Pre-flight prep
Startup Validation Checklist
This isn't a tool that scores your idea. The scoring happens in Launch Control. This is the prep list — what to bring to the session so the diagnosis you get out is sharp, not generic. Walk in with these seven inputs ready and you'll surface real gaps in 30 minutes.
This page is the prep list, not the diagnostic. The diagnostic — pressure-testing your idea against the seven pillars of launch readiness — happens inside Launch Control, the guided session that walks through 13 structured questions and tells you which pillars are weak.
What you do here is gather the inputs that make that session worth running. Walk in with sharp answers to the seven items below and you walk out with a sharp diagnosis. Walk in cold and you walk out with a clear homework list — both outcomes are useful, but the first one moves your idea forward faster.
How this list is organised
Seven items, one per pillar. Each item names the thing to bring, the format that works, and the failure mode if you skip it. None of these items require building the product. All of them require leaving the building.
1. Problem language, in customer words
Bring: two sentences describing the problem, written using language pulled from real customer conversations — not your pitch deck, not your imagination, not ChatGPT.
Format that works: direct quotes from at least five interviews. Phrases like "every Friday I spend three hours" are useful. Phrases like "we help teams streamline workflows" are not.
Why it matters: the moment your problem statement sounds like marketing copy, every downstream answer will sound like marketing copy too. Launch Control catches this early — but the fix is sitting in your interview notes, not in the session.
2. One named ICP, not a category
Bring: the name of one real person (or company) who fits your target customer. Their job title, company size, industry, and the situational trigger that makes them a buyer.
Format that works: "Heads of Finance at 50–200-person SaaS companies who just hired their first finance lead." If you cannot name an actual person who fits, the customer is still abstract.
Why it matters: "everyone" and "SMBs" and "founders" are not ICPs — they are categories. Launch Control will push your answer until it is specific enough to write a job ad for. Bringing one named person short-circuits that loop.
3. Behavioural demand evidence
Bring: anything that shows real behaviour — pre-orders with cards on file, paid pilots, signups with conversion data, repeat usage of a prototype.
Format that works: numbers. "Twelve people pre-ordered at £29." "Three pilots ran for four weeks; two extended." "Landing page converted at 6.2% from cold traffic."
Why it matters: demand signal is the heaviest pillar — the one most-cited in real failure cases — and stated interest does not count. If all you have is "people said they would use it," that is the work you need to go and do, not the work you bring.
4. The three closest substitutes
Bring: the three things customers use today instead of your product. One of those three is almost always "doing nothing," which counts.
Format that works: for each substitute, one sentence on why a customer would switch and one estimate of the cost of switching (time, money, retraining, data migration).
Why it matters: customers do not have a vacuum where your product goes. They have a workaround, however inadequate. If you cannot name the three substitutes, you cannot name what makes you better — and "better UI" is not the answer Launch Control accepts.
5. An honest execution plan
Bring: the technical critical path, the one piece you do not yet know how to build, and the budget plus timeline you actually believe (not the pitch-deck version).
Format that works: bullet points with rough effort estimates. Include any regulatory or platform-policy risks if they apply.
Why it matters: strong frontend teams shipping ML-systems-engineering products is a recurring failure pattern. Launch Control surfaces the mismatch — but it cannot read your team. The honest assessment has to come from you.
6. One primary channel with a CAC hypothesis
Bring: one acquisition channel (not a list of three "maybe" channels) and a back-of-envelope CAC estimate with the source of the number.
Format that works: "Cold outbound to FinOps leads on LinkedIn; CAC ~£250 based on 2% reply rate, 20% meeting-to-pilot conversion, 50% pilot-to-paid conversion." Numbers can be wrong; the discipline of writing them down is the point.
Why it matters: distribution is the second-most-cited failure mode, and "we'll figure it out" is not a strategy. Bringing a channel hypothesis means Launch Control can pressure-test it instead of telling you to come back when you have one.
7. A price point and payback math
Bring: what you plan to charge, the unit economics at that price, and the payback period at your assumed CAC.
Format that works: £X/seat/month, gross margin Y%, payback Z months. The numbers do not need to be defensible yet — they need to exist.
Why it matters: "we'll monetise later" is the line on a lot of failure post-mortems. Pricing is product. Launch Control checks whether the numbers survive contact with reality; you need numbers for that to happen.
What if you don't have most of these yet?
Run Launch Control anyway. The system is built to handle thin inputs — it will mark the underdone pillars as Insufficient and explain exactly what evidence would move them. That output is your homework list, and it is more useful than a generic "go validate" tip from anywhere else on the internet.
Most founders run Launch Control twice. The first run surfaces the gaps; the second run, two to four weeks later, confirms they have been closed. The system is designed for that loop.
What you'll get out of Launch Control
A signal-strength reading on each of the seven pillars (Insufficient, Emerging, Strong), a signal strength level for the idea overall (Insufficient, Weak, Emerging, Strong Validation), and a written diagnosis explaining which gaps are blocking the launch and what evidence would close them.
You will not get a number out of 100. You will not get "your idea is awesome." You will get the same kind of structured diagnosis a senior PM would give you in a discovery review — except it runs in 30 to 45 minutes, and it does not flatter you to keep you using the tool.
Frequently asked questions
What goes on a startup validation checklist?
The honest answer is two layers. The first layer is the prep work — customer interview notes in their language, one named ICP (ideal customer profile) person, behavioural evidence of demand, a list of substitutes, an execution plan, a primary channel with a CAC (customer acquisition cost) estimate, and pricing math. The second layer is the diagnostic itself, which is the seven pillars of launch readiness. The prep is what you bring; Launch Control is where the pressure-testing happens.
Do I need everything on the prep list before running Launch Control?
No. You can run Launch Control with whatever you have right now, even a one-line idea. The system will flag the inputs you are missing as Insufficient signals — that is the system telling you exactly what to go and gather. If you bring everything on the prep list, you get a sharper diagnosis on the first run. If you bring less, you get a clear list of homework instead.
How long does the prep take?
If you have already been building for a few weeks, an hour to organise what you have. If you are starting fresh, two to four weeks of customer conversations and channel tests is realistic. The prep itself is part of the validation — gathering it forces the conversations and experiments that turn opinions into evidence.
What is the difference between this and a generic startup checklist?
Most online startup checklists are tip lists ('make sure you have product-market fit'). This one is structured around the seven dimensions where startups actually fail, and it is designed as preparation for a diagnostic — not as the diagnostic itself. The thinking happens in Launch Control. The gathering happens here.
What if my idea is too early-stage to have most of this?
That is normal. Most ideas come in with strong material on Pillars 1 and 2 (problem, customer) and very little on Pillars 3 to 7 (demand, differentiation, execution, distribution, monetisation). Launch Control is built for that — it surfaces which pillars are early and tells you which evidence to go and collect first. Do not delay running it because the prep feels incomplete.
Stop reading. Start pressure-testing.
ReadySetLaunch's Launch Control walks you through thirteen structured questions across the seven pillars. Three free trial credits, no card required.
Start Launch Control