The NESTA Connect blog is probably one of the two best innovation blogs going, and Jonathan Crowley wrote a terrific post there today, which nicely summarises a lot of issues that we’ve been discussing here. He starts with this picture:
which shows the spectrum of innovation. He then goes on to say that most people tend to think of ‘disruption’ when they think about innovation, but that he’d rather consider both ends of the spectrum.
The other key points in the post are:
- There are benefits to working at the small end of the spectrum – you can experiment quickly and cheaply, and cycle through many ideas at once.
- The main benefit to working at the smaller end of the spectrum is that it’s easier to take action.
- And finally, it makes good sense to have projects across the entire spectrum.
The last point reflects the issue that John was trying to get at in discussing the three horizons framework – which he ties in to the idea of time. You should read the whole post, and it’s also worthwhile to spend some time on that blog – they are consistently interesting. I just wish they posted more frequently!
Tim,
But is this curve descriptive or normative? (i.e. does it describe how things in the world of innovation are or how they ought to be?)
It strikes me that many presumably disruptive innovations are neither hard nor risky (think Microsoft’s DOS in the early days of the PC.) Conversely, some innovations that were touted as future disruptions ended up being expensive and risky without significantly shifting the terrain (think Segway).
We can go back and forth about historical details, of course; but it seems that the generalisation is -at best- glossing over important exceptions.
It also strikes me that the word “Risk” is a weasel word dropped in there to justify the correctness of the graph. If we come up with counter-examples, the authors can very well say “well, at this end of the spectrum things are very risky, and this is just an example of failure”.
But why do innovations on the left side of the scale have small risk? In fact, it seems to me that if we assess risk as a percentage of total investment, then they would contain as much risk (or even more risk) than their more disruptive siblings.
On the other hand, if the curve is supposed to be normative… why?
Well, I think that in general risk correlates fairly well with disruptiveness. On the DOS example, it wasn’t necessarily technologically risky. But in the early days, there were a lot of competing OSs, and it was not at all clear that Microsoft would come out on top. And their level of risk was awfully high in that if they hadn’t prevailed, there would be no more MS, because DOS was all they had at the time.
As you say, we can go back and forth on historical details, but I think that the bigger point is that the drawing is by far the least important part of that post. I agree that it isn’t perfectly thought out – I think it would go better if it used an actual spectrum just showing the level of discontinuity in an innovation ranging from incremental to highly disruptive and that would have made the point more effectively.
I am just not sure how to interpret risk in this context.
Suppose I am a baker who sells, among other things, cross buns. Arguably, if I start selling sugar-free cross buns (assuming this is possible), then this is pretty low on the scale of innovation and disruption (after all, it is not like making wheat-free cross buns or like using an entirely new manufacturing process or …)
Imagine I produce and sell 100 buns a day. But because I switched to sugar-free buns, I only sold 10 buns for some reason (they taste bad, no consumer demand, cross buns are supposed to be fattening …) I’d say this is pretty significant risk since my earnings have dropped by 90%.
Now suppose I am one of the largest bakeries in the country (Woolies, for example). If I switch to sugar-free buns nation-wide, arguably this is a more disruptive change since I am changing the landscape of the cross bun industry. But my earnings will not drop as much a sole trader since consumers will often have little choice but to buy the new sugar-free buns.
It seems that in this case risk is correlated with market disruptiveness rather than technological disruptiveness. But also, contrary to what the graph says, that risk is inversely proportional to market disruptiveness.
Where did I go wrong?
Have you read the original post that I linked to Marco? He addresses a few of these issues there. The main point is that he’s talking about risk within one firm. Woolworth’s and the corner bakery are ALWAYS going to face different risk profiles based on the size differences, so you can’t really compare across firms. He’s really focusing on the choices that face a particular firm.
And with the sugar free buns example, the recommendation that he’s making is that in that situation, you would maybe start with a small extra batch of those to see how they go (a quick and dirty incremental innovation), and then expand on that if it works.
“And with the sugar free buns example, the recommendation that he’s making is that in that situation, you would maybe start with a small extra batch of those to see how they go (a quick and dirty incremental innovation), and then expand on that if it works”.
A bit like Toyota playing around with Hybrids in the Prius before rolling them out to the Lexus etc….
Tim,
I agree that there are different risk profiles. Ceteris Paribus’ing in this situation is inadequate as too many other factors come into play.
So is it legitimate to represent risk as a smooth curve that goes from high to low? as per the post. I think that the answer is no.
Can these risk profiles be ranked according to their disruptiveness and observe some kind of monotony? I think that the answer is also no (or at best no evidence was given.)
I re-read the original post, to see why I misunderstood. It defines: “RISK – lower both for resource and the implications of taking a new innovation to a *market or society*” (emphasis mine)… so I am still confused. But that’s no different from how I normally am 🙂