Andrew Crossley

    What framework should founders use?

    SHORT ANSWER

    Founders need one sequence, not a shelf of frameworks. A workable one: discover the problem, validate willingness to pay, prototype the flow, build the MVP, launch narrow, get first revenue, then scale what works. Borrow specific tools where they help — jobs-to-be-done for framing, opportunity solution trees for discovery — but a framework that produces artefacts instead of decisions is overhead.

    Answered by Andrew Crossley, Fractional Chief Product Officer · Updated 2026-08-01

    Why it matters

    Framework shopping is a common form of productive procrastination. It feels like strategy and produces no evidence.

    One shared sequence, on the other hand, makes it obvious which stage you are in and what would let you move on.

    How it works in practice

    1. 1

      Discover

      Interviews about past behaviour with a specific segment. Output: one clearly stated problem.

    2. 2

      Validate

      A commitment signal — deposit, pilot, letter of intent. Output: evidence someone will pay.

    3. 3

      Prototype

      A clickable flow tested with five users. Output: a locked scope.

    4. 4

      Build the MVP

      The single flow, instrumented, in four to six weeks. Output: a working product.

    5. 5

      Launch narrow

      Twenty to fifty matched users. Output: activation and retention data.

    6. 6

      First revenue

      Charge and measure repeat usage. Output: the persevere-pivot-stop decision.

    7. 7

      Scale

      One channel, one metric, hire against the constraint. Output: repeatable growth.

    Common mistakes

    • Adopting enterprise frameworks like SAFe in a ten-person company.
    • Using OKRs to describe activity rather than outcomes.
    • Running discovery without ever deciding anything.
    • Switching framework when the real problem is a lack of customer contact.

    FROM EXPERIENCE

    Why a sequence beats a library

    The Crossley Method exists because founders kept asking which framework to use, when what they needed was to know which stage they were in and what evidence would let them leave it.

    Seven stages, one output each. If you cannot produce the output, you are not ready for the next stage — regardless of how good the deck looks.

    Frequently asked

    Are OKRs useful for startups?

    At ten-plus people, yes, if the key results are outcomes. Below that, one metric and a named owner is enough.

    Is jobs-to-be-done worth learning?

    The framing is genuinely useful for discovery interviews. The formal methodology is heavier than most startups need.

    What about lean startup?

    Still sound in principle. In practice, teams skipped the validation half and kept the build-fast half.

    IN SHORT

    • One sequence: discover, validate, prototype, build, launch, first revenue, scale.
    • Each stage has one output — no output, no progression.
    • Borrow tools for framing; avoid frameworks that produce artefacts instead of decisions.

    Explore the Crossley Method

    Seven stages from first idea to first revenue, with one output each.

    Explore the Crossley Method

    THE FRAMEWORK

    The Crossley Method: idea to first revenue in seven stages

    See the full method
    1. STAGE 1DiscoverWeek 1
    2. STAGE 2ValidateWeek 2
    3. STAGE 3PrototypeWeek 3
    4. STAGE 4Build MVPWeeks 3-4
    5. STAGE 5LaunchWeek 5
    6. STAGE 6First RevenueWeek 6
    7. STAGE 7ScaleOngoing

    MORE ANSWERS

    Startup founders

    How do I start a SaaS company?

    Start a SaaS company by choosing a narrow, expensive, recurring problem for a specific group you can reach, validating that they will pay before you build, then shipping the smallest product that solves it end to end. Get to first paid customer before hiring, raising or automating anything. Distribution — how you reach that group repeatedly — matters more than the software.

    How do I find a technical co-founder?

    Technical co-founders come from people who already know you — former colleagues, communities you contribute to, and open-source or startup circles — far more often than from matching platforms. What makes them say yes is evidence: a validated problem, early customers, and a prototype you built yourself. In 2026, many founders should first ask whether they need a co-founder or a contractor plus AI tooling.

    What mistakes do first-time founders make?

    First-time founders consistently build before validating, scope too wide, hire too early, price too low, chase funding instead of revenue, and measure activity instead of outcomes. Each individually is survivable. Combined, they burn a runway cycle before anyone learns whether the core idea works. The counter-move is uncomfortable focus: one segment, one problem, one metric, one channel.

    How do you validate a startup idea?

    Validate a startup idea by testing willingness to pay, not enthusiasm. Run ten to fifteen interviews about what people did last time they faced the problem, then ask for a commitment — a deposit, a paid pilot, a signed letter of intent. Set your pass thresholds before you start. Two weeks of this routinely prevents six months of building the wrong thing.