Are In-House R&D Labs Outdated?
from the maybe-possibly dept
Has start-up culture completely changed the idea of research and development? This isn’t a new idea, as many companies have made this point in the past. However, here’s an article raising the idea again. These days, with a (mostly) healthy venture capital industry, and a strong startup culture in the US, many tech companies are realizing that in-house R&D just isn’t worth it. In some ways, it’s the typical “decentralized vs. centralized” argument at work. With an in-house R&D system, you have a centralized solution, where limited resources force researchers to focus on specific problems, and they tend to focus on improving core products, without cannibalizing existing lines too much. This shows why they often get blind-sided by more disruptive technologies that change the market. With the decentralized startup model of research, lots of companies (funded by VCs) do all different types of research, and the bigger tech companies get to pick the more successful ones and buy them out. They may pay more, but the results are more proven in the marketplace, and they end up spending less on completely wasteful projects. Of course, it doesn’t always work out that way, and there are plenty of stories of acquisitions gone bad (though, in many cases, this is because bigger companies losing marketshare tend to buy companies that simply prop up their obsolete offerings, rather than really jump into more revolutionary areas). The big downside of this model, of course, is that it makes integration much more difficult (another reason why many of these deals fail), but that doesn’t seem to offset the upside. It will be interesting to see if the pendulum shifts back, or if the focus will remain on this more decentralized means of R&D.
Comments on “Are In-House R&D Labs Outdated?”
Doesn't solve everything
Some R&D projects cost a lot of capital, and that money may not come through VC. Some of Bell Labs’ work on the transistor, for example, was too speculative at the time for business funding. Not to mention their work on information theory, which was pretty abstract an diffuse in its application.
Sometimes
This works for some stuff. Particularly the ‘new paradigm’ type of research. There is still a lot of basic research that would never get funded this way though.
Also a lot of very targeted reasearch will only ever be done by companies that are already in an existing market. Some vertical markets are too narrow and too expensive for a real startup to get into them. Some types of research can take many millions of dollars a year for a decade or more before you see results.
R & D can be worth it
The American business climate is too focused on short term goals. Inhouse research is still a good way to develop market-leading products 5 to 10 years in the future. Managers are (sorry to say) not rewarded for being patient. Good inhouse research includes these factors:
– within a well-defined area, researchers get free reign to make discoveries.
– the company also has marketing people capable of understanding what the researchers are doing, and seeing how to make future products from their work.
I consulted at AT&T in 1985, 86 and 87, the period during which Bell labs started to die. It was particularly clear that Marketing people in the AT&T computer business had neither the patience nor the understanding to mine the labs for future products. Much valuable work died in this period, because, with no one to appreciate its (imaginative) commerical future, the work was just an expense.
Disruptive technologies are not attractive to big
Disruptive technologies are not attractive to big companies – that’s part of the definition. A disruptive technology is usually lower performing and has a limited market – so it is only of interest to small companies looking for a niche without much competition – while the big company is looking for a way to grow 10-20%. Then, later, when the disuptive tech matures it becomes “good enough”, and it steals the market from the establish tech because it’s cheaper or smaller, better in some key attribute – and good enough in the others.