What's wrong with innovation programs?

Whether its TCS, Infosys or IBM - companies have at different times set up innovation cells and programs to encourage innovation. However, hardly any new innovations come out of such programs - leave alone transformational ones. The reason for failure of these programs - started with the most well meaning intentions - is highlighted in a talk by Google's CIO Douglas Merrill on Innovation at Google (video below).

The video is not about these companies, or a criticism of innovation management practices. It simply is about defining innovation and then describing how innovation is managed at Google.

The key highlight is towards the middle of the video when Douglas explains that innovation at Google is primarily chaotic. There are no committees to assess projects, no formal budgets, no approvals nothing! Here's what they do (PS: this is not a quote from the video):

Google simply lets its engineers spend 20% time on their projects - thereafter it monitors these projects based on simple rules - like no of users, or no of CPU cycles saved - depending upon the objective of the project.

If a project is doing well - then the engineer may be given more resources (which includes more than 20% of his own time and more engineers, plus technical resources like servers and stuff) to the project. If the project is doing bad then it may not receive these benefits, but there are NO PENALTIES.

This approach is in line with the most important statement Douglas makes in the video - "Reward risks and don't penalize failures".

There are other important aspects of Google's approach - 'listening to users' for their problems - Google does this by studying usage patters, conducting eye movement tests, receiving lots of email from users etc. Once the problems are identified, comes the next approach - crowdsourcing. The problem identified is thrown open to Googlers at large and crowdsourcing is used to come up with many solutions. Most solutions are tested on experimental basis and the successful one is implemented.

And finally, "East your own dog Food" - meaning all new Google products (or new features) are first released to Googlers inside - another form of crowdsourcing. This means that Googlers with their different sized monitors, different browsers, different settings and different intentions of using the product are able to quickly post a barrage of bugs from compatibility to use case testing. All this without the need for a "well formed process" of testing.

The above also means that Googlers are not just allowed but encouraged to customize their usage - a monitor, browser, settings etc of their choice. This is opposite of what Indian tech companies resort to: Standardization and that too Hitler style - "Don't visit this site", "don't use 'x' browser", "we decide what hardware you need" .. and so on. In several companies I know of, even the "Settings" menu on browsers is disabled.

It is towards the end that Douglas himself says that having an organized process like approvals, making of business cases, evaluation committees - only act as gatekeepers making it difficult for the enthusiastic risk taking innovators to even start their work, leave alone complete it.

I would recommend this video to anyone who is even remotely interested in innovation! (It's a long video though - 57 mins- see it in a quite time).

PS: 100% credits to Shubham for pointing me to this video

.