ValidationJuly 14, 2026By Maarten Follaets

What is innovation? The definition most founders skip half of

Cover reading what is innovation, showing a new idea card moving from a drawer into someone's hands, with the second half of the definition highlighted and the Foxy mascot in the corner

Quick answer: Innovation is something new or significantly improved that is actually put into use. The OECD's Oslo Manual, the international standard for measuring innovation, defines it as a new or improved product or process that differs significantly from what came before and that has been made available to potential users or brought into use. Both halves are required. Founders remember the first half and forget the second, which is where almost everything dies.

Someone asked me last month what innovation actually means.

Not in a philosophical way. He was writing a grant application and needed to say his thing was innovative, and he wanted to know if it counted.

I gave him a bad answer at first. Something about doing things differently. Then I went and looked up how it is actually defined, and the definition is much stricter and much more useful than the one in my head.

The definition, both halves

The Oslo Manual is the reference the OECD and Eurostat use to measure innovation across countries. Their definition, roughly: a new or improved product or process that differs significantly from your previous ones, and that has been made available to users or brought into use.

Read the "and" carefully.

The first half is about novelty. The second half is about adoption. It is not innovation while it sits on your laptop. It becomes innovation at the moment someone else uses it.

That is a brutal little clause, and I like it a lot. It means the market decides whether your work counts, not you.

An idea is not an innovation. Neither is a working product.

Three different things that get called the same word:

An idea. Free, infinite, exciting. I have ten a day and most of them are bad.
An invention. It works. You built it, it functions, you can demo it.
An innovation. It works and someone uses it, repeatedly, because it made their life better.

Most founders I meet, myself included on my worse days, are stuck between two and three, telling everyone they are at three.

The uncomfortable truth is that a beautiful product with no users is not an early innovation. It is just an invention that has not been chosen.

Innovation does not require new technology

This surprises people. Nothing in the definition says "technology".

Charging per seat instead of per licence is an innovation, if it changes what customers do. Moving a service from an office to a phone is an innovation. Removing a step is an innovation. Half of what we call innovative technology is really just a familiar thing wired into a workflow that finally fits.

So if you are not building something technically new, stop apologising for it. Being new to the world is not the bar. Being used is the bar.

Which cuts the other way too: your AI feature is not innovation because it uses AI. It is innovation if somebody's Tuesday changed. I have written more about that in AI innovation is not the hard part.

Why so much "innovation" never gets used

Here is where I get personal.

I have built and sold a startup before. It worked, and I am proud of it. But if I am honest, I was figuring a lot of it out as I went, and some of it was luck dressed up as strategy. I do not want to run on instinct again.

Because the way most of us fail the second half of the definition is very predictable:

We fall in love with the thing we are building.
We show it to people who like us.
They say encouraging things, because they are kind.
We take that as demand.
We build for six months.
Nobody changes their behaviour.

Nothing broke. The technology worked. The problem just was not painful enough to move anyone, and we never checked.

How to make sure the second half is true

You cannot know for certain. But you can massively improve your odds, and it takes weeks, not months.

Write down the assumption your whole idea rests on. Usually it is: this problem hurts enough that people will change what they do, and pay.
Go find ten people who have it. Ask them about the last time it happened. Do not pitch. The second you pitch, you have poisoned the answer.
Ask for something that costs them: a pre-order, a paid pilot, a deposit, a real slot in their calendar. Applause is free. Commitment is data.
Then build the smallest version that could prove you wrong.

That is the whole loop, and it is the one I wish I had run properly the first time. There is a longer walkthrough in the honest guide to validating a startup idea.

The part I need help with

Knowing this and doing it are different skills. When it is my own idea, I am the worst judge in the room. I hear the yes and skip the shrug.

That is the reason we are building an AI co-founder. Foxy is not there to be nice about your idea. It is there to ask the question you are avoiding, look at your customer conversations, and tell you when the evidence is thin. If you want to put your idea through that before you build it, start here.

So, your idea. Is it novel, or is it used?

And if it is not used yet, what is the one thing you could test this week to find out whether it ever will be?

Frequently asked questions

What is innovation, in simple terms?

Innovation is something new or significantly improved that is actually put into use. The OECD's Oslo Manual, the standard reference for measuring innovation, defines it as a new or improved product or process that differs significantly from what came before and has been made available to users or brought into use. If it is not in someone's hands, it is not innovation yet.

What is the difference between an idea, an invention, and an innovation?

An idea is a thought. An invention is a thing that works. An innovation is a thing that works and that people actually use. Most founders have ideas and call them innovations. The gap between those two words is usually about six months of building and a lot of money.

What are the main types of innovation?

Broadly two: product innovation (what you offer changes) and process innovation (how you make or deliver it changes). Business model innovation is a process innovation with better PR. Notice none of these require new technology. Charging monthly instead of once is an innovation if it changes what customers do.

Does innovation have to be new to the world?

No. It has to be significantly new to the unit doing it, and it has to be used. Something can be old news in another industry and genuinely innovative in yours. Founders waste a lot of energy trying to be original when they should be trying to be useful.

Why do most innovations fail?

Because they never clear the second half of the definition. Somebody builds a real thing that works and nobody adopts it, usually because the problem it solves is not painful enough to change anyone's habits. The failure is in the demand, not the build.

How do I know my innovation will actually be used?

Test it before you build it. Talk to ten people with the problem, ask about the last time it happened rather than pitching, and try to get one real commitment: a pre-order, a paid pilot, a deposit. That is what tells you the second half of the definition will be true.