an independent tutor with 1-10 tutors and 20-200 students, already selling mock tests for a high-stakes exam. The trigger is concrete: a new batch or the start of mock-test season, when five to twenty mocks must exist in a few weeks.
The painful task is production. The owner assembles a 60-question paper by hand from scattered PDFs, counting questions per section to match official weights, guessing at difficulty balance, writing an answer key, formatting an OMR bubble sheet, and checking duplicates. Generic AI tools write plausible questions but no blueprint compliance, so the tutor still does the assembly and cannot trust the set.
The wedge is the Blueprint Lock. The tutor supplies section counts, question types and a target difficulty mix, plus one to three past papers or personal notes. That becomes a machine-checkable constraint set. Generation is rejected and retried until every section count, difficulty target and syllabus tag passes. The output is a print-ready PDF pack with branded cover, instructions, question paper, OMR sheet, answer key and a rationale appendix tagging each question by syllabus node and difficulty.
The center prints it and sells or assigns it that week. That is why $99 is defensible against free ChatGPT: the buyer is purchasing compliance and packaging, not prose.
MockSmith is explicitly not a student tutor, not live test delivery, not auto-grading, and not a question marketplace. It wins one claim: blueprint compliance plus print-ready packaging in minutes. The JEV profile matches this shape - speed to first revenue and reachable customers score well; operational simplicity is the weak point, which is why this plan keeps generation quality and cost bounded from day five.
2. Smallest sellable MVP
One page. Enter center name and exam, fill or upload the blueprint, upload past papers, pay, receive an email with the pack. Everything else is out of scope for thirty days: no student accounts, no live delivery, no autograding, no LMS integrations, no marketplace, no human review service, no mobile app.
The success test for the MVP is not feature completeness. It is whether a stranger can pay $99 and receive a pack they would print without editing.
3. Technical implementation plan
Next.js on Vercel, Supabase for auth, storage and project records, PDF.js with a managed OCR fallback, OpenAI for structured generation, React-PDF for assembly, Stripe for payment, Resend for delivery.
Build order: blueprint schema, then parser, then validator, then generator with per-slot retry, then PDF template, then payment. Day one sets up Stripe and TrustMRR even though checkout goes live on day five, so no revenue is blocked later.
The hardest risk is quality under constraint: a model can satisfy counts while producing duplicates or trivial items. Mitigation is a validation gate, a duplicate detector, and a human spot-check of the first ten packs before they are sold.
4. Pricing and monetization
Single Mock $99 one-time. Exam Season Pack $499 one-time for eight mocks with two revision rounds. Coaching Center Monthly $199 for ten mocks across up to three blueprints. Blended expectations are $149 conservative, $199 base, $259 upside.
The first ask is always $99, because it clears an operating budget without approval. The $499 pack is pitched only after a delivered pack, and referrals carry $100 credit toward the next pack.
5. First-customer acquisition
One channel: personalized sample mocks by Instagram DM and WhatsApp to center owners in one metro, sourced from Google Maps and Instagram. Filter to 1-10 tutor operations with visible mock-test activity. Build 300 prospects by day 7, generate 20 branded ten-question samples by day 8, send 30 DMs on day 9. The message leads with the sample and the center's own name, never with a price list.
6. 30-day roadmap
Days 1-5: build. Payment rails, blueprint parser, constrained generator, PDF pack, live checkout. MVP deadline day 5.
Days 6-8: quality pass on three real blueprints, cost per pack measured, 300-prospect list and 20 personalized samples.
Days 9-14: outreach, first close (target day 12), delivery, referrals, generator v1.1.
Days 15-21: season pack push, second metro list, 200+ total touches, referral engine, one public walkthrough, midpoint review that kills the weakest message.
Days 22-30: warm-segment sprint, unit cost reduction, monthly plan offers, refund prevention, testimonial collection, coordinated batch push, deadline close, and reconciliation of payouts, refunds and fees before submitting verified profit.
7. Revenue and profit scenarios
All three scenarios are projections, not verified revenue.
Conservative: 3 customers at $149 blended ARPU, $447 gross, $20 processing fees, $25 operating cost.
Base: 10 customers at $199 blended ARPU, $1,990 gross, $90 processing fees, $90 operating cost.
Upside: 25 customers at $259 blended ARPU, $6,475 gross, $199 refunds, $324 processing fees, $200 sponsored posts, $220 operating cost.
Hackathon profit deducts revenue, refunds, chargebacks, processing fees and paid marketing only. Business profit additionally deducts hosting and AI delivery. Every scenario keeps delivery cost per pack around $5 against a $99 minimum price, well clear of the below-cost disqualification line.
8. Risks and validation experiments
The three decisive risks. First, cold DM response rate: send 30 and count replies in 72 hours; fewer than two replies means change the segment, not the volume. Second, question quality: run three real blueprints on days 6-7 and reject anything over 20% validation failure before selling. Third, Stripe review: complete verification on day one, run a real $99 purchase and refund on day 5, and confirm payouts clear.
Two secondary risks. Copyright on uploaded papers is handled by a rights confirmation at checkout and by preferring blueprint plus tutor notes. Free-alternative objection is handled by selling the packaged artifact, with a fallback $49 thirty-question entry offer if ten prospects name price versus free tools as the blocker.
9. First 24 hours
Hours 1-2: domain, Next.js skeleton on Vercel, Supabase project. Hours 3-5: Stripe account verified, TrustMRR connected, three products created. Hours 6-8: blueprint schema and parser draft against one real past paper. Hours 9-11: 50 prospects from Google Maps with contact handles. Hours 12-14: one branded ten-question sample and the DM script, ready for day 9.
Revenue and profit figures in this report are hypotheses derived from stated assumptions. No customers, partnerships, sources or verified revenue exist at the time of writing.
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 →Kickoff Proof
Turns an agency's onboarding Loom or YouTube video into a branded, answer-gated certification page so a new client has to prove they watched before work starts.
Crash Course Press
Turns a licensing-exam instructor's own notes plus the official exam content outline into a branded, print-ready practice booklet and timed cohort quiz links in under 30 minutes.
SpeakGrade Report Cards
SpeakGrade turns two short phone recordings of a child reading aloud into a logo-branded, parent-ready spoken-English report card that shows exactly which words and habits improved since the last term.