Originally published at Forbes
I will cut straight to the chase. You do not “come up” with good ideas—at least, not in the way popular business culture would have you believe.
Sometime many years ago, probably during a painfully quiet brainstorming session, a well-meaning facilitator uttered the now-infamous line “There’s no such thing as a bad idea.” Combine that with Edison’s endlessly misquoted line about inspiration being 1% genius, 99% perspiration, and an entire motivational industry has grown up around the notion that persistence alone will redeem any concept, no matter how detached from reality.
No experienced founder actually believes this. The data is explicit—the single most common reason startups fail is lack of market need, accounting for roughly 42% of failures. In other words, the idea was not wanted in the first place. No amount of grit compensates for solving a problem nobody has.
Uri Levine, cofounder of Waze, summarizes this better than any slogan: “Fall in love with the problem, not the solution.” Every successful business he has built exists to solve a real, persistent frustration.
I am from the same camp. My company was founded on intractable high-friction cross-border payments that I encountered—and naturally discovered I would not be the only business owner to suffer.
Apart from the familiar advice about meeting customer needs, let’s consider how viable ideas actually formulate—not how they are pitched on social media but how they survive.
All ideas start off ugly.
There is a romantic belief that good ideas arrive fully formed. They do not. They arrive awkward, incomplete and usually wrong in at least one important way.
Steve Jobs once observed that new ideas are rarely new in isolation. They are recombinations or splits of existing concepts. In philosophical terms, they are evolutionary rather than revolutionary. This does not mean that any arbitrary mash-up is valuable. A banking theme park or a pension lottery may be novel, but novelty alone is meaningless.
Pixar cofounder Ed Catmull describes early ideas as “ugly babies.” They are fragile, easily damaged by premature judgment and almost never impressive at first glance. Originality, as Catmull notes, is delicate. The danger is not that early ideas are flawed but that they are killed before they can improve.
The real challenge is knowing how to protect an idea without deluding yourself into nurturing something fundamentally unwanted.
Ideas should be stress-tested, not judged.
The mistake many founders make is exposing an early-stage idea to the harshest possible audience far too soon. An investment committee, a public launch or a hostile pitch environment will not refine an idea. It will simply reject it.
Effective idea development is staged. In the earliest phase, ideas should circulate only among a small group of trusted colleagues or mentors whose role is not to declare it good or bad but to identify where it breaks. At this stage, the goal is shape, not validation.
As the idea improves, the audience should expand. Each new group should be marginally more critical than the last. This progression mirrors how robust systems are built. You do not test a prototype aircraft by flying it into a storm on day one.
Crucially, you need to detach your identity from the idea. Catmull’s reminder that “You are not your idea” is not soft advice. It is operationally essential. Feedback cannot be processed if every criticism feels personal.
This mindset is visible across high-performing organizations. Satya Nadella’s emphasis on a “learn-it-all” culture at Microsoft replaced a defensive environment with one that treated feedback as fuel.
Test small, learn fast, repeat.
Once an idea has survived internal refinement, it should be tested in the real world, but in controlled conditions.
This is where minimum viable products, pilot programs and limited beta releases earn their keep. The objective is not growth. It is learning. Eric Ries captured this succinctly in The Lean Startup: The only way to win is to learn faster than competitors.
Small-scale testing reveals truths that no internal debate can surface. How people actually behave when faced with your product will almost always contradict your assumptions. That contradiction is valuable.
Airbnb’s early story illustrates this perfectly. When the founders first proposed that strangers would rent rooms in each other’s homes, the reaction was near universal skepticism. Rather than abandon the idea, they ran a small experiment during a conference. It worked, just enough to justify continuing.
Iteration is the work.
The uncomfortable reality is that most ideas do not improve linearly. They lurch forward through cycles of testing, embarrassment and revision.
Pixar’s creative process is instructive here. Even after a story “works,” teams actively look for what is still broken. Mistakes are treated as part of creativity, not evidence of incompetence.
This mindset is equally applicable to startups. Reaching product-market fit is not an endpoint. Markets shift, customer needs evolve and competitors adapt. The work of iteration never stops.
Adobe’s Kickbox program formalized this logic by giving employees small budgets and a structured process to test ideas cheaply. The outcome was not a flood of new products but a cultural shift. Experimentation became routine rather than risky.
The lesson is simple. Treat ideas as hypotheses, not beliefs. Each version exists to be disproven or strengthened by evidence.
Stop asking if an idea is good.
Forget good or bad. Better questions to ask: What part of this works? Where does it fail? What assumption is most fragile? What evidence would change my mind?
LinkedIn cofounder Reid Hoffman captured this bluntly: If you are not embarrassed by the first version of your product, you launched too late. Early discomfort is a sign you are learning in public rather than perfecting in isolation.
Good ideas are not discovered. They are built, dismantled and rebuilt until the problem they address becomes undeniable.
Most ideas will die, but a few will survive long enough to matter. Your job is not to predict which is which but to create the conditions where learning happens faster than failure.
Comments are closed.