Small Business Analytics : Insights Without the Ph.D.

Custom_reporting Early this morning, the Vancouver Board of Trade and UBC Sauder School of Business hosted a talk by Neil McGuigan titled Gaining a Competitive Advantage Through Business Analytics. The premise of the talk was getting answsers to questions such as:

  • What factors influence your customers buying behavior?
  • How can you predict behavior based on past performance?
  • Who are your most profitable customers?

Neil provided numerous examples of how larger companies such as GE and AT&T used business analytics to answer these questions and drive increased profits.

Unfortunately, when the topic of small businesses came up Neil really didn’t have much in the way of advice. While this was disappointing, it’s also understandable. By the definition, business analytics is all about mining massive amounts of data, which most small businesses just don’t have. Of course, small businesses still want insights into the above questions.

And here’s where semantics came into play. Neil’s talk was on data mining, and the questions it can help answer. His talk was not on the questions, and the ways they can be answered.

For large companies, the answers are buried in massive amounts of data. They need data mining and business analytics tools to get at them, and generally have teams of researchers with Ph.D.’s to actually do the work.

Smaller businesses have less data, and lack the raw material for effective data mining. For the very reason that it generally doesn’t make sense for small businesses to use data mining, they don’t need to!

When it comes to making decisions and taking action based on real-word data, small businesses benefit far more from good standardized reports, proper key performance indicators (KPIs), and tools to aggregate and drill-down into their numbers.

Neil gave an example of how AT&T used data mining to identify events that caused a customer to leave, allowing them to proactively take steps to prevent it. AT&T has millions of customers interacting with tens of thousands of employees. There’s just no feasible way for them to identify and respond those events without mining and analyzing those interactions.

On the other hand, if you’re a small business owner you’ll more often than not just know what causes you to lose a client. For a B2B company it might be letting a long period of time pass without checking in on a client. Sure, you could perform a detailed analysis and learn a client is 84% more likely to leave after not having been contacted for 6 months, but why waste time and money proving what you already know? You’d be far better off developing a report telling you what percentage of customers haven’t been contacted in X months and who they are.

A restaurant manager doesn’t need data mining to know beer sales will be up during the Canucks’ run to the Cup. What they could benefit from is a report showing sales per man-hour worked. This simple KPI not only provides great insight into profitability, but makes it immediately and empirically clear if they’re over staffing on game nights.

While it wasn’t they focus of his talk, Neil did have one great piece of advice for small businesses:

Collect and store as much data as possible, even if you don’t have the ability to do anything with it today.

This point cannot be emphasized enough. Technology is constantly improving, and it’s only a matter of time before the tools become more accessible. You’re far better off storing data that you have no way of using, rather than wishing you could use data you no longer have.