Among the major mistakes that fixed-line and mobile telecom operators have repeated has been to chase big technology ideas and forget about the customer. There is an engineering-driven “build it and they will come” mentality that often puts technology and product ahead of customer and marketing. But the technologies being put in place today—like IMS and content delivery systems—will rely on a much more sophisticated understanding of customers and their behavior in order to deliver the revenue their proponents promise.

According to standard industry rhetoric, content services—which are costing carriers billions of dollars to prepare to deliver—are all about niche marketing and personalization. However, service providers generally have a poor understanding of who their best and worst customers are, and what makes them tick.

Understanding Profitability

Telecom operators do not lack data. The lifeblood of their business is data—from billing and provisioning to customer interaction and repair—and the fact is, they have more data than they know what to do with. Most major operators are in the early phases of learning about or performing useful analyses on data to generate customer and business intelligence. For many service providers, customer profitability analysis seems to be the first logical step. “If you let everyone in, how do you insure profitability? You can’t. This has to be very targeted, and traditionally carriers have been lacking in that area,” says Jay Bowker, vice president of global sales and marketing for Coastal Technologies.

Understanding which customers are most profitable is critical for several reasons. First, it’s just good business to know not only which customers are most profitable, but what characterizes those customers so that one can go out and find more of them. Second, profitable customers should be the ones spending the most money, which means they are most deserving of loyalty rewards—if such incentives are offered. Third, and perhaps most important, is that if a carrier understands which customers are profitable, then it will also understand which customers are not profitable. Put in simple terms, there’s no reason, for example, for a wireless carrier to subsidize a new phone just to get a customer into a two-year contract that is guaranteed to lose money.

Capturing this basic knowledge about each customer and feeding it into the customer care channel is not done consistently or universally. “I would agree that most carriers don’t do a great job of profitability analysis,” says John Georgesen, senior director of decision sciences for Convergys. “It’s because it involves pulling together data from many different parts of the enterprise.” Once again, organizational and IT disparity contribute to a lack of visibility across the business and into the customer, particularly in the large North American incumbents (see sidebar, “A Look at Europe and Asia,” p. xx).

The key here is the level of detail involved. Profitability analysis is conducted, but only at a high level with summary data. “All carriers conduct profitability analysis,” says Susan McNeice, director of marketing for Vibrant Solutions. “But the heritage of a lot of the systems creating profitability data is financial. These reports, or outputs, are highly summarized.”

The problem with using summary data to drive niche marketing it that it lacks the detail necessary to identify and target specific groups or behavior types. “When you summarize or aggregate at various levels, you lose another layer of nuance,” says Phil Francisco, director of product marketing for Netezza. Part of the reason carriers have relied on summary data, he argues, is a basic lack of technology that could deal with billions of CDRs, customer interaction data points and other elements. Just because data warehousing technology seems to have caught up doesn’t mean carriers are using it or have mastered it yet.

However, the problem isn’t necessarily systems-oriented, but also ties into what disparate systems reveal about disparate organizations. Carriers often “don’t have the internal relationships to really understand overall usage and overall costs,” says Bowker at Coastal Technologies. “We have a heck of a time getting all the teams together to understand all of this.”

Determining profitability isn’t a basic equation, but it’s not rocket science either. In addition to all of the estimated per-subscriber costs of acquisition, management and support are interconnect costs and other potential cost factors, like how much time the customer spent on the phone with a call center. The data points are finite and already exist in various databases around the enterprise; however, they simply aren’t brought together consistently. Mobile and fixed-line operators “do a good job of cranking out bills, but we haven’t seen real in-depth analysis of … who are the most profitable customers under contract, or of those who are coming up to the end of their contract that the carrier should retain,” says Bowker.

Changing the Mobile Model

The mobile market provides some good examples of the dilemmas that operators increasingly are facing. First, offering a mass-market service with a legacy that stresses customer acquisition results in something like an open-door policy. Even when limiting credit risks, the open-door policy gives the same treatment to all customers, regardless of their relative value. In a market where the operator is floating most of the cost of the customer’s handset just to win their business, this relatively indiscriminate approach means the operator is playing catch-up from day one with every customer and is unlikely to make its money back in many circumstances.

Now, this approach wouldn’t be such a problem if an increase in usage necessarily meant an increase in revenues and thus helped to offset the up-front costs. But it doesn’t. Because of intense negative price pressure, revenue is declining or remaining flat while traffic multiplies, meaning it is becoming more difficult to recapture the up-front handset expense without steady revenue from uptake of value-added services. A more discriminating approach would seek to determine which customers are worth the up-front investment, which ones a carrier can take or leave, and which ones are wholly undesirable.

By failing to understand customers’ relative value and offering every customer essentially the same subsidies, operators put themselves in an inflexible position from the start of the relationship. They’ve given away all of their promotional dollars just to win the customer, which means there is little left to give away to help drive new service uptake. Instead of viewing follow-on promotions and free trials and loyalty tactics, they are vilified as revenue cannibals.

“The fear of revenue write-down is a killer of many new ideas,” says Duffy Mich, CEO for Aperio CI. “Operators cannot take the risk of giving something else away, like an added promotion, because they are just writing down revenue,” he says. This particular approach to economics has created a mindset among carriers where almost nothing is given away for free or on a trial basis after the handset subsidy. Customers are not rewarded for their loyalty—or at least not rewarded in familiar ways, such as with redeemable points or airline miles.

The move to content alters the competitive playing field quite a bit and puts mobile devices in competition with consumer electronics products, like Apple’s iPod. Apple has rejuvenated its business with devices that cost anywhere from $75 to a few hundred dollars. Consumers have willingly spent the money to own them and they seem to be visibly prevalent. Mobile operators like Verizon Wireless have taken notice and attempted to jump on the bandwagon, but with their traditional subsidy model.

Verizon Wireless’ new Chocolate device, for example, is a mobile phone and an MP3 player, “but they got it wrong,” says Mich. For starters, he says, it is very difficult to put one’s own music on a Chocolate, though it is easy to buy music from the Verizon Wireless music site. “People are not going to buy this phone just to then buy music they already own,” he says. Verizon Wireless “is thinking about the data usage, not about the music or the customer.”

A better approach would be to recognize the consumer electronics market for what it is and sell the Chocolate as an iPod competitor to customers who are identified as those most likely to want one. “Charge an iPod price for it, make it easy to put music on and to buy more, but charge for the phone to relieve the handset subsidy on the marketing group and free them to be more aggressive in other areas,” says Mich. Now, asking mobile operators to change the fundamental models and economics of their business might be asking a lot, but the point is that other models that utilize both customer and market knowledge need to be considered in the new content business.

Without becoming too fancy too fast, Mich suggests that carriers start simple in order to become smarter and appear more customer-savvy. The place to begin, he says, is in analyzing and understanding existing customers’ behavior. “Being more sophisticated actually means thinking about the basics, thinking smaller,” says Mich. “You need simple analytics based on a handful of values. You don’t need thousands of data points to do this.”

For example, customer churn is something all carriers spend time and money trying to combat. Often churn is fought with win-back campaigns and other promotions after the fact, but without an understanding of the factors that are driving high churn rates. “I believe most phone companies don’t have a clue why people leave them,” Mich says. “They think they know, but … when I look at the data and measure what types of customers are with them, I find in the wireless space that overwhelmingly the people who leave are people who had an older phone and were paying far more than they should have been.”

The mindset that keeps a customer on an aging phone and paying too much for service also stems from the fear of revenue write-down. Operators are often unwilling to offer proactive plan advice or new phones for loyalty because of the costs involved. Adjusting all customers to optimal rates plans can reduce revenue in the short term. Giving away a new subsidized handset means the whole cost recovery game starts all over again.

Yet if customers are going to be encouraged to do more with their mobile phones, then they need to have the right device in their hands, and the right incentives in place to encourage them to try new services—most likely without having to pay for the first few tries. Making this model happen will squeeze some of the economics of the mobile business. It can work, however, if operators know the relative value of their customers and thus exactly how much, if anything, they can afford to give away to win their business.

Is Loyalty Valued Anymore?

In a world where companies are cutting pensions and benefits; contracts are broken without a second thought; celebrity athletes play only for the highest bidder; and politicians change their stances with every shift of the wind, it’s worth asking whether loyalty means anything anymore. Without dissecting the cultural fabric, one thing seems clear: loyalty counts when it comes to spending money.

This is where telecom needs to pull its head up out of the wiring and pay attention to other industries. Consumers are now habituated to loyalty programs. People go out of their way in many cases to earn loyalty points and in doing so stick with their preferred providers. “I will take a connection to fly United even if there’s a non-stop on another airline, but my loyalty to my telecom provider is pretty loose, and no one is trying to come up with reasons for me to stay,” says Steve Bamberger, vice president of communications, media and utilities for Oracle. “I think carriers are putting too much stock in triple play being the thing that keeps people around.”

This is one of the great mysteries of the content business—the idea that loyalty is fostered by getting a customer locked into a multi-service package. Technology is making it easier for customers to shift from one carrier to another, and offerings are bound to be similar from carrier to carrier. As a result, loyalty has to go beyond the multi-service offering and live up to or exceed the expectations set by credit card companies and airlines. Currently, the most common loyalty programs that mobile operators offer are simple referral schemes, which promise one-time billing discounts for each referred customer, up to a specified limit. These programs stand alone, however, and are not associated with any sort of all-encompassing loyalty program. They don’t combine referrals, for example, with specific discounts or trials on value-added services. They are neither sophisticated, nor well advertised.

Moving Up the Analytics Chain

Understanding customer behavior to prevent churn is important, but what will be more important is having the tools in place to understand changes in customer behavior as a response to specific stimuli. “Operators need to get marketplace feedback in rapid order. Let’s say I’m putting a promotion out there for 90 days on which I won’t make a whole lot of money. I want data in 10, 20, and 30 days to know if there’s uptake. If it’s wildly successful, I may want to extend the offer, but I also want to know if it is successful for the reasons I think it is successful” before taking that step, says Vibrant’s McNeice.

The ultimate end state is a model that can account for product, changes in pricing, changes in bundling, stimuli like promotions and up-sales, and other factors that will influence demand. In extremely short launch windows and product life cycles, it will become critical for carriers to have rapid feedback from the market to optimize the variables that drive service uptake and thus to maximize revenue from any given offer or bundle.

Satellite providers have been extremely aggressive in their customer acquisition and retention offerings, particular around exclusive content like NFL Sunday Ticket. While the offers to this point have been largely price- and bundle-oriented, satellite providers are making investments in “real-time decisioning technology to operationalize sophisticated offer management, but they aren’t quite live yet,” says Oracle’s Bamberger. In this case, rather than a difference in technology, it’s a simple difference in mindset. “I am impressed with how aggressive satellite providers are with their offers. They’d rather break their systems but get their offers right. It’s a business mentality, not an engineering mentality,” he says. Often it is an engineering-centric mentality that causes traditional telecom players to focus on technology first, and customers and marketing second.

Staying within the current mindset, it is possible to get carriers on the road to better customer knowledge, and once again it starts with churn management. “If we were to design a customer analytics or retention analytics focus that is broad and encompassing, we’d say there are three things to look at,” says Georgesen at Convergys. The first, he says, is to use traditional billing and usage data to create profitability models and to determine which customers are most likely to churn. This will highlight which subscribers or prospects are worth going after because they are profitable but also likely to leave. “You can’t neglect that in your strategy,” he says.

The second step is to bring in demographic information that helps to create a better picture of what customers look like, so that they can be segmented in ways that are meaningful to continued marketing and customer support. It may also mean reaching into areas of the business that don’t typically get involved in churn analysis. “One key to a recent engagement was to move past traditional sources like billing, usage and customer satisfaction data and pull in provisioning, repair, and information about what happened during some of the customer save calls,” says Georgesen. This step doesn’t come without a price, however, and he admits that “there are technical challenges,” that it requires “a good deal of perseverance,” and that one needs “buy-in from high-level stakeholders to make it happen.”

The third and perhaps most important step, born from the aggregation of all of this enterprise-wide information, is to understand customer behavior. In terms of churn it is a matter of understanding why customers leave. “You have to pull in customer intelligence to understand what customers need in that particular segment and how satisfied they are, because that gets to the root causes of why they are behaving in certain ways,” Georgesen says. Not surprisingly, the best way to understand why customers behave certain ways is to ask them. “You have to reach out to them to understand why,” he says.

Behavior and Demographic Data

Georgesen’s second and third steps do not go uncontested. While it is agreed that both behavioral and demographic data can be useful, not everyone agrees about which should come first and which has the most value. “I have yet to see the use of demographic data work very well in the phone business,” says Aperio CI’s Mich. “Most implementations have been a complete waste of money. It works well in marketing automobiles, catalog sales and lots of other things, but … I haven’t seen anybody use that demographic data effectively and generate a high rate of return.” He explains that at best he’s seen $1.15 returned for every dollar invested, but not consistently enough to make such a minimal return worth pursuing.

The problem may be that typical demographic data doesn’t apply well enough to communications usage. “There doesn’t seem to be any linkage between spending pattern, usage pattern and demographic information,” says Mich. But arguments that telecom is too unique to learn from other industries’ methods generally do not hold water. What’s more likely is that the cart is being put before the horse. “I wouldn’t spend any money on socio-economic data until I understand the behavioral data I already have,” says Mich.

If demographic data is only useful if it’s combined with the right behavioral data, it would flip Georgesen’s second and third steps. Georgesen agrees that once carriers get to a level of true personalization and complex one-to-one marketing, “demographic data might tell the characteristics of the area in which a customer lives, but it doesn’t say much about him.”

Mich agrees, and suggests that while he and his two next-door neighbors would be grouped into the same demographic code based on the socio-economic rating of where they live, their communications habits could not be more different. He says that while he himself is a heavy voice user, he uses almost no data services at all. One neighbor, however, uses voice and data services constantly from three different devices, while his other neighbor has a cell phone purely for emergency use. These three users are radically different, but would be grouped into the same demographic. All the same, it would be the demographic information that pointed the carrier to the right neighborhood in the first place to turn up two of three people who are likely profitable customers.

Other Third Party Data

Sometimes the term “demographic data” is used too loosely, and what people are really talking about is various forms of third-party data combined with the aggregated data operators already have. “Consumer product retailers know who, what and where, as well as the circumstances and factors that drive customer behavior,” says McNeice at Vibrant Solutions. “Wal-Mart is the king at this. They understand not just who is buying and what is being sold per store, but they also know what is being bought, what combinations are being bought, and if there is a causal relationship between things like weather and what’s being purchased.”

Third-party information relating to events like weather, holidays or sports championships can reveal new information about customer demand, but that event information itself is not specifically demographic. Setting the semantics aside, the point is that the analysis of various forms of third-party data in conjunction with behavioral information will reveal “how different factors converge to drive consumer behavior,” says McNeice.

In the end, the good news for telecom carriers is that most of the processes behind gaining strong analytical knowledge of customers are scientific and mathematical. “Carriers are not unaccustomed to living in that world, and they can do well exploiting the math relationships between these data sources,” says McNeice. But the analysis can happen only when the right data is accessible. This means that carriers need to improve in two key areas to make the math work: organizational communication and customer communication. Until organizations are willing to share data and conduct regular fact-finding outreach to real customers, all of the math in the world won’t come up with the right answers that encourage customers to increase their spend and their loyalty to one carrier.