Back in the early 00s, as a result of my work with Internet services at the commercial arm of the corporation, I was lucky to be able to spend some time at the BBC’s Research and Development organisation. Nestled in the middle of a leafy Surrey estate, Kingswood Warren was a mansion house scarred with 1950s and 60s public works extensions, installed with mad inventors running down corridors shouting “eureka” at regular intervals. Kingswood Warren was famed as a place of invention, churning out early colour TV, Teletext and NICAM stereo amongst other things, as well as an being used as an occasional set for the filming of Doctor Who.
R&D in big, patriarchal, benign monopolies was a thing. It was a thing that great institutions like the BBC, AT&T or Xerox did at scale. It was something that led to products (even if, in Xerox’s case, most of those products were built by somebody else).
But these days, although big technology companies do R&D, how much of it actual filters through into products that can be bought? Most mid-term innovation seems to be done through acquisition rather than research paths. What purpose does R&D serve, and what can companies who are looking to increase their innovative capabilities learn?
As part of my research project on Collaboration #sharingorg, the theme of innovation has been coming up again and again. Organisations want to become more innovative. Increasing their collaborative capability appears to be one the the levers.
That has led to the model above – a way of thinking about how different styles of collaborative working fit into the big organisation mix to achieve different ends.
“Innovation” itself is a loaded term, and to me can express a breadth of meaning from marginal gain, optimization of existing ways of working all the way through to quantum leap disruptive shifts. The style in which people can work collaboratively scales from hierarchical control of the industrial era through to leaderless networks as expressed in modern movements like Holocracy.
The history of the industrialized world can be summed up in the top right hand box. Controlled optimization is the expression of how the industrial model works. And it’s great for scaling. It can be an issue for innovation, though, as it really gears against anything different ever happening.
Traditional R&D functions tend to be a form of controlled disruption. Neatly parked out of the main organisation, the traditional challenge has been to enable anything good to make the transition from R&D into product. Xerox’s PARC famously failed to get things to make that transition.
In big tech companies today, I’d also argue that most R&D activity isn’t actually about innovation. It’s much more about giving the impression of an innovative company, and of creating an anti-innovative patent factory, churning out IP to be used to defensively block opponents from doing anything interesting. For all of the talk of self-driving cars, most of Google’s new products in recent years have come from acquisition or incremental improvement rather than R&D.
Where there is much interesting change happening is in the networked optimization box. With the rise of Agile, but even through earlier models like 6 Sigma, control is pushed out to workgroups for them to make decisions about how to develop products or improve processes. As organisations increasingly look to improve customer experience, those boundaries of control need to be further loosened to give scope to reach outside of the organisation.
The final box is the world of startup. And the more that I look at this the more I believe that there’s not much that the world of established businesses can learn from how startups operate other than with a view to acquisition of the successful ones. The chaotic, high failing, loosely-controlled world of thousands of startups is the antithesis of a big organisation. Hack days are playing at it, and rarely evolve into anything other than a team building exercise. Open innovation techniques might be the exception to prove this rule.