Data Mining or Data Whining

When point of sale data first poured in on tape (remember those days), it was relatively easy to improve gross margin and inventory turns by visually analyzing the data and making obvious decisions that were to up to that time somewhat hidden in PMA’s (Product Movement Analysis). Annual improvements in gross margin, of which I am personally aware, ranged from a low of single digits to a high of twenty-five percent, and I am sure that there were c-store companies that did better. Today, in most companies, the low hanging fruit has already been picked and the challenge is to dig deeper into the data and seek out solutions that are hidden within.

The following is a quote from Dr. Martin Block, Professor of Integrated Marketing and Communication at Northwestern University: “Analysis of point of sale data is key to effectively managing convenience stores in today’s competitive environment. Data analysis and data mining force emphasis on profitability and marketing response. Analyzing product ranges, store formats and types, pricing and service can all be done to support strategic marketing decisions.

Data analysis focuses attention on profitability and establishes a natural marketing program evaluation tool. Examining sales over time also provides the opportunity to develop a forecasting tool for the purpose of both planning and budgeting. Adding marketing communication spending, including promotion, advertising and special events, allows the estimation of market response and all the advantages that contemporary marketing mix modeling can bring. If consumer level data is added, such as loyalty card programs, best market segments can be identified and much more efficiently marketed”.

I agree it is a long quote, but that is exactly what I like about it. Dr. Block touches on a lot of areas that are usually ignored or, at least, not fully explored. Analyzing sales without factoring in seasonality, promotional programs, pricing, competitive response and communication spending will not yield a usable number to go forward with, or even better yet, to look back at for future planning.

Incorporating all this data into a series of annual barometric research reports is where most “data whining” occurs. It requires a major investment of someone’s time and thought to set up and manage over time. The days of just doing a conjoint variable analysis study or a factor analysis on sales data are over. Please don’t misinterpret what I’m saying. If you haven’t already done those two studies, at a minimum, you still may have a twenty five percent increase in gross margin dollars waiting for you out there. We as an industry can’t afford to get away from doing the basics.

But what about the future? Dr. Block mentions that data can guide the marketing response. I believe that is true especially when companies tie in proprietary consumer research and internal/external databases into data mining.

Due to my involvement with two universities, I am privy to seeing the impact that consumer research can have when coupled with data mining. It can re-direct corporate marketing strategies overnight. When done correctly it re-energizes businesses. It changes how companies think and how they go to market.

The combined information directs the marketing response – and by this I do not mean just pricing strategies. It identifies key customers and key customer segments that are continually evolving over time.

Companies that incorporate data mining as part of their corporate culture recognize the need for customer retention and think in terms of what a customer contributes in terms of his or her lifetime. This leads to enhanced CRM (Customer Relationship Management). They re-calculate their customers’ overall value to the organization both short term and long term. What is a cigarette customer worth to the organization today, and five years from now, versus a foodservice customer or any other? Given that number, how should that change that category’s pricing and promotional strategies that are already in place? How would it change the firm’s sales and profitability forecasts both short and long term?

In many cases, those companies that don’t make data mining part of their culture continue to play catch up. They are always trying to attract “new customers” because they are not satisfying/managing the inherent profitability of the ones they already have. They do reactive, not pro-active, marketing.

In my opinion data mining provides the foundation for pro-active marketing. It positions companies that use it to “change the rules” (see Michel Roberts’ book Strategy Pure and Simple II).

Data mining should not be viewed as an expense to the organization. To the contrary, it has become an exceptional value. The huge capital investment in systems has already been made. Incorporating data mining into your company’s culture is like giving someone a membership in the “fruit or jelly of the month club”. It’s the gift that just keeps giving and giving.