Product Strategy
Building the wrong thing well is still building the wrong thing. Validation is how you find that out while it's cheap.
Validation isn't a survey, it's evidence of behaviour
Asking friends and family "would you use this?" tells you almost nothing — people are polite, and stated intention is a poor predictor of action. Real validation means finding evidence that people will change their behaviour: join a waitlist, pre-order, hand over an email address for something specific, or pay a small deposit. Talk is cheap. A wallet or a calendar entry isn't.
The distinction matters because the failure it prevents is the expensive one. CB Insights analysed 431 VC-backed company shutdowns and found poor product-market fit behind 43% of them — and two-thirds of those were early-stage companies that never found a market at all. Not companies that built badly. Companies that built the wrong thing, competently.
43% of shutdowns trace to poor product-market fit · 15–20 customer conversations before you commit · 2 weeks is usually enough to know
Start with the problem, not the product
Before testing your solution, test whether the problem is real and painful enough that people are already trying — badly — to solve it. Look for existing workarounds: a spreadsheet someone built themselves, a WhatsApp group standing in for a booking system, a manual reconciliation someone complains about every month. In South Africa the WhatsApp one comes up constantly, and it's a genuinely good signal: it means the need is real enough that people are running a business through a chat app rather than going without.
If nobody is improvising a fix already, that's a signal the pain might not be sharp enough to pay for a solution. Not proof — but a reason to keep testing before you commit.
Cheap ways to test demand before writing a line of code
A landing page describing the offer with a genuine call to action — sign up, join the waitlist, book a call — tells you more in a week than months of assumptions. A concierge version, where you manually deliver the outcome for a handful of early users with no software at all, proves the value before you automate it. And talking to 15–20 potential customers about their actual current process, not your idea, surfaces language and objections you'd never guess sitting alone with a whiteboard.
One rule holds all three together: ask about what they did last time, not what they would do next time. Past behaviour is evidence. Future intention is a wish.
The goal of validation isn't to confirm you're right. It's to find out cheaply if you're wrong — before "cheaply" stops being true.
A practical spec: two weeks, in order
Validation drags on when it has no shape. This is the version we'd actually run:
- Days 1–3: write down the riskiest assumption in one sentence. Usually it's "people with this problem will pay to solve it," not anything technical. Everything else is subordinate to testing that.
- Days 4–8: ten to fifteen conversations, about their current process only. No pitching, no demo. You're collecting the words they use, which later become your landing page copy.
- Days 9–12: a single landing page with one specific call to action and a real price on it. A price with no payment is still a filter; free tells you nothing.
- Days 13–14: decide. Enough signal to scope a build, enough to justify one more test, or enough to stop. All three are wins. Only "let's keep researching" is a loss.
Two weeks, and the cost is your time plus a domain name. Compare that to the cheapest real build.
What validation can't tell you
It won't tell you the exact feature set, and it shouldn't — that's what a proper scoping process is for once demand is real. Validation answers "should this exist," not "how should this be built." Conflating the two is how founders end up with a fully-built app nobody asked for, or stuck endlessly researching instead of shipping something small enough to learn from.
It also can't validate a market that doesn't exist yet. If you're genuinely early, the honest test isn't demand — it's whether you can afford to be wrong for long enough to be right.
When you're ready to move to building
Three things are true at once. You have real signal — people paying, signing up, or showing up — not just polite interest. You understand the problem well enough to describe it in the customer's words rather than your own. And you can say clearly what the smallest useful version of the solution needs to do.
That last one is exactly the input a scoping session turns into a plan. At Produsell every engagement is fixed scope and fixed price before it starts — no hourly billing, no surprise invoices. A two-week Sprint produces discovery, a clickable prototype, a technical feasibility review and a fixed quote for the build, all of it yours to take anywhere. Full builds include weekly demos of working software, and hand over the source code and documentation at the end. We'd rather talk you out of features than sell you a build you don't need, and this is the stage where that conversation is cheapest.
If you're weighing platforms rather than validating demand, the SEO tooling question is the wrong one to start with too — and if you're already trading and just want to be found, here's what SEO services actually cover. Everything else runs through product strategy first.
FAQ
How long should validating a business idea take?
Days to a few weeks for a landing-page or concierge test — not months. Two weeks is a realistic full cycle: conversations, a page with a real price, a decision. If validation is dragging on indefinitely, it's usually avoidance of the harder step, which is building something.
Do I need an MVP to validate an idea?
No. A landing page, a manual concierge version, or direct customer conversations can all validate demand before any software exists. An MVP is what you build once demand is proven and you need to learn how people actually use the thing.
How many customer interviews are enough?
Fifteen to twenty is the point where answers start repeating and you stop hearing anything new. Fewer than ten and you're generalising from noise. Ask about their last three attempts to solve the problem, not about your idea.
What's a red flag that an idea isn't validated?
Lots of polite enthusiasm — "that's a great idea!" — and nobody willing to pay, pre-commit, or take a concrete next step. Compliments are the sound of a test that hasn't been run yet.
Can Produsell help with validation, or only the build?
Our product strategy Sprints often start here — working out what's actually worth testing before anyone commits to a build. It's a fixed-price engagement and the output is yours whether or not you build with us.
Sources: CB Insights, why startups fail (431 shutdown post-mortems)
The short version: spend two weeks and a domain name proving people will act, before you spend months proving you can build.
Tell us what is holding your business back.
A short, fixed-price scoping session with our marketing lead or a backend engineer — whichever your business needs. No pitch, just a straight answer.