I defined innovation in 2018, and I'd keep the definition. The word is the problem.

In 2018 I wrote a piece on this site called "What on earth is innovation? Why a rat in your kitchen means a change is needed", which is still in the Internet Archive. Its argument was that people flinch at the word innovation because nobody defines it, and that the working definition should be a plain one: any new activity that adds value to your business. I'd still keep that definition exactly as it was. What has changed is that hardly anyone says the word any more, and the job it used to do, sounding busy without saying what you're actually doing, has passed to another word.

What I wrote then

The piece said people have an issue with the word because "we don't like what we don't know", and that in my early years in digital marketing I used words like innovation and product development and got funny looks from clients. Then it gave the definition:

"Innovation is any new activity that adds value to your business."

It also made the point that innovation doesn't have to come from marketing at all. A new piece of software for whoever does your HR counts. A new process for your bookkeeping counts. A better script for the person answering the phone counts.

Why that still holds

It holds because it sets a test anyone can apply, which is the only thing a definition is for: is it new, and does it add value? Something doesn't count because it's interesting, or because it was expensive, or because it ended up in a press release.

Most of what gets sold to small and medium businesses as innovation fails the second half of that test, and the useful thing about the definition is how quickly it tells you so. A tool that automates something nobody was doing adds nothing. A rebuilt website that looks better but converts no better adds nothing. Meanwhile, the dull changes that pass the test, such as answering the phone faster or writing down the three questions every customer asks, rarely get called innovation by anybody.

What the word turned into

In 2018, "innovation" was the word people used when they wanted to sound as if they were moving forward without saying what they were doing. "AI" has taken over that job almost exactly. The pattern is the same: a real technology, a genuine set of uses, and a far larger number of people attaching the word to whatever they were already selling.

The same test cuts through it: is this new for us, and does it add value we can see? If a tool writes your social posts in half the time but nobody was reading them, you've innovated your way to the same place, only faster. We've put that question to each channel in turn in what AI actually changes about marketing your business, and to the loudest claim of the lot in whether SEO is dead now that everyone uses AI.

The part I'd cut

The 2018 piece hung its argument on a television chef shouting at restaurant owners, which was fun to write and did no work. The point underneath it was that businesses have a life cycle and that change can revive it. That's true, but a metaphor isn't an argument. I'd now want the test, two examples that pass it, two that fail, and no celebrity.

How to use the definition

Take anything you're being sold, or anything you're thinking of doing yourself, and ask both halves of the question out loud.

• Is it new for us? It doesn't have to be new in the market, only new here. Plenty of businesses have never done the ordinary thing properly, and finally doing it counts.
• Does it add value we could point at? That means something like more enquiries, time saved, lower costs or fewer mistakes. If the answer is "visibility" or "presence", keep pushing until you get a noun you can count.

If both halves are a yes, do it, whatever it's called. If only the first one is, you've bought a new thing. We'd write the answer into a one-page marketing plan, and if you'd like somebody to apply the test with you, our 90-day marketing plan is that conversation with a price on it.