When is an innovation not an innovation? One of our perennial problems when trying to establish whether an innovation attempt has transitioned to become an innovation is working out how to account for integrated clusters of innovation attempts. The most visible forms of attempt-clustering being those found in new cars.

Because OEMs tend to inject new models onto the market in three, four or five year cycles, what usually happens is that some or all of their Tier 1 and beyond suppliers turn up before design-freeze day with a catalogue of new ideas and concepts. Some of these will be adopted into the new car design and some won’t.

So far so good. Here’s the problem though. Let’s assume that the new car becomes a hit, and that the OEM ends up repaying the cost of developing the vehicle and, miracle of miracles, gets to turn an actual customer-money-in-the-bank profit. Does that then mean that all of the innovative features present on the vehicle get to call themselves innovations?

The extent of the problem recently became apparent when a friend succumbed to some heavy OEM advertising propaganda and went out and bought a new car. Being his only engineer friend, I was forced to sit in the car so that he could inflict his version of the showroom demonstration on me. There were so many ‘innovations’, the demonstration lasted several hours. Or maybe that’s just what it felt like. Grinning can give you cramp after a while.

After a while, when I’d start become conscious of impending smile-muscle spasms, I thought I’d try and apply a bit of TRIZ Principle 22 to calm my misery. Turn new-owner-pride lemons into research lemonade.

So, after each new feature demonstration, I started asking whether he thought it was something he could see himself using after the initial novelty had worn out.

That worked for a while. Then, when we got onto things like ‘automatic handbrakes’ or ‘heated indicator switch’ (I’m joking – everything was becoming surreal by this stage), the answer to my question turned into variants of the sentence, ‘err, yes, because, I can’t turn it off/get rid of it’.

To cut a too-long story short, the upshot of the analysis was that 98% of the new features were precisely that. Features. As in features with no discernable benefit.

Now, I have to say that this 98% number is rather convenient. Because that’s precisely the percentage of all innovation attempts that end in failure.

Only now, we’ve got to take into account that, because the overall car under consideration has sold well, all of the new features contained in it get to call themselves innovations. Even though 98% of them clearly aren’t. Which means our 98% number is actually recursive.

Stepping back a little before we get ourselves stuck in a nest of recursions, the situation is rather worse than it might already sound. That’s because when my poor friend purchased the car, he had little or mostly no option to not have the feature. Which basically means that he’s paid for a pile of digital and hardware crap that has zero utility. Which, when brought into my Yorkshire-Economics, no-way-am-I-paying-for-that mind, means I’ll probably never be able to buy a car made after, by current reckoning, 2008.

From a more pragmatic innovation-accounting perspective, it means that we need to be very careful about labelling ‘all’ the novel features in any kind of integrated cluster as innovations just because the overall cluster turns into an innovation. Rather, we need to take into account a 98% nested-dysfunction effect and recognise that the large majority of those features are junk.

Which then just leaves the final challenge of somehow isolating them so the OEMs can gain a better understanding of which of the 2% of the new features they’re inclined to keep adding to the cluster are the ones worth spending time and money on. Sounds like a job for PanSensic. Or diverting some of the OEM research funding towards asking their customers better questions. Hopefully in time for the day I have to finally replace my (2007) car.