Recording customers in motion

Jan 01, 2007 10:30 PM  By

During the next five years, retail store sales are projected to grow by 2.4%. In contrast, e-commerce sales are expected to climb by 13%-14%. Equally important, sales where a consumer researches online before making a purchase in a store could grow to represent at least 17% of retail purchases and as much as 30% for categories such as electronics.

To make the most of these trends, you should be leveraging your customer insights — particularly those from your online customers. The majority of online marketers know that they’re sitting on a gold mine of customer data: click-stream information, demographic details, transaction histories, sources of traffic, effectiveness of search marketing, changes in conversion, drivers of average order values. Yet too many multichannel merchants miss opportunities to increase — sometimes dramatically — the value of their customers.

Of course, these merchants understand that they need to move beyond looking through the traditional mass-market retailing lens. But what holds them and their companies back isn’t the lens; it’s the camera. You can gain the greatest insight by using an online customer’s data to create a kind of full-motion marketing video that shows how behavior has changed over time, rather than a still-photo camera showing an isolated single point in time.

By “recording” customers in motion, you can identify how their value to your business has changed and what forces are driving changes in their behavior. Then you can direct your resources accordingly. By analyzing the video, you can construct targeted marketing programs that address the changing behaviors and needs of specific segments of your target audience. With the marginal cost of making unique offers in an online environment close to zero, this approach offers you the chance to earn a substantial return on assets and capabilities that you’ve already invested in.

THE SIX-STEP SOLUTION

Through our work at McKinsey & Co. with a number of clients, we developed a method to transform the data equivalent of still photos into virtual video. When used by traditional subscription-based companies such as banks and telecommunications providers, this approach suggests a potential EBITDA improvement of 5%-7%. But among product-focused retail companies that have adapted these principles in their online business, their efforts to monetize personalization have led to 25%-30% increases in revenue per customer.

We’ve broken the approach into six distinct but closely linked steps that connect the strategy with results:

  1. Visit the data mine

    The process begins by creating a data table from a random sample of the customer base. Every piece of information about each buyer may be important, from transactions, order size, and order frequency to repeat purchases, click-stream, demographics, and service history. You want to use what you know and to learn what you don’t.

  2. Conduct unbiased analysis

    Next, run these data through a CHAID (chi-squared automatic interaction detector) analysis. This is a classical algorithm whose purpose is to work through very large data sets and construct relationship “trees” that identify, in this case, which factors about customers have the most statistically valid correlation to the customers’ value. Not only is this an efficient way to analyze the data, but it also keeps biases out — again, you’re looking for what you don’t know yet. The outcome of this hard-data analytics is the development of value-based customer segmentation.

  3. Add what you already know

    A crucial component of this process is linking the segmentation data with the results of existing (or new) attitudinal and brand-focused research. The analytical work has determined the most-valuable customers to pursue and behaviors to encourage; it’s also important to know how each of these customers views your company and your brand. Then you can decide how to try to change behaviors. Being hyperanalytical is good for making mailing lists but not for delivering specific products and price points that are personalized to a given consumer.

  4. Test, test, test

    Sure, you know the importance of testing by now — but it bears repeating. There are plenty of potential tools, from e-mail offers to A/B Web tests, to help you better understand variations within consumer subsegments before you invest in full-scale programs.

  5. Personalize and automate

    Using insights garnered from the successful test programs, develop statistical models for driving deep personalization in large-scale programs that will deliver results to the bottom line. Automate the process of applying the insights by creating dynamically generated online offers or ensuring that the most effective retention offer is automatically generated during a customer service encounter. Just as it’s important to see the customers in motion, it’s also important to move quickly to capture them. (A number of packaged software solutions are available to assist with this automation, and even developing a custom solution isn’t cost-prohibitive.)

  6. Refine, refine, refine

    This may be the most important step. While you will reap the largest returns in the first iteration — because your company will be learning to capture value that was previously unrealized — there will still be opportunities to generate significant additional value through refinement. It’s important to look at this refinement from the technical/analytic perspective as well as from the organizational point of view.

Use daily scorecards that show how customers are continuing to change. Just as important, however, is to mesh the structure of the marketing operation with the conclusions generated by watching the virtual video. It’s vital to use the insights gained in the first five steps to ensure that people have the right roles and skill sets and to create a performance management system that has a clear linkage between customer segment objectives and the overall financial targets for a category.

EVERYTHING OLD IS NEW AGAIN

Some of the leading e-commerce and subscription-based players have been using this strategy for many years, yet the majority of retailers continue to apply mass-marketing techniques to their online customers. By changing the way in which you look at data and incorporating the “virtual video” into marketing programs and organizations, you’re all but certain to see financial gains, regardless of your product and customer mix.

In a world where e-commerce will grow six times as fast as retailing overall, and where personalization and highly segmented marketing are the keys to success, this strategic shift is central to the success of multichannel merchants. You already have the resources and the insight. All it takes is the right touch with the right camera.


Amy Guggenheim Shenkan is a consultant and Sean Collins is an associate partner in the San Francisco office of McKinsey & Co., the international management and marketing consultancy. Jeff Schumacher is a partner in the Los Angeles office.