Last week, I swung through the United States to spend time with four different Fortune 100 companies whose revenues add up to more than $300 billion. One common theme that cut across the discussions was how best to use data to support innovation.
I was particularly struck by this question: “How do I know when I have too much data and I can stop doing research?”
If you regularly read this blog, you probably can guess how I responded. The odds are pretty high that if you’re thinking about that question, you already have too much data, and need to shift your focus to acting on the data.
I’ve spent much of the past year on the applied side of innovation, helping our venture investing arm to spot and shepherd high-potential opportunities and helping our incubation team to lay the foundation for exciting growth businesses. Almost none of these efforts have involved large-scale, detailed quantitative research.
It’s not that the research wouldn’t provide insight. It’s just that most of these efforts are resource constrained, and it’s clear that teams will get more bang for their buck talking to customers, developing prototypes, sharing those prototypes with customers, filling key leadership team gaps and so on, than they will from rigorous data.
The teams are conducting tons of research — such as qualitative discussions with current and prospective customers or quick-hit surveys to gauge reactions to specific ideas — but it’s focused, and the results are immediately incorporated into the strategic plan.
Still, large companies like the ones I met with last week have luxuries that startups don’t. The cash flow from core operations provides funds to invest in exploratory opportunities. Done properly, this can help companies optimize existing markets or discover opportunities that might otherwise be hidden.
Unfortunately, that’s not how I see many companies using research. Instead, quantitative research is done, and done, and done again to act as a sort of security blanket (typically involving enough paper to provide warmth even in cold Winter nights!). It reminds me of the old Winston Churchill line: “Statistics are like a drunk with a lamppost: used more for support than illumination.”
Here’s a suggestion. Look back at the last 10 things your company launched. Compare the actual performance of the product or service to pre-launch projections. Odds are the resemblance is loose at best, particularly for ideas that are innovative with a capital “I.” Use that to advance the argument that perhaps you should divert resources from quantitative research to other forms of learning.
It’s not that I don’t like data. I’m actually a data hound. Heck, I have an undergraduate degree in economics and an MBA. And I truly think that writers like Tom Davenport who talk about Competing on Analytics are right — the smart use of data can really be a source of competitive advantage.
But I’ve seen too many people confuse detailed data with truth.
Read the rest at Scott’s Havard Business Review blog.
Scott D. Anthony is managing director of Innosight Asia-Pacific.