Can you share examples of innovation programs in companies that you are aware of? How do they measure the outcomes of such programs?
December 29, 2012 11:29 PM   Subscribe

Can you share examples of innovation programs in companies that you are aware of? How do they measure the outcomes of such programs? I know Google used to or still has a program where in employees work on projects of their choosing for 20% of their time. Any thoughts around steps that can be taken to increase innovation in an IT organization would be helpful.
posted by r2d2 to Work & Money (7 answers total) 5 users marked this as a favorite
Look around for the Valve employee handbook.
posted by oceanjesse at 1:08 AM on December 30, 2012 [1 favorite]

"Open Innovation" is a big trend. You should be able to find examples of different company's activities with a search.
posted by infini at 5:03 AM on December 30, 2012

You might like this article about 3M's 15 percent program, which was started all the way back in 1948. The most famous result of that work-on-your-own-projects time so far has been the Post-It note. I believe 3M is considered to be the first company to try this sort of arrangement, and the article talks about how the idea originated.
posted by vytae at 8:20 AM on December 30, 2012

I don't think you'll have too much trouble finding academic case studies, this was a rising topic in b-schools a decade ago and I doubt it has gone away.

I think that for Google, one of the outcomes they likely measure is success in hiring and retaining otherwise qualified employees.

As for steps to increase innovation in an IT organization a big one: reduce the distance between IT implementors and architects and the end-users their work serves. Allow and encourage communication and equip your IT people with qualitative user research skills. Look for ways to reduce the cost and risk of trying new things.

A lot of IT problems are, at their core, communication problems. IT can help facilitate communication through apps and infrastructure, and those apps will better fit the communications needs of the organization and its end-users if there is a tight communication loop between those parties and the IT staff.

There is a tendency to overspecify communications between IT and other groups. I'd argue that this tendency developed for a number of reasons, some of which are no longer appropriate. First, once upon a time, IT was expensive. That expense could be justified by the huge return on investment delivered by a successful IT project. Therefore, it made sense to do a lot of up-front investment in planning to increase the likelihood of success. Second, the people who worked in IT were often, um, not as skillful at communications with the types of people who tended to be the end users of the work they were skillful at. They were also relatively rare, so IT organizations employed layers of interlocutors. Third, developing software and hardware is a creative act, and like any creative act, involves a certain amount of unpredictability and chaos. Large organizations seek to manage unpredictability and chaos through the application of process and management. Taken together, all of these things lengthen communication cycles and create the opportunity for important nuances to get lost.

The downsides of these measures may have made sense back at a time when IT was expensive, but each successful IT project, and every unsuccessful IT project since has made them increasingly obsolete. The costs of IT have fallen and the cumulative results of all those successful IT projects of the past have accelerated the pace of change. People have learned from the successes and failures of past projects. The practitioners of IT projects have become more diverse and mainstream, and some of the stereotypical introverted developer types figured out that they were better served by emerging from their dev caves and talking to end users than by relying on interlocutors.

Google's 20% time is, in many ways, an acknowledgement that innovation is often messy and emerges from the leaves of the organizational tree. It is a signal that "management" knows that they've hired smart people who can manage themselves. I don't know how it is received internally, but I suspect that it is most successful if employees also understand that it means that they have an obligation to take on individual responsibility for innovation.
posted by Good Brain at 12:19 PM on December 30, 2012 [2 favorites] a counter-example, leadership at my company have decided to host a 24-hour code-a-thon in a contest format to 'spur' innovation. Which I think is pretty much way off track.

I predict that most activities will be narrow, egocentric pushes to demonstrate individual expertise in a specific technology.

I argued that 'innovation' evolves from a rich understanding of a cross-section of current technologies, and the recognition of a gap in the business or technological space. Synthesis is pretty much a required ingredient.

Deaf ears...

I think something with a sustained commitment like the Google 20%, even five or ten could provide a fertile context for ideas to organically arise from.

Innovation/inspiration occurs via a bolt from the blue. The point of debate is not just, 'How to we create an environment with more *bolts*?' but also 'How do we move that original thought forward into reality?'

My experience suggests that sustained cross-discipline activities generate more *bolts*. What comes next is harder: letting the would-be innovator work at-risk to run the idea to ground, knowing that most will flounder.
posted by j_curiouser at 1:19 PM on December 30, 2012 [2 favorites]

Microsoft has their Garage, as a response to Google's 20% time.
posted by jacalata at 6:13 PM on December 30, 2012

vytae: "I believe 3M is considered to be the first company to try this sort of arrangement, and the article talks about how the idea originated."

I think it was 3M that I heard also has a policy of rotating engineers/scientists between labs every 5 years. The idea being that part of innovation is applying what you know to a new area. OP asked about quantifying outcomes, and 3M's approach appears to be by measuring the revenues from products introduced in the past four years.

But it also depends on what you mean by innovation. If you mean patents and new products, your challenge is more difficult. Basically by definition, innovative IT technologies originate from places with a build-it-not-buy-it attitude. Kerberos came from MIT's Athena project. HTML/HTTP came from CERN. Google started as a master's thesis from Stanford. Firefox (and SSL, and javascript) has it's distant origins at University of Illinois Urbana-Champaign via NSCA Mosaic. There's a pattern here of major innovations being

Given that you're asking us instead of your coworkers, I'm guessing your company, however, is not a build-it firm. And that you're hoping not for some successful new IT product, but to advance adoption of new technology, software, concepts and services from outside. In that situation, its up to you to determine metrics that matter to the company based on what it is you're trying to achieve: waste reduced, costs saved, sales made, etc.

Also keep in mind, that for however awesome Google's 20 percent time is, my takeaway from the Spanner paper was that it still took them 10 years to migrate the stuff that matters (revenue generating things like AdSense) away from MySQL. So no matter how awesome your engineering is, your stakeholders are still risk adverse.
posted by pwnguin at 8:16 PM on December 30, 2012

« Older What do I need to buy to be able to use my fairly...   |   How long does my aircon unit need to be upright... Newer »
This thread is closed to new comments.