Move High-Value Customers to the Head of the Call-Center Line


Share on LinkedIn

Today’s consumers are hyper-informed. They research potential purchases in surprising depth, comparison shop on the web, and then buy from just about any merchant on the planet. The call-center agents who field their inquiries, however, are typically flying blind. When a call connects, they have no clue who is on the line. And without good information, the wrong agents often make the wrong pitches to the wrong people. As a result, the best prospects often slip through their fingers. In a customer-service situation, the best customers often go unsatisfied.

Imagine you’re a credit-card issuer trying to sign up subscribers to a gold card. You publish a special toll-free number in golfing and fashion magazines. When the ads hit the news stands, the calls start coming in. Caller A has no chance of qualifying for the card, but your agent spends 10 minutes on the phone with him before she figures that out. Caller B is qualified for the card and ready to sign up, but after nine minutes on hold, he abandons the call because the agent took so long with Caller A. Caller C is qualified not only for the gold card but also for your platinum card. Your agent should be trying to sell her the platinum, but, hey, this is a phone line for the gold card.

Not very good, is it? To compete, consumer-facing companies need to get a little smarter. They need to take all their customer knowledge and market research and apply it to brand-new prospects before the agent even picks up the phone.

The bulk of the effort is the initial customer analysis.

With this capability, companies will know—even before calls connect—who is most likely to convert, who has the highest value and which offers will be most compelling for each customer. They can move high-value callers ahead in the queue, route each caller to the most appropriate agent, customize sales scripts and direct low-value customers to interactive voice response (IVR) systems—all based on a subsecond snapshot of the caller. This is real-time segmentation, a process that applies powerful analytics at the point of contact—the inbound call, a web site or brick-and-mortar store visit—not just to the existing customer database.

To get real-time segmentation capabilities, you must first lay the groundwork:

  • Identify your most valuable customers. Take a look at purchase histories, recency/frequency/monetary statistics and socio-demographics to create a handful of profiles of your customers segmented by their value. What are the common traits of your best customers? This analysis is the foundation of your real-time intelligent call routing program and should be deep and granular. This analysis can take anywhere from a day to a month, depending on the size of your customer base and the depth of data you possess.
  • Identify your best salespeople. Rank them in a way that parallels the profiles you developed in the customer analysis step—say, gold, silver and bronze.

Now you need some technology pieces that conclusively identify callers in real time, match callers with the customer profiles you have developed, and route the callers to the right salespersons. These key pieces are:

  • Advanced Caller ID to automatically confirm name, address and phone number—information that unlocks demographic and lifestyle data—all in milliseconds. For existing customers, you can add in purchase history and anything else you know about them. Advanced Caller ID is available as an online service and doesn’t require a software systems upgrade. You can activate it while you’re analyzing your customer base.
  • Automatic profile-matching technology that captures the traits of new inbound callers and compares them against those of your gold, silver and bronze customers, segmenting callers on the fly. Your best customers may be highly educated 50-somethings who rent lofts in Sun Belt cities with populations above 500,000. You need to know if a caller fits the profile before you pick up the phone.
  • Call routing. With action codes assigned to each profile segment, callers who fit the gold profile go to the gold salespeople. The silvers go to the regular team. The bronzes might go to the overflow team. Sales scripts can be tailored for each segment. Companies can also segment and route callers based on the products a consumer would be likely to buy. Sort them out before the agent says hello and route them to agents with the right training for that product.

Although all this sounds like a huge undertaking, the bulk of the effort is the initial customer analysis. The quality of that analysis will dictate the quality of the solution and, ultimately, customer satisfaction. The matching and routing pieces are more straightforward software components.

One major PC manufacturer puts these tactics to good use. A consumer who has never phoned the PC manufacturer calls in. The enhanced Caller ID system automatically serves up the number, name and address before the call connects, without any dialogue or button-pushing. That information triggers a wealth of commercially available demographics and lifestyle intelligence, including credit rating, income and household makeup. The system matches the data on the caller with the coded customer profiles, directing the call to the right agent—all within 400 milliseconds.

The system may identify a first-time calling consumer as a 30-year-old professional from New York with an advanced degree, a $120,000 salary, a penchant for sailing and no debt and then route him to a gold agent in the consumer division ready to pitch high-end home desktop computers with full multimedia capabilities. The agent knows only that she has a gold-level customer on the line who is likely to want the high-end system and can afford it.

The PC manufacturer handles 10 million calls a year this way via several dozen call centers in the U.S. and overseas. Call data is cached for 60 days, so it’s easily retrievable, and underlying contact data is updated daily. Accuracy has been outstanding. The system scaled though the 2007 Christmas peak without a problem. In the past, the company randomly assigned callers to agents who spent an inordinate amount of time identifying callers. The wrong agents made the wrong pitches to the wrong consumers.

These agents are no longer flying blind. They are marketing selectively to identified consumers on a foundation of sophisticated, relevant and actionable information.

Real-time use of segmentation is on course to get even more sophisticated. The next generation of actionable analytics will match callers to customer care agents based on, say, the customer’s psychographics. You don’t have to look into the distant future to see this kind of improved customer satisfaction—and business performance.

Robert Nascenzi
Robert Nascenzi is the chief operating officer for TARGUSinfo. He was previously the president and CEO of Claritas.


  1. Robert

    Great post.

    As someone educated at the school of information-based CRM, I found your article very interesting. But it raises the question of whether the agent on the Gold line should be selling the Platinum rather than the Gold card just because the customer was eligible. The key question for me is what product was best for the customer, not what was best for the company in the short-term. There is plenty of evidence that aggressive selling whilst value-creating in the short-term is value-destroying in the longer-term, as customers are mis-sold the wrong product, become quickly dissatisfied with it and defect. And that isn’t even including more recent sub-prime mis-selling disasters!

    As call centres are finding out, it is high-EQ agents that are usually best at making this trade-off on behalf of customers and at successfully closing the transactions that analytics and technology makes possible. As much time needs spending on the selection and training of teh right agents as on on the bells and whistles that surround them in today’s call centres.

    Graham Hill
    Independent CRM Consultant
    Interim CRM Manager


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