Reading the 2012 predictions around big data -- a big data pile in itself -- you can't help but wonder if yours is the only organization not employing data-driven decision making in all major business capabilities and processes.
Well, according to McKinsey's latest Global IT survey, the majority of organizations are reluctant to entrust decisions on new products and services to realm of data analytics. In fact, the majority of respondents are relying on data purely for blocking and tackling:
"Despite the promise of big data to reshape strategy and decision making, more than 75 percent of respondents to this survey report that their organizations’ greatest benefits from data use flow from clear and timely reporting of financial and performance metrics. Only about half say they seek to use data to provide new business insights or develop new information-based products and services."
However, the use of data to at least support decision-making is becoming more prevalent in all areas and levels of organizations:
"Most respondents say that managers and executives in their organizations are setting expectations for using data and analytic support in decision making. Indeed, one-third say all of their senior executives demand data, while almost two-thirds indicate that midlevel managers do the same."
Yet, there are many successful cases of data-driven decision-making. So, what's the problem?
Apparently, we are:
"We believe these simpler aspirations reflect the difficulties and barriers to more effective use of data and analytics in decision making. Respondents highlight three barriers in particular (Exhibit 3): a cultural preference for experience over data; a lack of skills in synthesizing and translating the analytics and data for decision makers; and concerns that the data quality is poor. All three of these barriers are recurrent themes in our discussions with clients."
Culture, mistrust of the data, lack of interest. These very human factors are barriers for 46% of the respondents. Yet, these barriers aren't new. Nor, confined to big data and advanced analytics.
In business-technology, we face these same barriers for every trend, every cycle. Beyond the Sisyphean toil, revisiting these fears and uncertainties takes valuable time. Time that could otherwise be applied towards early adoption and competitive advantage.
Of course, taking a "damn the torpedoes" approach can be equally problematic. Invested time and money are at risk, as well as downstream implications of poorly chosen or implemented business-technology.
The odd thing here is our industry replays this adoption scenario all the time. Each trend. Every company. I find it ironic that an industry predicated on introducing change is still so bad at it. But, we don't have to be. Especially now.
The biggest variable in technology adoption is the degree of certainty. How can organization be certain that employing data analytics for decision-making will provide positive returns? And how can an organization be positive as to the return without employing the technology?
For years, solving the uncertainty factor was a mostly academic exercise, focused on benchmarks, case studies, referrals and single purpose proof-of-concepts. Given the academic nature, these exercises were often drawn out, and inconclusive.
To change a culture, you need to bring proof to the table. And proof requires hands-on experimentation and real-world data. We need data to prove that we need data. How will we get that?
By changing how we introduce change. We need experimentation platforms that are collaboration hubs for business and technology professionals to evaluate the appropriateness (cost, risk, opportunity, time) of any significant business-technology introduction.
Coincidently, because of this year's greatest hyped technologies, we now have technology and techniques to build on-going, low-cost experimentation platforms. We can acquire low cost computing capacity via the cloud, we can use evidence-based techniques to run controlled tests over an assortment of open data sets, and we can analyze both expected and data exhaust outputs using open source languages and data stores.
So, perhaps in 2012, we can solve a few problems at once:
1. Demonstrate the value of cloud computing for speciality processing
2. Determine the fit of big data and advanced analytics for individual organizations
3. Accelerate change, via experimentation and fact-based conversation
Alternatively, we can push that rock up another hill.
