Image source: http://slideplayer.com/8194821/25/images/16/The+Evolution+of+Data+Mining.jpg
The next chapter in the history of business is focused on the integration of analytics into all facets of corporate life. Analytics isn't about massing data, and analytics is not "reporting." It's the work of leveraging applied statistics to develop the underpinnings of effective measurable strategies. This applies to all industries. This applies to all functional areas. Examples include anything from how we abate risk in consumer finance, to how we route calls in a call center. It could include how we create a realistic financial forecast, or how we customize point-of-sale promotions within analytically-derived sub-chains of a retail chain.
This evolution is happening as there is an unavoidable reality that if you are looking at any business opportunity, the fusion of good data/predictive analytics/good intuition/experience will outperform good intuition/experience alone. To further explain, intuition and experience may develop strong business strategies based on the correct and proven belief that consumer demand is driven by each consumer's:
Recency of last purchase
Number of past purchases
Dollar value of purchases
This intuition and experientially based strategy may segment customers on these three factors and develop customized treatments to each segment.
The analytic approach would improve on the example above by accurately attaching quantified value to each of these three variables (what is the percent contribution of each of the three factors to consumer demand?). Further, through analytics we can develop, evaluate and manage ratios among all of the variables and conduct mathematical exercises to tease out additional information. Beyond that, hundreds more variables can be easily explored to see if they add value as well. Best of all, the final product won't be groups of customers, but rather a ranking of each individual customer. A parsimonious solution that is more robust and more easily managed as compared to the business as usual process.
The 3-variable intuitive/experiential strategy in this example is profitable and proven. But, merging additional data and deep strategic analytics to this example (and, in fact, to all that we know today) can measurably take a business' strategies and tactics to the next level.
To actively design organizations that will thrive in this new environment try to objectively assess your current organization for individuals that play on both sides of the qualitative/quantitative fence. If these individuals are available, carefully plan where they can be strategically placed to gain immediate value (A key project? A functional area facing a specific challenge?).
If you need to hire externally for individuals with these skills, and your organization has no senior colleagues with these backgrounds, strongly consider retaining consultants with qualitative and analytic backgrounds and project experience to be on your search committee. There are endless stories of organizations who have built new sophisticated data warehouses (some great - some not so great). These organizations then assembled interview panels comprised of qualitative business people and Information Technologists to seek out and hire the talent that could mine the data and build the strategies. A key flaw in this hiring process is obvious.
"Decision Scientists" are the quantitative/qualitative individuals that are responsible for making the data warehouse pay off. If there are currently no Decision Scientists in your organization, it is not likely that you can effectively identify this talent. If, as an example, you only speak English, you will not be very qualified to hire a professional foreign language translator. Similarly, if you haven't been responsible for developing business strategies that have strong advanced analytic underpinnings, you will be at a major disadvantage in identifying (let alone developing) appropriate key talent.
That said, your hiring committee will be seeking out individuals that have managed to merge their analytic and traditional business skills. These are people that can clearly demonstrate this ability via detailed "case study" examples that review all facets of a project (e.g., for a direct mail campaign, clarity should be gained on the individual's contribution in all project phases, from computer coding that led to a targeting strategy all the way through final creative development). Ask questions that will allow your prospective candidates to showcase their step-by-step analytic, strategy development, and operational skills. And, as with any candidate, these individuals should also display well-developed interpersonal and communication skills (the strategy will have no value if you cannot sell it into the business and provide clear instructions relative to its use). In short, an ideal candidate will have the ability to easily understand and embrace the "big picture" and the logic and value of what-is...prior to merging in the quantitative rigor needed to develop what-is-next.
Once your team is staffed, they will likely face a number of common challenges. Not the least of which is having the company accept resulting strategies for testing. Interestingly, I have found that the likelihood of an organization to embrace analytically-based strategies is inversely proportional to the level of fear that pervades the organization. The fear can be related to the loss of a job, fiefdom, responsibility...or just the perceived loss of stature that results from having an individual from the "outside" find improvement opportunities.
It will be incumbent on your new, well-diversified, analytically-talented and qualitatively astute team to present themselves as internal consultants (...individuals who will only look good if the client looks good). It will also be incumbent on senior management to actively support the responsible testing of promising new ideas.
It is important to build a strong culture of testing in your organization. Have a bias in favor of testing. Manage the tests so that the financial liability is minimized, but so that the results are still statistically valid within a published margin of error. As a centerpiece of this testing culture, underscore that all work needs to be quantified, and that all test results (positive or negative) are wins for the business. By continuing to learn, we continue to improve.
