Three questions sit behind almost every conversation with a would-be founder: where do strong business ideas actually come from, how do a rare few attract serious funding and valuation, and how should you think about a problem so that a genuinely valuable idea emerges? None of them is answered by waiting for inspiration. They are answered by putting yourself in the flow of information, running cheap experiments, and being ruthless about what the evidence says.
The approach below is deliberately practical. Each of the five practices is something you can begin this week without funding, a team, or permission. Together they behave like a funnel: you pour in five to ten rough candidates and let deliberate exposure and small tests narrow them down until one idea stands out on its merits.
Build a deliberate network
The fastest route to a better idea is proximity to people who have already walked the path. Networking here has nothing to do with multi-level schemes — it means engineering regular contact with operators, builders, and decision-makers.
Do this
Go where the operators are — in person and online
Two channels compound over time. The first is events: meetups, demo nights, industry conferences, and local founder gatherings, where a single hallway conversation can reframe a problem you've been stuck on. The second is professional platforms rather than consumer feeds. Time spent scrolling general social media rarely returns an idea; time spent on a professional network — reaching out to founders, directors, and senior leaders with a specific, respectful question — frequently does.
Send considered connection requests to people running the kind of company you find interesting. Most will ignore you; a meaningful minority will reply, and those replies are where advice, introductions, and pattern recognition come from. Good operators tend to know other good operators, so each genuine connection quietly widens the surface area of what you can learn.
A network of experienced people is less a contact list than an early-warning system: they surface problems worth solving, sanity-check your thinking, and occasionally open a door that no amount of solo effort would.
Spend weekends inside early-stage ventures
You cannot read your way to an intuition for how new things get built. You have to stand next to them while they are being built. A recurring habit among senior operators in the most active startup ecosystems is to give discretionary weekend time to young companies.
Do this
Volunteer time with a startup — as an intern, advisor, or hand
Whether you're a student, an employee, or already running a business, a standing arrangement to help a startup on evenings or weekends puts you inside the machine. You see the innovations before they're announced, the culture that produces them, and the breakthroughs that don't yet have a name. Household-name companies are a useful reminder that scale is not a prerequisite for significance — several were started in a spare room or a garage, by people who simply began.
For an owner or aspiring founder, this proximity pays three ways. You learn what is being built next, you absorb a different working culture, and — not trivially — you sometimes find something worth investing in. If a venture you've come to understand from the inside grows, its rising value can grow your own stake alongside it.
Validate demand with A/B testing
When you can't yet tell whether an idea is high-potential or a dead end, stop debating it and let customers decide. A/B testing is the simple discipline of putting two versions in front of real people and measuring which one performs.
What it looks like in practice
Suppose you're launching a simple consumer product. You prepare two versions: one in plain, low-cost packaging at a modest price, and one in premium packaging at a much higher price. Both sit in front of real buyers. Whichever sells better, earns more per unit, or returns more revenue is the version you carry forward — and, just as usefully, you now know something concrete about what your market values.
The same method scales down to almost anything. Before writing a line of code for an app, sit five or six prospective users in a room, show them two screen designs, and ask which they'd rather use and why. Authors increasingly test two possible titles the same way. In every case the mechanism is identical: a real choice, a measured result, and the confidence that comes from evidence instead of hope.
A/B testing converts an argument you can't win — “is this a good idea?” — into a question the market answers for you. Each test either strengthens your conviction or saves you from an expensive mistake.
Ship a minimum viable product
A minimum viable product (MVP) is the smallest thing you can build that genuinely solves a customer's most pressing problem. Not a stripped-down disappointment — a focused product that does one important job well.
The remote-control test
A traditional television remote carries thirty or forty buttons, the vast majority of which no one ever presses. A modern streaming remote gets by with a handful. That gap is the difference between a feature-heavy product and an MVP: the streaming remote is often more valuable precisely because it removed everything that added no value.
Features are not the point. Solving the customer's burning problem with the least possible surface area is the point.
Fail fast, fail cheap
Because an MVP is small, you haven't sunk much money or time into it. If it doesn't land, your losses are contained — and you're not really losing anyway, because you've bought something valuable: honest feedback from real customers that tells you exactly how to build a better version next.
A product that older people and children can both use effortlessly beats a cluttered one that impresses no one and helps no one.
| Dimension | Feature-heavy product | Minimum viable product |
|---|---|---|
| Goal | Impress with everything it can do | Solve one burning problem well |
| Cost & time | High up front, before any feedback | Low; feedback arrives early |
| Risk of failure | Expensive when it misses | Contained; losses are small |
| Customer value | Diluted by unused features | Concentrated where it matters |
| What you learn | Little, and late | A lot, and soon |
Read for weak signals
Ideas that look obvious in hindsight usually announced themselves quietly, years earlier, to anyone reading closely. Wide, deliberate reading is how you pick up those signals before they become common knowledge.
Do this
Read biographies, industry reports, and founders' own words
Biographies and autobiographies show you how earlier builders thought when the outcome was still uncertain. Industry writing tells you where each sector is heading. And founders' interviews and essays are worth reading twice — once for what they say, and once to decode what they're signalling about the problems they think matter next.
Pair that reading with the growing set of platforms that track emerging companies and sector trends, where research reports let you see where capital and talent are quietly gathering. Two questions turn passive reading into idea generation: what is actually happening across these industries? and what are the people building them trying to tell me?
Keep a running note of recurring themes across everything you read. When the same unmet need surfaces from three unrelated sources, you've likely found a problem worth taking seriously.
Converge with a one-page canvas
The five practices generate and pressure-test candidates. A one-page business model canvas is how you hold them side by side and converge on a single idea. It forces every assumption behind an idea onto one sheet — where gaps become obvious.
How to use it to choose
Write each surviving idea onto its own canvas, then run the five practices against it. The idea that fills every block with real, evidence-backed answers — and leaves the fewest blanks — is almost always the one worth pursuing. Applied honestly, this is how five to ten candidates collapse into a single high-potential idea.
The same tool works inside an established business. Ask senior team members to map what they're working on and the ideas they're pursuing onto a canvas each. When people self-assess their best ideas this way, attention stops scattering across ten half-formed initiatives and concentrates on the one thing that matters most.
A possible action in the field
Reading this changes nothing on its own. Here is a concrete first pass you can complete without funding or a team — the smallest set of actions that puts all five practices into motion.
- List five to ten rough ideas.Write them down as they are — unpolished is fine. This is the top of your funnel.
- Send ten thoughtful connection requests.Reach out to founders or senior operators in areas you care about, each with one specific question.
- Offer a startup a weekend hand.Volunteer time with one early-stage venture to learn how new things actually get built.
- Run one A/B test.Put two versions of your strongest idea in front of five or six real people and measure the response.
- Define the smallest MVP.Describe the least you could build that solves the customer's core problem — nothing more.
- Read three sources and note the themes.A biography, an industry report, and a founder interview — then record what keeps recurring.
- Fill a canvas for your top idea.Complete all nine blocks. The blanks tell you exactly what to test next.
The essentials
- Ideas come from flow, not waiting. Put yourself where information, operators, and problems circulate.
- Network deliberately. Events and professional platforms — reaching senior operators with specific questions — beat passive scrolling.
- Get close to the work. Weekend immersion in a startup teaches an intuition no book can.
- Let customers decide. A/B testing replaces opinion with evidence about what people actually value.
- Build the smallest thing that works. An MVP fails cheap, learns fast, and concentrates value where it counts.
- Read for signals, then converge. Wide reading surfaces problems; a one-page canvas turns candidates into one idea worth building.