Most founders don’t fail because their idea is bad. They fail because they never identify the right customer, or worse, try to build for everyone. When CB Insights analyzed why startups fail, running out of capital topped the list at 70%, but in many cases it was the final symptom rather than the root cause. The real issue was that teams were building something the market didn’t need.

Over the past few weeks I’ve been working closely with early-stage founders building AI startups. They care about the problems they want to solve and are thoughtful about the technology. But, I keep seeing the same pattern.

A team begins with an idea for a solution. Then they go looking for customers willing to buy. When conversations don’t produce the expected results, instead of going deeper into the issues causing the customer problems, they settle for someone close enough and keep building. The discipline should run in the other direction: understand the customer’s problem deeply before deciding what to build.

I’ve spent almost thirty years inside large multinationals watching this same sequence repeated at every scale. It doesn’t matter whether the team has four people or four thousand. When the customer definition hasn’t been seriously tested, everything built on top of it is fragile.

The Startup Genome Project found that 70% of high-growth startups scaled prematurely along at least one of five core dimensions: customer, product, team, business model, and financials. They identified the customer as the primary indicator of progress. Startups that scale properly grow roughly twenty times faster. Most of that premature scaling begins with the mistake of moving forward without truly knowing their real customer.

The Ideal Customer Profile is where that clarity starts. But founders often treat ICP as a description they write once for a slide deck, rather than as something they continuously discover through disciplined conversations with the market. That misunderstanding leads to several misconceptions that quietly derail otherwise promising startups.

Misconception #1: The ideal customer profile is a demographics exercise.

Founders default to age, income, geography, and company size because those categories feel concrete. But demographics describe who someone is, not how they behave. Two 35-year-old business owners in the same city can have completely different pain points, buying urgency, and willingness to pay.

What drives adoption is situation and urgency. What triggered the problem? What workaround exists today? What consequences occur if the problem continues?

Allan Dib makes this case in The 1-Page Marketing Plan. The right target market is identified through the PVP Index, which scores personal fulfillment, value to the marketplace, and profitability. None of those are demographic. The question that matters is behavioral: who is already acting on this problem and spending money or time trying to solve it? That behavioral lens is what separates a useful ICP from a demographic sketch.

Misconception #2: A broad Ideal customer profile means a bigger opportunity.

It feels counterintuitive to narrow your focus when you’re trying to grow. But Dib is direct: a good niche is “an inch wide and a mile deep.” When you target everyone, you end up competing on nothing because your message resonates with no one specifically enough to drive action.

I watch teams try to serve two segments at once rather than commit to one. The intent is to keep options open. The result is that neither segment gets enough attention to generate real evidence. When limited interviews are split across two groups, neither one accumulates enough pattern density (enough repeated signals) to support a confident decision. Every conversation is half-relevant, and no clear pattern ever emerges.

The CB Insights data referenced earlier reinforces the point. Of those 431 shutdowns since 2023, poor product-market fit appeared in 43%, making it the top root cause of failure and the reason most of those companies ran out of capital in the first place. You can’t achieve product-market fit without first identifying a real customer with a real problem.

Misconception #3: Positive feedback validates your Ideal customer profile.

This may be the most dangerous misconception founders have about the ICP. A founder walks out of five conversations where people said “that sounds great” and concludes the ICP is confirmed. But as Rob Fitzpatrick argues in The Mom Test, compliments are the fool’s gold of customer learning. People are polite. They tell you what you want to hear.

Real customer insight comes from something different: concrete stories about what people have already tried, what they currently do to deal with the problem, and what the problem has already cost them. An ICP begins to emerge when those stories repeat across multiple conversations. Patterns matter far more than opinions.

Adam Grant’s Think Again explains the psychology behind this. Confirmation bias leads us to see what we expect, and desirability bias leads us to see what we want. Founders operating in what Grant calls “preacher mode” treat early enthusiasm as proof rather than treating their ICP as a hypothesis that still needs testing.

This isn’t just a startup problem. McKinsey’s 2024 Growth Leaders survey of over 500 executives found that only 15% consistently incorporate customer input into their business decisions, even though 63% cite customer feedback as a top source for growth ideas. The gap between saying you listen and actually listening is universal.

Misconception #4: More features make the Ideal customer profile easier to win.

When founders feel uncertain about their customers, they often respond by expanding the product. They add features, additional capabilities, and expand use cases. The logic seems straightforward: if the product solves more problems, more people should want it. In practice, the opposite tends to happen.

Anderson, Narus, and Van Rossum made this argument in Harvard Business Review: most value propositions list every possible benefit, leading to what they call “benefit assertion.” The value proposition lists numerous advantages that the customer doesn’t actually care about. The strongest propositions have what they call “resonating focus”: they concentrate on the one or two points of difference that matter most to the specific customer.

Each additional feature dilutes the focus. Instead of learning which capability the customer truly values, the team ends up testing nothing clearly. The product grows wider while the customer’s understanding stays shallow.

When the ICP is unclear, founders try to serve everyone. When the ICP is clear, founders know exactly which value matters.

Misconception #5: Once you define the Ideal customer profile, the work is done.

Teams often treat ICP selection as a box to check in week two and never revisit. But the best ICP work is iterative. You start with a hypothesis, run interviews, surface patterns, and refine. What you learn in week three should change what you believed in week one.

Grant’s scientific mindset applies directly here: if you are not willing to revise your ICP in the face of new evidence, you have turned a hypothesis into an ideology. Dib reinforces the sequence: dominate one niche, learn from it, and then expand. The ICP is a living document, not a finished one.

Founders who treat ICP as a static answer often scale too soon and fail. The teams that succeed treat it as an evolving discovery process.

Bonus: ICP and value proposition are separate exercises.

Many founders approach ICP definition and value proposition design as two separate steps. In reality, they evolve together.

You cannot fully understand your customer until you understand which value matters most to them. And you cannot define a compelling value proposition until you know exactly who it is meant to serve. Customer conversations sharpen the ICP. The ICP sharpens the value proposition. The value proposition guides the next round of discovery.

Eventually, something changes in those conversations. Prospects stop responding with polite curiosity and begin responding with recognition: “That’s exactly the problem we have.”

When that sentence starts appearing consistently, founders are getting close to their real Ideal Customer Profile.

How do you know? Put the message in front of real people who match your ICP and measure what happens. Did they respond? Did they ask a follow-up question? Did they forward it to a colleague?

A response is a signal. A follow-up question is a better one. A forward to a colleague is the strongest of all. Any other response is hope, and as Rick Page writes, Hope is not a Strategy.

This is the discipline many founders at MEST are practicing today. They are not just building technology. They are learning how to listen to markets, refine problems, and narrow their focus until the right customer becomes unmistakably clear.

That discipline is what turns a promising idea into a real company.

Contributed by Dr. Tim Schneller, DBA | Founder, CollabInspire

AI was used as a writing and research tool in the development of this article. All ideas, source selection, and editorial decisions are the author’s.

5 Misconceptions Founders Have About the Ideal Customer Profile (ICP) was originally published in MEST Africa on Medium, where people are continuing the conversation by highlighting and responding to this story.

In the first few weeks inside Makola Market, time bent in every direction.

We’d arrive at 6am when the first trucks unloaded. By noon, I’d watched hundreds of kilograms of plastic move through networks I couldn’t see on any map. By evening, I’d learned the names of women who sort more material in a day than some formal recycling centers process in a week.

We were there to understand why Ghana’s informal recycling economy, which moves enormous volumes daily, leaves almost no structured data trail.

A conversation with one waste picker led me to an aggregator. The aggregator led me to a buyer. The buyer led me to a manufacturer who said:

“We’d buy local recycled plastic if we could trace it. But we can’t. So we import.”

That’s when I realized: the problem isn’t lack of activity. It’s invisibility.

What’s Happening Right Now

Three forces are converging:

Regulations are coming. Extended Producer Responsibility laws will soon require manufacturers to use verified recycled content. Most can’t prove sourcing because traceability systems don’t exist.

Capital is looking for deals. Investors want to fund circular economy infrastructure. But they can’t underwrite operations without material flow data.

The system already works but stays invisible. Informal operators move massive volumes daily. But because transactions aren’t recorded in formal systems, cities can’t plan around them, banks won’t lend to them, and climate funds can’t verify their impact.

That gap represents a multi-billion-dollar infrastructure opportunity.

What We’ve Built (So Far)

We spent 15 months inside Makola before writing a line of code.

CircularTech captures what informal operators already do and converts it into data that cities, investors, and manufacturers can use.

We’re pre-revenue and pre-dataset. But we’re post-validation:

● 20 operators onboarded for April pilot

● Accra Metropolitan Assembly partnership

● Monthly World Bank technical reviews

● Ghana Ministry of Climate Change endorsement

● Top 1,000 at ChangeNOW Paris 2026

By June, we’ll deliver what we believe will be Ghana’s first verified informal waste dataset of this kind.

Why This Matters

Before payment rails existed, you couldn’t send money digitally. Before GPS infrastructure existed, you couldn’t navigate efficiently. Before carbon accounting standards existed, climate finance couldn’t scale.

Before CircularTech, you couldn’t finance informal recycling at scale; because the ground-truth data doesn’t exist.

We’re not asking anyone to change their business. We’re making what already happens visible, measurable, and financeable.

On March 30, I’ll be at the Grand Palais for ChangeNOW Paris; taking this to one of the world’s largest climate innovation summits.

Not to pitch recycling but to show that African-built infrastructure can set the standard for how circular economies become investment-ready.

If you’re attending ChangeNOW, find me. If you know someone deploying capital into African circular economy systems, send them this. And if you’re building at the intersection of hardware, software, and real-world systems, let me know. I’d love to compare notes.

From Accra to Paris to Makola Market.

Contributed by Nabeela Abubakari, Founder, CircularTech

Nabeela Abubakari spent 18 months inside Makola Market before writing a line of code. That ground-up approach became CircularTech, the data infrastructure she now leads. Making Ghana’s informal waste economy visible to investors, cities, and corporations. Her work has earned her the Sustainability Woman of the Year 2026 award, selection as a top 1,000 innovator at ChangeNOW Paris 2026, and endorsement from Ghana’s Ministry of Climate.

We Have No Conclusive Data Yet. was originally published in MEST Africa on Medium, where people are continuing the conversation by highlighting and responding to this story.