Separating Angel From Devil Customers Can Be Trickier Than You Think

3
1375

Share on LinkedIn

I confess. I’m a high-maintenance catalog shopper. You direct marketers out there know my type. I’m the customer who always calls to place an order— shopping on the web is too much trouble for me. I’m the customer who buys three different sizes of one item and invariably returns two of them— and sometimes three. I’m the customer who says, “I bought this fabulous peach blouse 18 months ago. Can you look up the old item number? I want to order another just like it in turquoise.”

Just last week, I called Brand A. They’re one of my favorite catalog companies to torture when I need to spruce up the wardrobe. This particular call went a bit differently from previous calls, though. I tried to order a sweater from a ripped-out catalog page I had stuffed in my to-do pile. The Brand A customer service representative insisted that I must have a catalog number to place the order.

A reasonable request, to be sure, but Brand A had always happily accommodated me during similar calls. Sadly, my encounter with Brand A didn’t end well that day— the representative said she couldn’t help me. I didn’t place the order. The episode shook my catalog diva status to the core.

What changed? Did I just get a CSR who didn’t know how to navigate my beyond-the-pale request? Did the encounter signal a new tough-love customer service policy? Or did it mean that Brand A had really analyzed my purchase behavior and profitability?

An Angel with a Devil’s Pitchfork?

The premise of Larry Selden’s and Geoff Colvin’s book Angel Customers, Demon Customers suggests companies use data and analytics to discriminate between customers who really deserve extraordinary service and attention and those customers who are simply draining profits from the bottom line. Retailers have taken note. Some have started instituting stricter return policies— the authors list Royal Bank and Fidelity Investments as just a few of the companies employing this approach.

Too often, however, companies choose an all-or nothing approach to customer service policies and service levels. That’s why customer service centers have a great opportunity to institute true customer yield management policies. Catalog and Internet retailers have a wealth of data on individual customers: purchase history, spending history, product return history, complaint calls and service failure. That’s a lot more than your average bricks-and-mortar retailer. How can all of that rich data drive a more profitable– and effective– customer service strategy?

The first step is to get a handle on a customer’s true profitability through a Customer Equity Model. Like me, the Catalog Diva, there are a segment of customers who spend a lot with your company each year. A Customer Equity Model goes beyond segmenting customers by spend, however. The model first examines each customer’s revenue. Next, it allocates the cost of servicing each customer. For you other divas out there, this stage is where a problem in our self-perceived status can begin.

Servicing costs fall into two categories: direct and indirect costs. Direct service costs include the cost of goods sold, shipping and handling costs, order call costs, service call costs and other costs that can be tracked to a specific individual. For simplicity, many companies calculate an average cost for each call to place an order. That average cost assumes all customers spend roughly the same amount of time on the phone. More advanced techniques apply a variable cost per call depending upon the nature of the call. For example, if a customer’s call history reflects three calls to place an order and one to track a missing shipment, the three order calls may take an average of four minutes each, but missing shipment calls may average twice that long.

Indirect service costs are trickier. For example, how much of the call center overhead should be allocated to each customer? How much of the general advertising budget should be allocated to each customer? The answers will be custom to each company’s situation. The art in Customer Equity Models is finding the optimal metric to drive indirect cost location; total dollars spent may not be a good metric. For example, if I order one $250 jacket, does that order take as much time and effort to service as an order for ten $25 sweaters? Probably not.

In the case of general advertising costs, it’s possible to refine the allocation further. This refining can be done by market or another general grouping of customers. For example, if the Cincinnati market has a general advertising budget of $1 million and Cincinnati customers purchased 100,000 units, then each Cincinnati customer would be allocated $10 of the marketing budget. But if the Dallas market has a budget of $5 million and Dallas customers purchased 1,000,000 units, then each Dallas customer would be allocated $5 of the marketing budget. Why allocate the same costs per customer over extremely disparate markets? Tailor them instead.

The end result of this detailed Customer Equity Modeling effort is that you calculate each customer’s profitability. You can then rank and segment customers by how much annual profit they contribute. The second step: score customers by their attrition risk. How likely are you to lose each customer? The third step: estimate each customer’s untapped potential spend with your company. This step isn’t for the faint of heart, but techniques exist to help you gain enough insight into this $64,000 question to make smarter budget allocations.

An Optimized Service Strategy

The fourth step is to create a segmentation strategy matrix on all three dimensions: profitability, attrition risk and potential spend (see chart). Once this step is complete, the strategy for your approach in the customer care center comes in focus. Take the Up Sell customers in the green zone. Not only are those customers high value, but they also have a lot more spend to give you. Could I possibly be this type of catalog diva?

For these high-value, high-potential, high-risk customers, toss out talk time as a measurement metric. Replace that metric with total average sales and total repeat sales. Use your IVR to route these true divas to your very best CSRs. Customize your screens to alert your CSRs of our true diva status. Train them to handle this elite group of your best customers. Ensure this segment receives your most aggressive trial offers to entice them to shift spend to your brand. If they need a little wiggle room in a return policy or other customer service approach— give it to them.

As a contrast, consider the strategy for customers in the Retain segments. They’re high-value, but they’re already giving you most of their spend. Do you want to lose them? No way. But, when tough budget choices beckon, it’s better to divert some of the customer care budget to the real Up Sell divas instead. So what’s the best way to handle these folks? Two words: kid gloves. Yes, route them to the best available CSRs. No, don’t rush them off the phone if they require special service. But you don’t need to dazzle them with the hottest offers, because they’re already spending as much with you as they can. Call center performance metrics for this segment, therefore, will be total average sales and longevity.

Now consider those Ignore customers. There’s a lot of noise in the marketplace right now about “firing your customers” when they aren’t profitable. We don’t advocate this scorched-earth policy. Instead, we advocate the wise allocation of scarce resources. When faced with limited funds, simply spend less on these segments. Encourage them to use self-serve channels such as the web and IVR. You might require these segments to adhere to your published return policy. And you most definitely won’t give them advance notice of your hottest offers or newest products.

The Devil Inside Me

All of this customer yield management talk raises a natural question, of course. Am I really the catalog diva I think I am? Or, is there a devil inside me that is truly tormenting the companies I choose to buy from? I’ll let you decide that one. But make the wrong choice about how to treat me, and you may regret it. Hell hath no fury like a diva scorned.

Kelly Hlavinka
COLLOQUY
A partner of COLLOQUY, owned by LoyaltyOne, Kelly Hlavinka directs all publishing, education and research projects at COLLOQUY, where she draws on her broad experience as a loyalty strategy practitioner in developing articles, white papers and educational initiatives.

3 COMMENTS

  1. Kelly

    A very interesting article about a tricky subject. It leads me to a few difficult questions.

    Which is the most important of the following two customers?

    Customer A, a socially-active elderly lady who has always been a loyal buyer, who has spent $45,000 with you to-date, will only spend $1,000 this year and no more, and will likely stay with you for a further four years before retiring in Florida?

    or

    Customer B, a young working single mother new to you, who has spent $2,000 with you to-date, will spend $1,000 with you this year with the potential to grow that to $2,000, but will only stay with you for three further years at most before moving on to another city?

    Customer A has a total lifetime value of $50,000 (ignoring discount rates, etc) with only $5,000 remaining. But she has been loyal to you through thick and thin including when you struggled to adapt to the changing world of big box retailing. Customer B has a total lifetime value only $10,000 but $8,000 remaining. She isn’t loyal and she will soon be gone. The calculating manager would say invest in growing Customer B to her full annual potential for her remaining tenure and ignore Customer A. She has shopped with you for the past 45 years and isn’t going to change now. But a human one would say invest in Customer B, but look after Customer A too. Her long years of loyalty has earned it. The calculating manager is economically correct to go after Customer B, but he probably doesn’t have any friends and he probably doesn’t even notice!

    What impact do you think the social networks of the two customers (and their recommendation value) would have on their value?

    Do you think that customer management decisions should just be based on customer valuation numbers alone? And which numbers? Or is customer management about more than just cold, hard numbers?

    Graham Hill
    Independent CRM Consultant
    Interim CRM Manager

  2. Identifying more valuable customers and treating them better is one of those core CRM principles that’s hard to argue with. Why spend time/attention/money if there is no payoff?

    The problem is, as Graham points out, in the blog-powered WOM world we now live in, it’s harder than ever to figure out most valuable customers. I suspect most companies will revert to value = those that will buy in the immediate future.

    The other problem is that I fear some companies seem to think that treating selected more-likely-to-be-valuable customers better means treating the other not-so-valuable customers poorly. Even a small disgruntled number can now post a video on YouTube or create a new version of [your-brand-here]-Hell via the blogosphere.

    In my view, the successful companies must develop a strategy to treat *every* customer well to avoid the negative WOM, and then strive to give something special to more valuable customers.

    Bob Thompson, CustomerThink Corp.
    Blog: Unconventional Wisdom

  3. Bob

    You are so right.

    The customer-cashflow relationship is often portrayed as a simplistic linear one, whereas in reality, it is more complicated and complex. This can lead to stupid things happening, like the Sprint 1,000 fiasco, which ultimately contributed to the CMO loosing his job.

    And the relationship is getting more complicated and complex. As you point out, the relatively linear Customer Consumption Value (CCV) component is not so difficult to estimate. (And it can only ever be a crude estimation, in contrast to publicists who wax lyrical about Return on Customer and other silver bullet measures say.) But the non-linear Customer Referral Value (CRV) component is more difficult to calculate, involving more difficult maths and more tenuous assumptions. And unlike the CCV component, the CRV component is reversible. That is, I influence other customers through what I say and do and other customers influence me through what they say and do in return. For example, I am looking for a new high-end laptop, but I won’t even consider a Dell because of their well-publicised Dell Hell debacle and their initially arrogant response. And I don’t need to know Jeff Jarvis personally to decide that a Dell is not for me.

    As the world becomes more connected, the calculating managers are going to become overwhelmed by the complication and complexity inherent in networked business. The old models will, indeed, already have become inadequate. And the brave new world isn’t so amenable to calculation as in the past. Irrespective of how much computing capacity or staff with PhD’s in quantitative finance you have. Just ask financial markets.

    Graham Hill
    Independent CRM Consultant
    Interim CRM Manager

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