ShelfShift
ShelfShift turns a DTC brand's Shopify, Amazon FBA and wholesale inventory exports into a 14-day 'Move / Markdown / Reorder' action plan with a dollar figure attached to every line.
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.
Founder-operator or ops lead of a DTC apparel, footwear, accessories or beauty brand doing roughly $50k-$500k/month, 2-50 employees, selling on Shopify plus at least one other channel (Faire wholesale, Amazon FBA, or 1-3 owned doors/pop-ups), based in the US, UK, CA or AU.
Market demand 5.6/10
Why this could work in 30 days: one founder can ship the MVP in 5 days; the plan targets a first sale on day 12; 7 sales at $149 clears $1,000 of collected revenue; strongest signal is speed to first revenue at 8.0/10.
Score and projections come from the generated report and are not guarantees. The $100K Hackathon is run by Ship or Die — read the official rules before you build.
Your first 24 hours
14 concrete actions. Tick them off as you go — nothing is saved or sent.
Product opportunity and positioning
Target customer
Founder-operator or ops lead of a DTC apparel, footwear, accessories or beauty brand doing roughly $50k-$500k/month, 2-50 employees, selling on Shopify plus at least one other channel (Faire wholesale, Amazon FBA, or 1-3 owned doors/pop-ups), based in the US, UK, CA or AU.
Why now
The contest window sits inside the single most decision-dense six weeks of the retail calendar: December sell-through triage and January clearance planning. The purchase is urgent this month without any manufactured urgency.
Problem
In peak season the founder has bestsellers sold out on the website while the same SKU sits at a pop-up or in FBA, dead stock eats cash going into January, and reorder/markdown calls get made from a pivot table on a Sunday night.
Solution
Upload three CSVs (Shopify variant export, FBA inventory/sales report, wholesale sheet), set a seasonality multiplier and per-channel transfer cost. ShelfShift normalises them, computes velocity and days-of-cover per SKU per channel, gates every candidate move on transfer economics, and returns a ranked 14-Day Move Plan as XLSX plus a one-page PDF.
ShelfShift — 30-Day Launch Report
Contest window: Monday 16 November 2026 13:00 UTC to Wednesday 16 December 2026 13:00 UTC. One founder, no team, no funding, no brand. Every revenue and profit figure in this report is a projection derived from stated assumptions — not verified revenue, not existing customers, not traction.
1. Product opportunity and positioning
The buying moment is concrete. A DTC apparel or beauty brand doing $50k–$500k a month sells on Shopify plus at least one other channel — Faire wholesale, Amazon FBA, or a pop-up. In the last three weeks of December those channels disagree with each other: a bestseller is sold out on the site while the same SKU sits in a 3PL or a pop-up rack, and slow stock is quietly becoming January's cash problem. The founder resolves this with three exports, a pivot table and a Sunday night.
The expensive errors in that session are not exotic. They are shipping units to a channel where the recovered margin never covers freight, handling and channel fees; and marking down a SKU that would have sold at full price in a different channel. ShelfShift is built around exactly those two errors. Every candidate action passes a transfer-economics gate — margin recovered minus transfer, handling and channel fee must be positive — and anything that fails is relabelled "liquidate, don't move" rather than quietly recommended.
The positioning against enterprise retail AI is delivery model, not features. Blue Yonder, RELEX, Lokad and Manhattan are integration projects with procurement cycles. ShelfShift is an uploaded file and a finished document. No OAuth, no Shopify app review, no implementation, no forecasting model to defend. That is a deliberate downgrade of ambition in exchange for speed: the buyer keeps full control of the decision and pays $149 to stop guessing.
ShelfShift is a new product, not a relaunch. It inherits the problem space of cross-channel inventory placement for lifestyle retailers and inverted the delivery model into a self-serve artifact.
2. Smallest sellable MVP
The sellable unit is one 14-Day Move Plan for $149. The minimum that justifies that price: a column mapper that accepts three messy CSV shapes, a velocity and days-of-cover calculation per SKU per channel, the transfer-economics gate, four action buckets with a dollar impact per line, and an XLSX plus one-page PDF the founder can act on the same day.
Anything that does not help produce or sell that document is out of scope for 30 days: no live Shopify or Amazon OAuth sync, no ML demand forecasting, no purchase orders, no supplier emails, no ERP/WMS integrations, no team roles, no approval workflows, no mobile app, no pooled benchmark data. The engine is deterministic rules plus a user-set seasonality multiplier. One LLM call writes a plain-English summary paragraph; if it fails, the plan still ships.
3. Technical implementation plan
Next.js on Vercel, Supabase for Postgres, file storage and magic-link auth, a TypeScript calculation engine as a serverless function, SheetJS for XLSX, React-PDF for the one-pager, Stripe Checkout, Resend for transactional email, and a single OpenAI call for the summary paragraph.
The build order is: schema and upload first, mapper second, engine third, export and payment fourth. Mapping is the hard part and gets the most time. Column-name guessing plus a manual remap screen for the fields the parser cannot infer is cheaper than trying to support every 3PL's export dialect. Everything runs synchronously; a plan must complete in under two minutes so no queue, no worker and no on-call rotation is needed.
The one thing most likely to break is mapping variance — a Faire wholesale sheet with merged headers, an FBA report with reserved and inbound columns, a Shopify export with 400 variants. The mitigation is a visible "we mapped these columns, fix anything wrong" step before generation, and a hard cap of three supported source shapes for the first 30 days.
4. Pricing and monetization
Three SKUs, all self-serve. Single Move Plan at $149 one-time is the default purchase. Peak + Clearance at $349 one-time sells three plans across six weeks plus a January markdown-and-liquidate calendar — it is the higher-intent offer and should be pushed to anyone who opens the sample twice. Weekly Rebalance at $99/month exists for brands that re-upload every Monday; it is a retention product, not a launch product, and should not be sold hard before the one-time offer converts.
The offer this week is the free 20-SKU sample plan: it shows the total dollar impact at stake across twenty real SKUs from the buyer's own export, and the full file, XLSX export and re-upload diff unlock at $149. Delivery cost per plan is targeted under $1.50, so gross margin on every sale is effectively the processing fee. Nothing is priced below cost.
5. First-customer acquisition
The only channel is direct cold email and Instagram DM to founders of DTC apparel, footwear, accessories and beauty brands, sourced from public Faire brand pages and Shopify storefronts that visibly show sold-out bestsellers. Target 300 brands on the list, 40–80 sends a day from one inbox, two follow-ups at day 3 and day 6, and a short screen recording of their own sample plan for anyone who opens twice.
Week one goal is not revenue, it is signal: 120 brands contacted, at least 8 sample-plan requests, and at least 1 paid plan. Mid-December the pitch pivots from restock to the January clearance calendar against the same list — which SKUs to hold, markdown or liquidate in the first three weeks of January.
6. 30-day execution roadmap
Days 1–5 are build-only: schema, mapper, engine, export, Stripe, launch page. Days 6–11 are list-building and first outreach while the mapper is still being patched from real files. Days 12–18 are the conversion push and the first paid plans, with the re-upload diff added once a paying customer asks for it. Days 19–26 pivot the same message to January clearance and push the $349 bundle to sample users. Days 27–30 are closing, reconciliation and a final send before the 13:00 UTC cutoff on 16 December. The day-by-day breakdown is in the roadmap section of this report.
7. Revenue and profit scenarios
Three projections: conservative 4 customers at $149 blended, base 14 at $165, upside 40 at $175. Base implies roughly $2,310 gross and about $2,057 hackathon profit after refunds and processing fees. Upside of 40 paid plans is a stretch and is well below the concept's own 70-plan target, which should be treated as an unreached ceiling, not a plan.
Processing is modelled at 4.5% blended (Stripe 2.9% + $0.30 domestic, more on international cards and Stripe Tax). Ad spend is zero in conservative and base because the channel is organic; the upside case includes $300 for a sponsored newsletter placement, labelled as optional. Operating cost covers Vercel, Supabase, domain, Resend, SheetJS/PDF rendering and inference, and is not deducted from hackathon profit but is deducted from business profit.
8. Risks and validation experiments
The four risks that can end this: mapping breaks on real exports; the free sample is judged in its first twenty rows and loses the sale; buyers decide a pivot table is good enough; and post-peak December kills urgency. Each has a cheap test this week — parse five real exports before writing the UI, run the sample on two friendly founders and time how long they look at it, ask five non-buyers what the pivot table already gives them, and A/B the restock angle against the clearance angle on the same list.
Kill criteria are explicit: if 120 contacts produce fewer than 8 sample requests, the wedge is wrong. If 20 sample plans produce zero paid plans, the output is not worth $149.
9. First 24-hour action plan
Hours 1–2 register the domain and deploy the empty Next.js shell with Supabase and Stripe test keys. Hours 3–5 lock the unified schema and build the upload screen. Hours 6–8 write the mapper against three synthetic exports. Hours 9–11 build velocity and days-of-cover plus the transfer-economics gate. Hours 12–14 assemble the first 40-brand outreach list from public Faire pages and Shopify storefronts.
Smallest sellable MVP
A paid $149 upload-to-plan flow: three CSVs in, a ranked 14-Day Move Plan out as XLSX plus a one-page PDF, with a free 20-SKU sample gating the full-file unlock.
Must-have features
- CSV mapper accepting Shopify variant exports, Amazon FBA inventory/sales reports and a generic wholesale sheet, normalised to one schema
- Velocity and days-of-cover per SKU per channel with a user-set peak-season multiplier
- Transfer-economics gate that rejects any MOVE whose recovered margin does not beat transfer, handling and channel fees
- Action classifier producing MOVE, REPLENISH, MARKDOWN-LIQUIDATE and HOLD buckets with a dollar impact per line
- Ranked top-10 actions page
- XLSX full SKU table plus one-page PDF summary, generated in under two minutes
- Stripe Checkout with magic-link access and a read-only share link
- Free 20-SKU sample plan shown before payment
Explicitly out of scope
- No live Shopify or Amazon OAuth sync
- No ML demand forecasting
- No purchase orders, supplier emails or ERP/WMS integrations
- No team roles or approval workflows
- No mobile app, no 24/7 support promise
- No pooled benchmark data across customers
- No multi-currency or multi-entity accounting
Demo flow (what the buyer sees)
- Founder uploads three CSVs to the upload screen
- ShelfShift guesses column mappings and shows a remap screen for anything unclear
- Founder sets peak-season multiplier and per-channel transfer cost
- Engine computes velocity, days-of-cover and candidate actions per SKU per channel
- Transfer-economics gate filters or relabels every MOVE
- Free 20-SKU sample renders with total dollar impact at stake
- Founder pays $149 via Stripe Checkout
- Full XLSX and one-page PDF generate server-side and are emailed plus shown on a share link
Technical implementation plan
Stack
Architecture
A single Next.js app on Vercel holds the upload UI, the remap screen and the results page. Files land in Supabase Storage and the parsed rows go into Postgres keyed by plan id. The calculation engine is a deterministic TypeScript function invoked synchronously after mapping; it returns buckets, dollar impacts and a ranked top-10. Export runs in the same request using SheetJS and React-PDF, writing the artifact back to storage. Stripe Checkout webhooks flip a plan from sample to unlocked, and Resend emails the share link. There is no queue, no worker and no scheduled job.
Build steps
- Create Next.js repo, Supabase project, schema for plans, skus, channels and actions
- Build CSV upload plus a header-sniffing mapper with a manual remap fallback
- Normalise three source shapes into one SKU-per-channel row model
- Implement velocity, days-of-cover and the peak-season multiplier
- Implement the transfer-economics gate and the four-bucket classifier with dollar impact per line
- Rank actions and render the top-10 page
- Generate XLSX via SheetJS and the one-page PDF via React-PDF
- Wire Stripe Checkout, webhook and magic-link access
- Email the share link via Resend and render a read-only plan view
- Add the re-upload diff comparing two plan versions
- Patch mappings against real exports from sample requests
- Deploy to a custom domain and run the full flow end to end
AI components
- One OpenAI call generates a plain-English summary paragraph per plan; it is cosmetic and the plan renders correctly if the call fails
Pricing and monetization
Single Move Plan
Entry offerFor Founder of a 2-3 channel DTC brand resolving one decision cycle now
- One 14-Day Move Plan, unlimited SKUs
- Ranked top-10 actions with dollar impact per line
- XLSX full SKU table plus one-page PDF
- Read-only share link for a co-founder or 3PL
Peak + Clearance
For Brand heading into January with cash tied up in slow stock
- Three plans across six weeks
- January markdown-and-liquidate calendar with per-SKU timing
- Re-upload diff between each plan
Weekly Rebalance
For Brand that re-uploads every Monday through peak and clearance
- Weekly regenerated Move Plan
- Diff alerts when a SKU crosses stockout or dead-stock thresholds
- Cancel anytime, no support SLA
Model: Self-serve one-time plan purchases with an optional monthly rebalance subscription.
First revenue offer: Free 20-SKU sample plan from the buyer's own export showing the total dollar impact at stake; the full XLSX, PDF and re-upload diff unlock at $149.
Expected ACV: $165
Upsell path: Sample users who open twice are offered the $349 Peak + Clearance bundle; one-time buyers are offered Weekly Rebalance at $99/month once a second plan proves the diff is useful.
First-customer acquisition
Direct cold email and Instagram DM to DTC founder-operators sourced from public Faire brand pages and Shopify storefronts with visible sold-out bestsellers.
Where to find them
- Public Faire brand directory pages for apparel, accessories and beauty
- Shopify storefronts showing 'sold out' on a listed bestseller
- Shopify 'About' and contact pages for founder name and email
- Instagram brand accounts with an active founder presence
- DTC brand roundups and newsletters that link public stores
- Public pop-up and market vendor lists for brands with owned doors
Exact outreach motions
- Cold email naming a specific sold-out SKU and the channel where it is still listed
- Instagram DM to the founder with the same SKU-specific hook, one sentence
- Day-3 follow-up with a different SKU found on a second channel
- Day-6 follow-up offering the sample plan as a no-commitment check
- Short screen recording of their own 20-SKU sample for anyone who opened twice
- Mid-December re-pitch on the January clearance calendar to the same list
- Warm intro request to every paying customer, logged as a referral source
Message template
Hi {first_name} — you're sold out on {sku} on your site while it's still showing stock on {channel}. I'll run a free 20-SKU reallocation check on your Shopify and FBA export, no integration, and send back a one-page plan showing what's movable and what's worth marking down. Want me to run it?30-day execution roadmap
Week 1 · Build & validate · 7 days
Repo, infra and data model
- Create Next.js repo, Supabase project and Vercel deployment
- Define plans, skus, channels and actions tables
- Provision domain and Stripe test keys
Deliverable: Deployed skeleton on a custom domain with schema applied
Upload and CSV mapping
- Build three-file upload to Supabase Storage
- Write header-sniffing mapper for Shopify, FBA and wholesale shapes
- Build manual remap screen for unmapped columns
Deliverable: Three messy CSVs normalised into one row model
Velocity, cover and the transfer gate
- Implement velocity and days-of-cover per SKU per channel
- Implement peak-season multiplier and transfer-cost inputs
- Implement the transfer-economics gate and four-bucket classifier
Deliverable: Deterministic engine passing tests on three synthetic datasets
Exports, checkout and access
- Generate XLSX via SheetJS and one-page PDF via React-PDF
- Wire Stripe Checkout, webhook and magic-link unlock
- Render the free 20-SKU sample and the top-10 page
Deliverable: End-to-end paid flow working in Stripe test mode
Go live and open the list
- Switch Stripe live, run a full purchase and refund test
- Publish the launch page with a working sample upload
- Assemble the first 40-brand outreach list
Deliverable: Live public URL and 40 qualified brands on the list
First send
- Send 40 SKU-specific cold emails
- Build the list from 40 to 120 brands
- Fulfil any sample requests and log mapping failures
Deliverable: 40 emails sent, list at 120, first sample plans delivered
Fix the mapper from real files
- Patch mapping failures found in real exports
- Send a second batch of 40 emails
- Review open rates by subject line
Deliverable: Mapper covering all observed export shapes, 80 total sends
Week 2 · Launch & first sale · 7 days
Second touch and DMs
- Send day-3 follow-ups to the first 80
- Send 20 Instagram DMs to founders
- Add a different SKU hook to every follow-up
Deliverable: 100 touches delivered, DM replies logged
Screen recordings
- Record a short screen capture for every contact who opened twice
- Send 60 new cold emails
- Tighten the sample plan's first page
Deliverable: First recordings sent, 140 total emails
Re-upload diff v1
- Ship the diff view comparing two plan versions
- Add the read-only share link page
- Fix the top three sample-quality complaints
Deliverable: Diff and share link live in production
List scale and second follow-up
- Send 60 new cold emails
- Send day-6 follow-ups to the first 80
- Ask two sample users what would make them pay today
Deliverable: 140 total sends, 20 sample plans delivered
Close the warmest leads
- Send a same-day offer to every opened-twice contact
- Push the $149 single plan with a share link in the email
- Fix any checkout friction found live
Deliverable: Target: first paid plan collected through Stripe
Turnaround and polish
- Verify every plan generates in under two minutes
- Email the paying customer for one warm intro
- Log the referral source in a spreadsheet
Deliverable: Sub-2-minute generation confirmed, first intro request sent
Two-week readout
- Compute emails sent, opens, sample requests and paid plans
- Keep the best-performing subject line and drop the rest
- Rebuild the list with 150 fresh brands
Deliverable: Two-week funnel readout and a fresh 150-brand list
Week 3 · Repeat & systematize · 7 days
January clearance angle
- Publish a clearance-focused landing variant
- Rewrite the email hook around January markdown timing
- Add a markdown calendar section to the PDF
Deliverable: Second landing page and clearance email template live
Clearance pitch
- Send 80 clearance-angle emails
- Re-pitch the first 80 contacts with the new angle
- Log which angle gets more replies
Deliverable: 160 clearance-angle sends, angle comparison logged
Instagram wave
- Send 30 founder DMs with the clearance hook
- Follow up on unanswered DMs once
- Collect any objections verbatim
Deliverable: 30 DMs sent, objection list written
Second plan and refund check
- Offer a second plan to existing buyers
- Check refund requests and log reasons
- Add a wholesale multi-door channel template
Deliverable: Multi-door template shipped, refund log current
Push the $349 offer
- Email the Peak + Clearance bundle to every sample user
- Add a bundle comparison block to the results page
- Send 60 new cold emails
Deliverable: Bundle offer sent to the full sample list
Public example plan
- Publish a synthetic example plan as a public page
- Link it from every email footer
- Send 40 new cold emails
Deliverable: Public example plan live and linked
Three-week readout
- Compute sample-to-paid conversion
- Kill any channel step with zero replies
- Hand-pick the 30 warmest untouched brands
Deliverable: Conversion number and a 30-brand warm list
Week 4 · Scale what works · 7 days
Existing contacts
- Send the bundle to everyone who paid for the single plan
- Send 60 new cold emails
- Request a one-line testimonial from each paying customer
Deliverable: Upsell sent, testimonial requests logged
Final list push
- Send 80 cold emails to new brands
- Send 30 Instagram DMs
- Follow up on all day-3 and day-6 threads
Deliverable: Full list contacted at least once
Sample-to-paid lift
- Rewrite the sample email to lead with the dollar total
- Reduce sample generation time to under 60 seconds
- Fix the two most common mapping complaints
Deliverable: New sample email and faster generation shipped
Real proof only
- Publish only real customer quotes collected so far
- Confirm no fabricated logos, counts or statistics on the site
- Check pricing page and checkout copy for accuracy
Deliverable: Site audit passed with zero unverified claims
Contest end framing
- Send a last-call email tied to the 16 December cutoff
- Offer same-day plan delivery on every reply
- Send the bundle to all openers who have not bought
Deliverable: Last-call campaign sent to the full list
Convert open threads
- Personally reply to every unanswered thread
- Send a final follow-up with the share link attached
- Log all payment links sent
Deliverable: Every warm thread answered
Deliver every paid plan
- Confirm all paid plans generated and emailed
- Fix any delivery failures
- Record every sale date inside the contest window
Deliverable: Zero undelivered paid plans
Days 29-30 · Consolidate · 2 days
Money in
- Reconcile Stripe payouts against TrustMRR records
- Record refunds, chargebacks and processing fees
- Compute hackathon profit and business profit separately
Deliverable: Final reconciliation sheet
Last push before 13:00 UTC
- Send the final email before 12:00 UTC on 16 December
- Verify no payments land outside the window
- Write the post-contest readout
Deliverable: Final submissions complete and readout written
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: 120 brands contacted, roughly 8 sample requests, 4 paid plans — the low end of the outreach funnel.
- Base: roughly 20 sample plans delivered at an 8-12% sample-to-paid rate implied by the concept, landing 14 paid plans.
- Upside: 350 brands contacted and 55 sample plans, converting to 40 paid plans — a stretch, still below the concept's 70-plan ceiling.
- ARPU is blended across the $149 single plan and the $349 bundle: $149 conservative, $165 base, $175 upside.
- Payment processing modelled at 4.5% blended (Stripe 2.9% + $0.30 domestic, higher on international cards and Stripe Tax).
- Ad spend is $0 in conservative and base because acquisition is organic cold email and DM; upside carries $300 for one optional sponsored newsletter placement.
- Refunds assumed at one plan in base and upside, zero in conservative; chargebacks zero throughout.
- Operating cost covers Vercel, Supabase, domain, Resend, PDF rendering and LLM inference, and is not deducted from hackathon profit.
- Per-plan delivery cost target is under $1.50 against a $149 price, so no sale is below cost.
- All figures are projections from these assumptions, not verified revenue.
Conservative
4 customers- Refunds
- $0
- Chargebacks
- $0
- Payment fees
- $27
- Paid marketing
- $0
- Hackathon profit
- $569
- Operating costs
- $60
- Business profit
- $509
Base case
14 customers- Refunds
- $149
- Chargebacks
- $0
- Payment fees
- $104
- Paid marketing
- $0
- Hackathon profit
- $2,057
- Operating costs
- $90
- Business profit
- $1,967
Optimistic
40 customers- Refunds
- $149
- Chargebacks
- $0
- Payment fees
- $315
- Paid marketing
- $300
- Hackathon profit
- $6,236
- Operating costs
- $160
- Business profit
- $6,076
Model your own numbers
Runs in your browser · nothing is saved- Collected revenue
- $2,086
- Payment fees (3.9%)
- $81
- Paid marketing
- $0
- Hackathon profit
- $2,005
- Business profit
- $1,915
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.
All scenarios are projections built from the stated assumptions. They are not verified revenue, not existing customers, and not a forecast of contest ranking.
Main risks and validation experiments
Risk 1CSV mapping breaks on real exports, consuming build and support time.
Ship a visible remap screen and cap support at three source shapes for 30 days.
Parse five real exports from sample requests before building the results UI.
If more than half of real uploads need manual column fixing after 10 submissions, narrow to Shopify-only input.
Risk 2Buyers judge the free 20-SKU sample worthless and never reach checkout.
Lead the sample with the total dollar impact at stake, not the SKU table.
Send 20 sample plans and time how long each recipient stays on the page.
20 sample plans with zero paid unlocks means the output is not worth $149.
Risk 3Founders decide their existing pivot table is good enough.
Name the specific transfer-economics error their pivot table cannot catch.
Ask five non-buyers what their current process already gives them.
If four of five report no painful decision last month, the wedge is wrong.
Risk 4Mid-December post-peak lull removes urgency from the restock pitch.
Pivot the same list to the January clearance calendar from day 15.
Split the list between restock and clearance angles on the same week.
If the clearance angle produces no replies across 160 sends, stop the pivot.
Risk 5Cold-email deliverability and DM limits cap reach for a solo sender.
Warm the domain before day 6, cap sends per day, vary subject lines.
Track open rate on the first 80 sends split across two subject lines.
Under 15% open rate on 120 sends means the domain or copy is dead; rebuild the sending setup.
Risk 6Recommendations look generic, triggering refunds and no referrals.
Require a dollar figure and a channel per line, and gate every MOVE on transfer economics.
Read every refund reason and every reply verbatim for two weeks.
More than 10% refund rate on 10 or more sales means the engine is not credible.
Hour-by-hour: the first 24 hours
Hours 1-2
- Register the domain and create the Next.js repo
- Connect Supabase and Vercel, deploy an empty shell
- Create Stripe test keys and a Resend account
Output: Live skeleton URL on a custom domain
Hours 3-5
- Write the unified schema: sku, channel, on_hand, units_sold_30/60/90d, unit_cost, price, channel_fee_pct
- Apply the migration and seed three synthetic exports
Output: Database schema applied with test data
Hours 6-8
- Build the three-file upload screen
- Write header-sniffing for Shopify, FBA and wholesale shapes
- Add a manual remap fallback
Output: Mapper accepting three CSV shapes end to end
Hours 9-11
- Implement velocity and days-of-cover per SKU per channel
- Implement the peak-season multiplier input
- Implement the transfer-economics gate and four-bucket classifier
Output: Calculation engine returning bucketed actions with dollar impact
Hours 12-14
- Collect 40 brands from public Faire pages and Shopify storefronts with visible sold-out SKUs
- Record founder name, email and the specific SKU to open with
- Draft the four-line email template
Output: A 40-row outreach list with a SKU-specific hook per row
Full 30-day launch report
Read the full generated report
ShelfShift — 30-Day Launch Report
Contest window: Monday 16 November 2026 13:00 UTC to Wednesday 16 December 2026 13:00 UTC. One founder, no team, no funding, no brand. Every revenue and profit figure below is a projection derived from stated assumptions — not verified revenue, not existing customers, not traction.
1. Product opportunity and positioning
The buying moment is concrete. A DTC apparel or beauty brand doing $50k–$500k a month sells on Shopify plus at least one other channel — Faire wholesale, Amazon FBA, or a pop-up. In the last three weeks of December those channels disagree with each other: a bestseller is sold out on the site while the same SKU sits in a 3PL or on a pop-up rack, and slow stock is quietly becoming January's cash problem. The founder resolves this with three exports, a pivot table, and a Sunday night.
The expensive errors in that session are not exotic. They are shipping units to a channel where the recovered margin never covers freight, handling and channel fees, and marking down a SKU that would have sold at full price somewhere else. ShelfShift is built around exactly those two errors. Every candidate action passes a transfer-economics gate — margin recovered minus transfer, handling and channel fee must be positive — and anything that fails is relabelled "liquidate, don't move" rather than quietly recommended.
The positioning against enterprise retail AI is a delivery-model argument, not a feature argument. Blue Yonder, RELEX, Lokad and Manhattan are integration projects with procurement cycles. ShelfShift is an uploaded file and a finished document. No OAuth, no app review, no implementation, no forecasting model to defend. That is a deliberate downgrade of ambition in exchange for speed: the buyer keeps control of the decision and pays $149 to stop guessing on a Sunday night.
ShelfShift is a new product, not a relaunch. It inherits the problem space — cross-channel inventory placement for lifestyle retailers — and inverts the delivery model into a self-serve artifact anyone with three CSV exports can buy without a call.
2. Smallest sellable MVP
The sellable unit is one 14-Day Move Plan for $149. The minimum that justifies that price: a column mapper that accepts three messy CSV shapes, velocity and days-of-cover per SKU per channel, the transfer-economics gate, four action buckets with a dollar impact per line, and an XLSX plus one-page PDF the founder can act on the same afternoon.
Everything that does not help produce or sell that document is out of scope for 30 days: no live Shopify or Amazon OAuth sync, no ML demand forecasting, no purchase orders, no supplier emails, no ERP/WMS integrations, no team roles, no approval workflows, no mobile app, no pooled benchmark data. The engine is deterministic rules plus a user-set seasonality multiplier. One LLM call writes a plain-English summary paragraph; if it fails, the plan still ships.
3. Technical implementation plan
Next.js on Vercel, Supabase for Postgres, file storage and magic-link auth, a TypeScript calculation engine as a serverless function, SheetJS for XLSX, React-PDF for the one-pager, Stripe Checkout, Resend for transactional email, and a single OpenAI call for the summary paragraph.
The build order is schema and upload first, mapper second, engine third, export and payment fourth. Mapping gets the most time. Column-name guessing plus a manual remap screen is cheaper and more robust than trying to support every 3PL dialect. Everything runs synchronously; a plan must finish in under two minutes so there is no queue, no worker and no on-call rotation.
The thing most likely to break is mapping variance — a wholesale sheet with merged headers, an FBA report with reserved and inbound columns, a Shopify export with 400 variants. The mitigation is a visible "we mapped these columns, fix anything wrong" step before generation, and a hard cap of three supported source shapes for the first 30 days.
4. Pricing and monetization
Three SKUs, all self-serve. Single Move Plan at $149 one-time is the default. Peak + Clearance at $349 one-time sells three plans across six weeks plus a January markdown-and-liquidate calendar; it is the higher-intent offer and should be pushed to anyone who opens the sample twice. Weekly Rebalance at $99/month exists for brands that re-upload every Monday — it is a retention product, not a launch product.
The offer this week is the free 20-SKU sample plan, showing the dollar impact at stake across twenty real SKUs from the buyer's own export. The full file, XLSX export and re-upload diff unlock at $149. Target delivery cost is under $1.50 per plan, so gross margin is effectively the processing fee. Nothing is sold below cost.
5. First-customer acquisition
The only channel is direct cold email and Instagram DM to founders of DTC apparel, footwear, accessories and beauty brands, sourced from public Faire brand pages and Shopify storefronts that visibly show sold-out bestsellers. Build a 300-brand list, send 40–80 a day from one warmed inbox, follow up at day 3 and day 6, and record a short screen capture of their own sample plan for anyone who opens twice.
Week one is about signal, not revenue: 120 brands contacted, 8 sample-plan requests, 1 paid plan. From mid-December the pitch pivots against the same list to the January clearance calendar — which SKUs to hold, mark down or liquidate in the first three weeks of January.
6. 30-day execution roadmap
Days 1–5 are build-only: schema, mapper, engine, export, Stripe, launch page. Days 6–11 are list-building and first outreach while the mapper is still being patched from real files. Days 12–18 are the conversion push and the first paid plans, with the re-upload diff added once a paying customer asks for it. Days 19–26 pivot the same message to January clearance and push the $349 bundle to sample users. Days 27–30 are closing, fulfilment, reconciliation and a final send before the 13:00 UTC cutoff on 16 December. The day-by-day breakdown sits in the roadmap section of this report; treat day 12 as the first-sale target and day 5 as the hard MVP deadline.
7. Revenue and profit scenarios
Three projections: conservative 4 customers at a $149 blended ARPU, base 14 at $165, upside 40 at $175. Base implies roughly $2,310 gross and about $2,057 hackathon profit after one refund and processing fees. Upside of 40 paid plans is a stretch and sits well below the concept's own 70-plan target, which should be treated as an unreached ceiling rather than a plan.
Processing is modelled at 4.5% blended — Stripe at 2.9% + $0.30 domestic, higher on international cards and Stripe Tax. Ad spend is zero in conservative and base because the channel is organic; upside carries $300 for one optional sponsored newsletter placement. Operating cost covers Vercel, Supabase, domain, Resend, PDF rendering and inference; it is not deducted from hackathon profit, but it is deducted from business profit. Two different numbers therefore exist and should never be blended.
8. Risks and validation experiments
Four risks can end this. Mapping breaks on real exports — test five real files before writing the results UI. The free sample is judged in its first twenty rows — put the total dollar impact at the top, not the SKU table. Buyers decide a pivot table is good enough — ask five non-buyers what their current process already gives them. Post-peak December kills urgency — split the list between restock and clearance angles in the same week and compare replies.
Kill criteria are explicit. If 120 contacts produce fewer than 8 sample requests, the wedge is wrong. If 20 sample plans produce zero paid plans, the output is not worth $149. If refunds exceed 10% on 10 or more sales, the engine is not credible and the pricing has to change before more outreach.
9. First 24-hour action plan
Hours 1–2 register the domain, deploy the empty Next.js shell, connect Supabase and Stripe test keys. Hours 3–5 lock the unified schema and seed three synthetic exports. Hours 6–8 build the upload screen and mapper. Hours 9–11 build velocity, days-of-cover and the transfer-economics gate. Hours 12–14 assemble the first 40-brand outreach list with a SKU-specific hook per row. By the end of day one there is a live URL, a working mapper and a list to send to.
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 →Idle Hours GPU Waste Autopsy
Upload your cloud/GPU billing export and get a dollar-quantified PDF autopsy of exactly where your GPU spend is being wasted, ranked by recoverable dollars per month.
LogSlim Audit Pack
A browser-local tool that turns one 24-hour log sample into a paste-ready exclusion config bundle and a projected monthly savings report for teams overpaying for Datadog, CloudWatch or Elastic log volume.
Delta Memo
Delta Memo sends Shopify store operators a one-page morning memo that explains exactly which channel, SKU, discount or cohort caused yesterday's revenue swing, with the numbers computed first and the prose written second.