Speed of Innovation

A basic problem when trying to project a possible future is defining some sort of metric relevant to the area you’re researching. In this post, I’m going to detail one such metric that I find interesting when looking at the future of [mobile] software platforms – Speed of Innovation.

No matter where you happen to work, it’s a sure bet to claim that the majority of innovation in operating system features (scheduler advances, memory allocation algorithms), applications (music players, games) and services (location based wikis, streaming media) comes from others. When not affiliated in a way that makes the choice of for you, you tend to choose a system where you have easy access to make modifications – often Linux and other unix derivates.

Open source is a bit like basic research in that there’s no immideate economical benefit for the person(s) sharing innovations with others, but everyone knows that building a better common base allows for greater later innovations (similar to applied research).

Thus, once innovation has happened – and it’s likely it will have happened on an open source system – it will spread to similar or compatible systems. If someone were to publish a better process scheduler, it would quickly spread to platforms where no or very small changes to the original invention are needed.

These platforms would score higher on the Speed of Innovation criteria.

Now, of course it’s possible to duplicate all the innovation (according to some measure of relevance) to a proprietary platform, but it’s quite expensive (and more so the larger the differences from the original). This can be quite hard to accept if you’re a company with an existing large investment into a proprietary system, but as Seth Godin says – we must ignore sunk costs.

Now, not everything has to do with low level operating system APIs. There’s been a shift towards open third party development lately, especially on the smartphones-that-aren’t-smartphones. This has happened partly due to a change in turnaround times from development to getting the application into the hands of actual users (via app stores) but also due to an increase in platform capabilities and better development environments. The future in this area is projected to be what’s called web application development, web apps, and thus in the Speed of Innovation metric we need to take that change into account as well.

Interestingly, it’s the same thing. There’s one web component available, open source, where much of the innovation in the field tends to happen – Webkit. As detailed above, that component is available on unix platforms and if you’re already working with such a platform all new developments benefit your system with no or minimal changes.

Combining an open platform where much code already exists with a modern web engine and display framework and you get a platform where third party innovation will happen at a rapid pace. So rapid, it suddenly becomes less interesting to look at actual support for feature X today, and instead plot a trajectory where feature X is likely to have been supplied by someone, within a certain time frame.

It’s thus less a game of writing long lists of requirements, and more a game of simply (hah) projecting general technological development. Us futurists love to do that. For everyone else:

If you’re a developer, you want to be where you can fulfill your vision.

If you’re a consumer, you want to be where you can do what you want to do.

If you’re a handset manufacturer, you need to be where that innovation happens.

Android™ by Sony Ericsson – the XPERIA™ X10

Android is a trademark of Google Inc. Use of this trademark is subject to Google Permissions.

More information:

Comments 3

Sort by: Newest | Oldest | Most popular

  1. By excarfase

    #1

    Hello! Just want to say thank you for this interesting article! =) Peace, Joy.

  2. By Dr Gordini

    #2

    If this is now happening as you say, why are you so late with the X10? You are entering the market with a Android 1.6 when others are at 2.0 or even 2.1?

    Also, worth noting is that the speed is so far not at all something you could see in the device itself, except for battery drainage. If you through open source have time over for other things, like the timescape fluff, how can you fail to make it run fast on a 1 GHZ processor (more like a 1 Mhz in fact). Also, why didn’t you have time to support multi-tocuh, which in contrast to the above mentioned fluff is something that is of real value to the end users.

  3. Pingback #3

    Social comments and analytics for this post…

    This post was mentioned on Twitter by troed: Speed of Innovation, a metric for evaluating software platforms. Here’s how it applies to the Sony Ericsson X10 http://tr.im/E2hC

1-3 of 3 comments.