Bringing Tech to Market: The Rules of Innovation

Everyone knows that best-quality plus first-to-market doesn't always equal success. A Harvard prof who specializes in this stuff has a great article in Technology Review that digs a lot deeper, called The Rules of Innovation. It's a look at why some technologies are marketplace success stories and some are forgotten failures -- and more, an attempt at rules which predict which will be which. There are lessons here for the entrenched companies (e.g. Sony) as well as for the disruptive upstarts (e.g. Sony 50 years ago). You have to understand the battlefield to win the war.

A couple real insights, a couple tired saws.

[Lemmy Caution]

The strongest insight in the article - and the best supported - is that well-managed companies that take care of their existing customers well are often not innovative, because processes and methods that are profitable are unlikely to be challenged, and because truly outre and novel ideas are too disruptive to be welcomed. The notion that innovation occurs in the context of and also creates disruption is a reasonable one. The rest of the article is questionable: his observations that lean projects are more adaptive than ones that have deep pockets which let them stick with a 'a bad strategy' begs the question of what a bad strategy is. It's the Right Thing To Say in a post-boom age (profits now! no vapor!) - but it omits the many successes that came of plodding along after initial disappointments. And, as soon as he used the word "leveraging" I know his article had run out of ideas.

Re:Like what he has to say...

[markmoss]

But I must quibble with a few points. I suppose he knows more about it than I, but does all innovation in a given field necessarily take place on the low end? What about new products that may or may not be higher-end than existing products? (Example: The runaway popularity of SUVs, which are certainly NOT low-end impulse purchases, and newer models of SUVs seem to subscribe to the Micro$oft bloatware model: more, more, more [useless] features and a higher and higher price tag. But people sure buy 'em.)

SUV's were sold by existing companies to existing customers. This makes them a "sustaining innovation" in the language of the article -- listening to the existing customers and making improvements to the product. And that's if you call them an innovation at all; they are not much different from the GMC Carryall my father bought used in 1963, drove for 15 years, and replaced with a new Carryall.

Sustaining innovations often do tend towards the high-priced end. The customer demand an established company is least likely to respond to with major innovations is "lower price" - you know your existing customers have the money, so making things cheaper just reduces the part of it you get, while making the product better and more expensive might milk more money out of them.

It's not impossible for a new company to be successful selling high-priced products -- think fashion designers and fancy restaurants. But note that you don't get GMs, IBMs, or Microsofts out of such markets. I can remember two computer companies that tried to start at the top: CDC (tried to sell bigger mainframes than IBM in the 1960's, went bankrupt, "refinanced" via an antitrust suit against IBM, and lost the money in supercomputers, IIRC), and Cray (supercomputers). There's also Amdahl, which made imitation IBM mainframes (a little faster or a little cheaper) so I'd call that starting almost at the top. None of these ever did very well. By contrast, starting at the bottom produced the mini-computer companies, at least two of which (DEC & Data General) were apparently quite successful until the PC companies found a lower bottom. And many PC companies have been very successful, although not at all secure - PC's are a nasty bottom-end commodity market where any company that lets its cost control or marketing lapse for a moment is dog-meat for momentarily more efficient competitors. Or possibly to competitors that have managed to lower the quality even further without getting buried in bad units...

Re:What about the SoundBlaster principle?

[Jonathan_S]

Soundblaster wasn't first.

I have an original SoundBlaster. 8bit ISA card, anchient, with box. It even sayes AT recomended (ie it will run in an IBM XT (8088) computer).

It also says 100% Ad-Lib compatable; becuase Ad-Lib came first and Creative Labs SoundBlaster killed them

And nVidia, ATI, S3, and Matrox were fighting it out for 3d acceleration long before 3dfx came on the scene. Of course they didn't hold a candle to the voodoo card, but they were the first. The nVidia Riva128, ATI Rage - Rage Pro, the S3 Virge (graphics decelerator), and matrox's m3d all predate the voodoo, and for quite a while 3dfx smeared them and took everying but the OEM market away from them. Except matrox for 2D only work...

Bad examples

Re:I Don't Believe Him

[happyclam]

Take his criteria for a successful disruptive technology. I can't help but observe that the light bulb, a successful innovation if ever there was one, satisfies neither.

It is instructive to note that he indicates there is a success rate, albeit very small (6%), for innovations that do not meet his criteria. Thus, simply finding an example of a success that does not follow his rules does not, by itself, invalidate the rules.

Plus, if I recall correctly, Edison had hundreds of innovations that were commercial flops. So perhaps the 6% rule is right along the proper lines for your example...