Twelve weeks. Pillar by pillar.
The full Moonlabs Academy curriculum. Every week broken into Coding, Commercials, Investment — what is taught, what is shipped, and the deliverable that goes into your operator portfolio.
This page is the full reference. No marketing fluff. If you are deciding whether to apply, read this. If you are a parent or a sponsor evaluating where the £6,000 goes, this is the answer.
The structure: every week has a goal, three pillars (Coding, Commercials, Investment) and a Friday-afternoon deliverable. Students do not move on until the deliverable is shipped and reviewed. Roughly 25–30 hours of contact time per week.
The wedge
A defensible one-line description of what you are building, for whom, and why it has to exist now.
Cursor + Claude Code setup. Production-grade Laravel/Next.js scaffold. Postgres + pgvector. Auth with Clerk or native. Vercel/Fly.io deploy. Sentry from day one.
A deployed scaffold, accessible on a real domain, hooked up to logging and Sentry. URL goes in the cohort tracker.
Wedge framing. The five-sentence wedge document. The first-person buyer description. Pricing intuitions — three reference points from comparable companies.
A one-page wedge document defended in front of the cohort on Friday. Three named candidate buyers.
How early-stage rounds actually work in the UK in 2026. SAFEs vs priced rounds. Cap table mechanics from scratch. The shape of an angel relationship.
A blank cap table opened in your tracker, ready to populate from week two onwards.
First shipped thing
A working, publicly accessible piece of your product live by Friday at 4pm.
Schema design with a long memory. Migrations that you will not regret in six months. Job queues. The deploy pipeline you actually want, not the one tutorials teach.
A working slice of the product live. Anyone with the URL can use it. Real, not staging.
The landing page as a sales document. What a sales-led landing page says vs a marketing one. Buyer-first copy. The five-line "what" / "for whom" / "why" / "evidence" / "ask" framework.
A buyer-focused landing page (no fluff sections) live on your domain.
Founder vesting and cliff. The one-page founder agreement. Mistakes that lock you into a bad cap table for years.
A founder agreement template adapted to your situation, ready to sign if you have a co-founder.
Five customer conversations
Five real buyer conversations completed by Friday, with notes good enough for week-four decisions.
Product analytics — PostHog or Plausible, set up properly. Event taxonomy that scales. The dashboard that tells you the truth.
Three named events instrumented and on a dashboard the whole cohort can see.
Discovery call structure. The five-question script. What to record. Anti-patterns (leading questions, confirmation bias, talking too much). Real call recordings reviewed in cohort.
Five completed discovery calls, recorded with permission, with summary notes uploaded.
Market sizing without bullshit. Bottoms-up TAM models. The honest-numbers approach that investors actually believe.
A bottoms-up market sizing for your wedge — one page, defended in cohort.
Reprice the wedge
A pricing experiment shipped against real prospective buyers, with results captured by Friday.
AI features as product surfaces — embeddings, RAG, eval suites. Prompt versioning. The repo convention we use across the portfolio.
One AI feature with a gold-set eval suite (10+ cases) and a documented prompt v1.
Pricing experiment design. Price points calibrated to reference customers. The one-week landing-page price test. Conversion-rate reading.
A pricing test live on the landing page, with the result documented.
Dilution modelling — what a £200k SAFE at £4m cap looks like four rounds in. The cap-table-projection exercise.
A dilution model for your company over the next three rounds, defended in cohort.
The deck, v1
A ten-slide deck defended in front of the cohort on Friday. No theatre, no growth charts off three data points.
Production hardening — error states, edge cases, rate limits, abuse handling. The "imagine 1,000 users tomorrow" exercise.
A hardened product capable of taking real traffic without falling over.
The elevator pitch. The thirty-second version, the two-minute version, the ten-minute version. Body language. Voice. The anti-pitch (everything you should NOT do).
Three versions of the pitch, rehearsed in front of the cohort and recorded.
Slide-by-slide deck construction. The eight slides that actually matter. Real decks deconstructed (raised vs not raised, side by side).
Deck v1 — ten slides, exported as PDF, sitting in the data room.
First paid commitment
A signed pilot agreement, paid trial, or LOI from a named customer by Friday.
Auth at the product edges. Billing — Stripe Checkout vs custom. Subscription mechanics. The first invoice rendered.
A working billing path. Someone can pay you. If they have, even better.
Pilot contract language. The two-page pilot agreement (we share the template). Pricing letters. The art of asking for the money.
A signed pilot, LOI, or paid trial from a named customer.
SAFE vs priced equity. When each makes sense. The mechanics of an angel cheque actually clearing.
A SAFE template populated with your terms, ready to issue to a first angel.
Mid-course honesty
A reset week. Refactor the codebase, rewrite the positioning, sharpen what is working.
The refactor: technical debt audit. Test coverage where it matters. The thirty-minute hot-path optimisation.
A codebase you would not be embarrassed to hand a senior engineer for review.
Positioning rewrite with real customer data. Repositioning vs pivoting. What earned conviction looks like.
Wedge document v2, materially different from v1 if the data demands it.
The data room — what goes in, what stays out. The folder convention we use. The two-line covering note.
A populated data room ready for diligence.
Demo day, first pass
A complete demo-day-quality pitch run in front of three independent operators.
Demo reliability — local vs hosted demo, fallback flows, the "WIFI is dying" rehearsal. The product walkthrough that lands in 90 seconds.
A 90-second product demo that works without internet, recorded as a backup.
The named-customer-reference call. How to secure one. What to ask for. How to use it on the pitch.
One named customer reference signed off, with permission to use in the pitch.
Financial model construction from our template. Stress-tests for the questions investors actually ask. The five-line P&L summary.
A financial model that has survived three operator stress-tests.
The cold campaign
A sharpened list of fifty plausible investors and a cold outreach campaign live by Friday.
Structured logs and observability. Sentry, Logflare, or equivalent. The dashboard you check every morning.
A morning-check dashboard with the five numbers you need to see daily.
Demand-generation playbook for sub-£1,000 ACV pilots. Outbound that does not feel like spam. Channel selection.
A documented playbook for your top channel, with the first batch of outreach sent.
Investor list construction. How to find the right fifty. Cold outreach written by hand. The follow-up cadence.
List of fifty named investors, twenty cold emails sent, ten warm intros queued up.
First investor calls
Five real investor calls completed by Friday, with notes captured.
Investor-readiness of the product itself — uptime SLA, the "is this real" test, the screenshare drill.
A product that holds up under live demo without theatre.
Objection handling drills. The twelve hardest questions investors ask. Practising the honest no-answer.
Twelve drilled responses captured in your own words.
Running the call itself. Setting agenda, controlling pace, asking for the close. The follow-up note that lands in 24 hours.
Five completed investor calls with follow-up notes sent.
Demo day rehearsal
A mock demo day in front of three independent operators, with written feedback acted on by Friday.
Final product polish. Loading states, empty states, the one-line tutorial. The thing your demo audience will notice in the first ten seconds.
Polished product, polished demo, polished landing page. All three reviewed.
Final pitch rehearsal. The fifteen-minute version. The cohort-internal critique session. The note pad after.
Pitch v3, rehearsed twice on Thursday and once on Friday.
Final deck and model audit. Diligence-ready data room. The "what happens after demo day" plan.
Deck v3, model v3, data room locked. Post-demo-day plan written down.
Demo day
Live demo day in front of an invited audience of investors, employers, and operators.
Demo-day game day. Backup plans for every failure mode. The "internet is down" version. The "presenter laptop dies" version.
A demo that works. Recordings of yours and others.
Follow-up management. The same-day thank-you. The 48-hour follow-up. The two-week cadence.
Follow-up cadence executed on the audience contacts that engaged.
Closing the round. Pricing the offer. The boring documents that actually move the money.
For a subset of the cohort: a closed round in the weeks after demo day.
Twelve artefacts in your operator portfolio.
- A deployed, monitored product on a real domain.
- A buyer-focused landing page that converts.
- Three event-instrumented dashboards.
- Five recorded discovery calls with summary notes.
- A wedge document, v2, rewritten with conversation data.
- A bottoms-up market sizing.
- A dilution model across three projected rounds.
- A ten-slide deck (v3) and a populated data room.
- A financial model that has survived three stress-tests.
- A signed pilot, LOI or paid trial.
- A SAFE template ready to issue.
- A pitch you can run cold, recorded multiple times.
You take these with you regardless of whether you raise. They are the operator portfolio you build in twelve weeks.
Summer 2026 cohort
Twelve weeks. Twelve seats.
Starts Monday 15 June 2026. Demo day Thursday 3 September 2026. Application deadline Friday 5 June.
Read the thinking.
All essaysYour next chapter starts here.
Tell us about the company you want to build. If we’re a fit, we’ll get back within a week.
© 2026 Moonlabs Incubator. All rights reserved.