There is an old cliché about business: Know your customer. But in a world where businesses can slice and dice each one of us based on the products we buy, the shows we watch, and restaurants we eat at, the concept has been taken to extreme heights. (Or is it depths?)
The concept of data mining, as it is known, is a simple one: capture and assemble all the information you can and apply some intelligent analysis to it.
But in practice, getting it right can be a complex, tough to handle task, fraught with potential problems if you get it wrong.
Here’s an example from my own world. I have a loyalty card from a major drugstore chain. I receive points that I can redeem for my purchases, but I also earn points when I go shopping with others. Now, if that retailer were to look at the purchases based on that data alone, they might be very confused as to my age, my gender, my preferences.
I can only imagine that the reason I don’t get a copy of their cosmetics offers is that the mismatch in purchasing habits, and products noted on the card, get kicked out by an application.
There are actually programs that can recognize the gender of a name and, one might imagine, disqualify a prospect from receiving certain offers, says Mike Fazackerley of T.O.P. Marketing / Matthew Scott Data Marketing Solutions of Mississauga, Ont., which has done work in the aftermarket for years with aftermarket companies including NAPA and Shell, as well as creating the Automotive Industries Association of Canada’s mapping tool.
While there are some examples of linking data mining to market action in the aftermarket, some of the most vivid examples are found outside, in the world of discretionary purchases.
“There is a group called the Society of the Crown, that has access to customized Crown Royal labels, and other special services,” says Fazackerley. “One example we worked on for [marketer Diageo North America, Inc.] was to have a customized event, by location of the member.” The outcome would be to have a special “private” event that would also allow these premium customers the opportunity to try other premium brands from the supplier.
What he is referring to is the practice of sussing out specific groups of customers and working in ways to maximize the return from those customers. It is the exact opposite of a cold call.
“Marketers learned long ago the importance of classifying customers into species and groups,” reported Terry O’Reilly, award-winning marketing guru, in his “Age of Persuasion” CBC radio show recently. “First dividing them by demographic–categories of age, gender, income, and geography–then by pychographics, grouping people according to values, attitudes, and lifestyles.
“More recently marketers started grouping consumers by ‘tribes’: like-minded groups and cultures with shared common interests.”
While much of the aftermarket is a way off from being able to execute marketing and merchandizing initiatives by “tribe,” it is well on the way to a more focused approach.
For any company or industry intent on selling a product–which one isn’t?–the goal of data mining is to be efficient and effective, in terms of both marketing efforts and processes.
Applied to the aftermarket, the concept is only just starting to take hold. Digging into the service habits, for oil changes, say, might reveal an opportunity.
“What guys do look at is days between service. And they all have that opportunity,” relates Fazackerley. He says that a recent campaign revealed that the time between service calls could be brought down significantly, adding an average of half an oil change a year from customers, if the right message was delivered at the right time to a shop’s customer.
“Response rates depend on the offer,” he emphasizes. And that means knowing the customer and at least making a good guess at what he might respond to. Which brings us back to getting the data in the first place.
This can be particularly difficult for many traditional stores that capture very little about their retail customers.
“There is a good opportunity. You have to start capturing the information, to begin with. Don’t ring it in as just a cash sale.”
For much of the traditional aftermarket in Canada, lacking the kind of aggregate point-of-sale information the U.S. industry has–courtesy of NPD–is problematic.
As an industry, on both sides of the border, the trade customer has been the justifiable focus, but even here there have only recently been strides made to allow detailed parts demand data to be used to create more efficient inventories.
Rod Bayless, Activant’s data expert, looks like you would expect an IT guy to look. Shaved head, scruffy goatee-type beard. Focused eyes. Detailed approach. Clipped speech.
“Today business intelligence and analysis solutions are a big deal. Something like 40% of the dollars spent on relational database technology is spent on business intelligence applications.
“It is a big deal with many customers in terms of what they are trying to do with technology. Making decisions properly and fact-based is becoming more prominent.”
Even today what are often considered sophisticated, in-tune companies consider their own decision-making to be flawed. They yearn for ever more accurate data. It’s an important part of winning in the market.
“If you have influence or knowledge of customers and apply that to servicing them better, you will typically be the one who wins the market segment.
“We have to deal with tens of thousands of different vehicles in order to better service our customers. Data warehousing and data mining helps us to sort all that complexity out. It helps us to put a sharp edge on what we do.”
Whether that opportunity is in the owners of dozen-year-old BMWs, or a whole class of GM Impala with water-pump failures, it is about a focused approach.
It is about attacking a specific opportunity in your business, he says, but to have intelligence, you need data, something there is a dearth of in the aftermarket.
“There just isn’t enough movement of items to learn from your own store. Traditionally what you need to do is get involved in data pooling. Such aggregate data can help determine real demand, and real failure rates.” In one example he offers, you can even make recommendations to a car owner about what service he might want to consider, or at least what components should be inspected, based on failure history.
“The more we understand, the better we can compete. What we need to do is to deliver this data to what we call the point of decision. It can’t be in a PowerPoint or a report. It needs to be part of what you do every day.
“If you are at the counter or working with the customer, you need to be able to articulate something about that car that previously was only available at the car dealer. This is about knowing your customer and being able to chisel the data so you can best compete.
“Taking data like this and mining these types of patterns, and getting them to the installer as tips or advisories, is where the benefits can really flow. You can increase the average repair, target and attract new customers, increase gross profit, and optimize the parts inventory.”