Start with a specific person and a specific pain, not a market or a feature. If you cannot say it in one sentence, you do not understand it well enough to build for it.

For [who, specifically], who struggle with [problem], today they [current workaround], which costs them [time, money or risk].

• Strong: “Practice managers at independent GP surgeries spend 6 hours a week chasing locum cover by phone and email.”
• Weak: “An AI platform that revolutionises healthcare staffing.”

Test it: talk to at least 20 people who fit “who”. Score the pain for frequency, cost and urgency, 1 to 5 each. Under 9 out of 15, sharpen the customer or find a sharper problem.

Investors, partners and early customers all ask the same silent question: why you? Domain experience shortens validation, opens doors and helps you spot what outsiders miss.

• Domain experience: years in the sector, the network you can call, the insight you have that others do not.
• Commitment: nights and weekends, part-time or all-in. Be honest: it sets how fast you can move and what help you need.
• Runway: how many months you can keep going before you need income from this.

Commitment works both ways. Compliments are not evidence. Climb the ladder: compliments, then time, then data, then money (a deposit, pre-order or paid pilot).

The opportunity check most tech founders skip. A crowded market is not a reason to stop, but “yet another CRM” needs a sharp answer to four questions.

• Is the market real? People already pay to solve it, and it is growing.
• Why now? What has changed: technology, regulation or behaviour.
• Why you? Access, insight or an unfair advantage.
• Can you reach them? Your first 100 customers, cheaply.

Score each 0, 1 or 2. Six or more out of eight: keep going. Under six: rethink before you build. List the top five alternatives, including spreadsheets and doing nothing.

Building is no longer the hard part; directing it is. Whoever writes the code, you own the product decisions, the code repository and the accounts.

• Vibe-code it yourself: days, lowest cost. Ideal for prototypes and validation. Watch security, data and maintainability.
• Freelancers: weeks, medium cost. Best for a clear, well-scoped first version.
• Agency or outsourcing: highest cost, less control. Best for complex builds with a fixed scope.
• SkilledUp Life volunteers: low cost, weeks. Best for bootstrapped tech startups that can manage a team.
• Technical co-founder: equity. Best when the product is the long-term advantage.

Rule of thumb: vibe-code to validate, engineer to scale.

Your funding path changes every decision on this canvas, especially speed and scope of the first version.

• Bootstrapped: v1 must earn money quickly. Pick problems with a budget holder, use the cheapest route that works, charge from day one.
• Fundraising (angels, seed): v1 must produce evidence investors trust: pilots, retention and a team that can execute.
• Venture-backed: v1 must prove a big market and a path to outsized growth; speed matters more than cost.

Revenue first keeps all three paths open. Raising early closes some of them. Model dilution before you raise.

Design the loop that tells you what to build next before you build the next thing. Feedback should arrive on its own, inside the product.

• Ask at the right moment: one question after a user succeeds, for example “What nearly stopped you signing up?”
• Watch behaviour: pick one number that shows real use, such as weekly active users or tasks completed.
• Close the loop: ship what users ask for, then tell them you did. That turns users into advocates.

Earn attention before you buy it. The best tech products spread because using them involves someone else.

When a user [does X], they [share or invite Y], because [they get value Z]. The new user lands on [where].

• Invite loop: the product is better with others: shared calendars, documents, workspaces.
• Content loop: what users make is public and points back to you.
• Referral loop: two-sided rewards for bringing someone in.
• Embed loop: “powered by” on every form, widget or report.