Customer Divisibility, A Metaphor For Personalization

3
3957

Share on LinkedIn

The more a company knows, or can find out, about its customers and then apply, the better it’s able to personalize and customize each experience, and so leverage loyalty behavior. Every customer touching process comes down to gathering, managing, and using data to optimize loyalty. At the pinnacle of customer insight is the opportunity to make the customer ‘divisible’ down to an individual level.

Most companies have some customer data, drawn principally from transactions. The feedback they get is largely reactive, where the customer initiates contact.

Studies have shown, however, that there are tremendous gaps between information and insight. Here’s my definition of Insight: An understanding of customer needs, problems, expectations and complaints that is thorough enough to provide value in the relationship with each customer. Companies need disciplined techniques to gather, store, share, and apply customer data so that it can be distilled into strategic insight.

When they do this optimally, they often find that customers have multiple identities, different demographic, lifestyle, purchase, and other characteristics which can be leveraged depending upon the marketing, communication, or promotional program, new product development effort, or process improvement or modification activity being conducted. We can say that the customer is divisible.

My earliest exposures to customer divisibility were over thirty years ago, first working with one of the world’s largest customer data management and direct mail companies. Although we defined our selection of households to receive product promotions as a form of ‘segmentation’, we were actually using prioritization algorithms created around our knowledge of multiple customer characteristics, defined on an individual or group basis. This knowledge was then applied to rudimentary customer selection software which helped determine the number of households selected for a promotional mailing.

Somewhat later, as the executive responsible for customer research at The Franklin Mint (a truly esoteric experience), I chose the criteria for selecting the optimum number of buyers to be mailed each of the eighty, or more, product promotions they did every year. We first went beyond segmentation in trying to identify what themes and products would interest customers. Programs were developed on a microsegment basis, in both planning and marketing, using the aggregated, and divisible, aspects of the customers’ purchase profiles as a guide for determining prospective program success.

If we knew that attractive numbers of Franklin customers had previously purchased items with an American History theme, for example, and had also bought innovative collectible items such as pewter sculptures, and one or more product series (also called ‘collections’, as opposed to a single item) such as collections of graphics of historic American sailing ships, U.S. President plates, etc., then we knew which, and how many, of the customers to select first and then in descending order of purchase proclivity.

The customers with the greatest purchase likelihood of buying a Civil War pewter chess set, then, would be those with all of the relevant characteristics just mentioned (even more pinpointed if they had purchased Civil War-related items), followed by customers with some of the characteristics, and finally by those with a single characteristic.

Again, using general divisibility ‘guesstimates’, we identified both the optimum number of The Franklin Mint customers who we thought could be promoted, for budget purposes, and we also estimated the total number of unit sales for each program. My recollection, though several decades distant, is that we were incredibly accurate in our estimates. This was probably more luck and general familiarity with the customer file than anything else.

Applying customer divisibility techniques in the pre-PC days was pretty primitive stuff, more about blending art and science, as I’ve discussed in an earlier column about predictive churn modeling. Data availability, data management and warehousing, and software sophistication for profiling has made divisibility much more possible today.

A good example of an industry which illustrates customer divisibility on a fairly large scale is the gaming business. This industry is both highly competitive and pretty advanced in how they gather and use customer data. Loyalty cards, with data generated when the players first become customers, and then updated at cashless slot machines and gaming tables enable them to overlay casino play with changing customer profiles.

Argosy Gaming Company commented on the leveraging power of customer data and personalization in one of their annual reports from a few years ago (prior to being acquired by Penn National Gaming in 2004, making it the U.S. third largest gaming entity):

With a clear understanding of each customer’s likes and dislikes, patterns and preferences, we can deliver targeted offers and incentives, and use the information to improve our operations. Knowing our customer better will give our company a distinct competitive advantage in the marketplace. And that advantage will mean increased customer loyalty.”

Most of us understand that some people are known to have multiple personalities. We’ve seen the subject treated on television and in the movies frequently enough. Multiple personalities is a useful metaphor for considering the opportunities represented by customer divisibility. The same individual may have several identities, depending upon the marketing, communication, or promotional program planned, product or concept being developed, frequency and type of contact desired, or customer-related process being created or modified. This is the essence of personalization.

With sufficiently detailed information about each customer, suppliers, marketers, sales, and service should treat divisibility as a competitive advantage. They can parlay customer data into the insight which enables them to optimize value for each customer, no matter what aspect of the customer’s individual profile, or personality, is involved.

Michael Lowenstein, PhD CMC
Michael Lowenstein, PhD CMC, specializes in customer and employee experience research/strategy consulting, and brand, customer, and employee commitment and advocacy behavior research, consulting, and training. He has authored seven stakeholder-centric strategy books and 400+ articles, white papers and blogs. In 2018, he was named to CustomerThink's Hall of Fame.

3 COMMENTS

  1. Michael, well put. I agree with you. Various identities of individuals and of “groups” is something most of us do not consider.

  2. Hi Michael

    What an interesting article.

    I agree with you completely when you suggest that we should use what we know about customers, their circumstances and their current context to personalise communications. Whether to the customer as a whole, or to the divisible customer. Increasingly that means providing customers with contextual support, service or even sales prompts to nudge them towards mutually valuable outcomes, rather than the traditional ‘buy my stuff’ campaigns which usually have little of value in them for the customer.

    There are however a couple of caveats I would like to add that limit the power of personalisation.

    The first is that what we know about customers is usually collected as part of an implicit trade-off of more information about customers in exchange for more relevant communications, better services and more useful new products. As numerous studies have shown, this trade-off has started to break down as marketers seek to gather ever more information about customers, but refuse to fulfil their side of the bargain.

    The second is that there are fundamental limits to personalisation. As an article by GfK’s Colin Strong on ‘Are Brands Entering an Uncanny Valley’ suggests, if you plot the degree of personalisation on one axis and the resulting brand liking on the other you find an approximate straight line relationship between the two up to a point, followed by increasing personalisation quickly driving brand liking off a cliff. Once communications become over-personalised they very quickly become too personal, intrusive or even downright creepy.

    These are just part of the dark side of personalisation. In response, customers are going dark on marketers. Just look at how the rapidly expanding use of programmatic digital marketing, with its ads that follow you from website to website, has spawned a new generation of customers who routinely browse the web anonymously using DuckDuckGo, who block webpage data collection using Ghostery and who block ads using Adblock Plus.

    I have seen the enemy, and they are us marketers!

    Graham Hill
    @grahamhill

ADD YOUR COMMENT

Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here