There has been much talk about "big data," data intelligence, and data science. Whether you are an experienced consumer of analytics or a neophyte, it is impossible to ignore the trend that those that:
a) mine the right data;
b) develop intelligent and simple algorithms to make sense of the data; and
c) integrate into actual workflows and decision support,
are going to win the data science war in their respective markets. As Marc Benioff eloquently put in his piece in Fortune "we are in an AI spring." While there is a bit of semantics in the big data and AI revolution (what is Artificial Intelligence really?); the statement is so resonant and self evident to us at FPO and AcclaimIP that we almost dismiss it.
The reality is, however, that many organizations and individuals fail to achieve one or more of the above criteria - for a variety of reasons. One look no further than the mortgage derivatives debacle we suffered in the last decade as a stark and stunning reminder that even something so fundamentally rooted in finance (money talks) and at such scale and consequence could escape the attention and understanding of so many incredibly intelligent people awash in data. Or alternatively, Baseball - a sport with massive financial stakes and a fundamental tradition of exhaustive statistical analysis continues to see an evolution in data science today - and most recently and famously in 2002 with the advent of "moneyball." The simple notion that walks were underrated in the market gave the Oakland A's a temporary, but important advantage over those that didn't understand. And today, even more is being done to evolve the statistical metrics that tell the story of our national pastime.
Part of the challenge is really data organization and strategy. Are we collecting the right data? Do we have access to good sources of information? How do we assure we aren't buried in noise to find insificant signals? Do we understand how data can help us craft and answer useful questions? A lot of this has nothing to do with deep analytical skills but rather deep analytical AWARENESS and openness to attack the challenge. The McKinsey report from 2011 asserted that by 2018 there would be a shortage of 140-190k people with deep analytical skills - but also that we would suffer a dearth of managers and analysts with the "know-how to use the analysis of big data to make effective decisions" to the tune of 1.5 million. You can read more at http://www.mckinsey.com/insights/business_technology/big_data_the_next_frontier_for_innovation.
To me, this really isn't a "big data" problem in most situations. It is much more about creating a framework and methodology (really a culture) to extract useful, actionable insights from available data (and be clever enough to expand that when appropriately). Then, build mechanisms to push those insights and decision support tools to the right people before they even realize they need to ask the questions (back to Benioff's "AI Spring"). This process can fail in many ways, frequently it is simply down to collecting "too much noise." In other words, often a default position is "when in doubt, collect data and we'll figure out what to do with it then." In theory, this makes sense - in the long-run, your systems and automation should tease it all out and give you the benefits from where they exist and you don't have to worry about making the right decision a priori. The practical reality suffers from at least 2 real problems: one is simple mindshare and focus and the propensity to succumb to the weight of your own data, as illustrated humorously below:
The other is that it pushes off the harder exercise of data strategy and management and is wasteful. And realistically, it is challenging to create a highly flexible and intelligent system to manage and make sense of disparate and heterogenous data. To be clear, I am not talking about the relatively simple and standard mining of customer data, behaviors, and buying patterns (for example). That data science isn't easy, but it is relatively simple conceptually - and it is a relatively mature area of applied analytics. But probably the biggest hurdle overall is that IP is a multi-disciplinary function with roles in a variety of different departments and therefor disparate accountabilities - not to mention internally generated IP (clearly an asset) is given no financial value on the financial statement.
When we talk about analyzing innovation, intellectual property, or other more complex concepts, especially in the context of the complete ecosystem, it requires a bit of circumspection in designing a strategy for harnessing data and data science (or AI). Returning to the world of sports, Ben Baumer, a professor at Smith College (and previously a statistical analyst for the NY Mets) was tasked by ESPN to categorize every major league team by their use of analytics (or in baseball parlance, sabermetrics). While the details therein and the methodology are fascinating, the results and presentation of results are also quite interesting. Each team was ranked and grouped into tiers:
- "All in"
- "One Foot In"
Keep in mind, this wasn't an assessment of desire or willingness, it was an assessment of behaviors and actions and results. You can read more at http://espn.go.com/espn/feature/story/_/id/12331388/the-great-analytics-rankings.
I think this same sort of methodology can be applied to other business arenas, including IP. Every company should look honestly at themselves and decide where they fit in that hierarchy with respect to IP data science. Our colleagues at SumoBrain.com are mining behavioral data locked in the IP landscape to understand some of this and you should expect to see more awareness and exposure associated with IP behaviors. We look forward to see more.