The answers about customer profitability exist within the information that providers already collect within their back- and front-office systems, says Daniel Kenyon, PeopleSoft’s vice president of communications industry strategy. “I truly believe they are measuring this information,” he says, “but what they aren’t doing is sending this information to an analysis engine.”
Carriers are just now starting to look more toward analytics to leverage their existing business information. “It’s a fairly recent phenomenon in the last year and a half,” says Richard Wolniewicz, vice president of engineering for analytics at CSG Systems. While he says the financial industry is farther ahead than telcos, many telecommunications providers are more advanced than other industries.
“ It’s happening now in some carriers but not in others,” Wolniewicz says. “It’s certainly larger carriers that have focused efforts on this.” These carriers are first looking to have the analysis in place and then take action on that analysis, he says. They are looking for ways to operationalize the information and make that information actionable by CSRs (for more on AT&T’s use of analytics, see “ProviderCase Study: AT&TMeasures Customer-Level Profitability”).
One problem many providers face when measuring profitability is the fear that they might take an incorrect action. “Carriers are very concerned about how they treat their customers,” Wolniewicz says. He has observed first-hand that carriers are very concerned about making sure an analytics product does not undo values the provider has already assigned to customers. For this reason, vendor products should include an opt-out feature for particular customer segments, such as high-value subscribers.
Customer Value vs. Profitability
Subscriber profitability and value are two attributes that providers want to be able to measure, but they are distinctly different. PeopleSoft’s Kenyon says customer value includes the predictive side, such as a subscriber’s propensity to purchase or use more products over time. It also often includes how many accounts and relationships a customer has with one provider.
But customer profitability reaches more into the financial and business aspects of a customer’s relationship with a provider. Both customer value and profitability are dynamic, fluctuating according to the changing nature of the subscriber’s relationship and service usage.
In the abstract, profitability is a simple equation of revenue minus costs. But for providers, determining true profitability is anything but simple. Providers must gather usage information and match that against the revenue generated. Kenyon says this is where providers truly face a level of complexity that most have not yet been able to deal with. The list of costs is long. Portal’s James Morehead, senior director of solutions marketing, says providers sometimes push out new services before they can actually bill for them, just to remain competitive. But they have to be careful about giving away these services, based on their actual cost to the organization. On average, Morehead says, it takes providers 8 to 9 months to charge for these new types of services. Lag time such as this must be taken into account when measuring profitability.
Other considerations include the cost of acquisition and subsidies on mobile handsets. “If you have a higher subsidy for the handset but you can’t charge for these services, you take longer to break even,” Morehead explains.
Commissions to sales agents and reseller channels add to the costs. And these are just the tip of the iceberg.
Subscriber churn is another factor that impacts profitability, as is package churn—when a customer signs up for a new package, but reverts back to an old package. With an annual churn rate of 25 to 30 percent, says Allan Koglmeier, vice president of Convergys’ unified customer management solution, “the economics are just absolutely brutal.” He says Convergys has tried a number of models to reduce churn with providers, but has had very little uptake. “They seem to consider churn a cost of doing business,” he says, and many providers dealing with churn don’t keep at it.
Network capacity used is another cost incurred, as it also affects network planning and build-out. As well, call center costs can certainly impact profit and make a profitable customer unprofitable, if customer care costs more than the revenue generated.
Revenue Assurance
Providers face one cost that often might get left out of the profitability equation: revenue leakage. “I think it has a negative impact on customer profitability that in many ways is hidden,” says Chuck Crenshaw, president and CEO of Connexn Technologies.
If the back office is not accurate, several provider costs can increase. Not only can billing mistakes, order fallout or provisioning problems directly affect revenues, but they can also generate calls into the call center, which translate directly into costs for providers. “Those calls are costly,” Crenshaw notes.
For example, if a customer receives a bill with errors, that customer’s calls to rectify the situation may end up bringing more cost to a provider than generating revenue for the month. “You can literally take that customer into an unprofitable state,” Crenshaw says, “just because of the cost you generate in the call center interaction.”
Then, if a customer short-pays the bill or just does not pay at all until the situation is handled, those costs could increase as days sales outstanding increases, or eventually top out if the account goes to collections. In many instances, this can result in churn.
PeopleSoft’s Kenyon says reconciliation between the switch and the bill needs to be considered in the profit equation. If the switch-to-bill reconciliation is typically leaking 2 percent, perhaps a provider could factor this into a customer profitability score to yield a more accurate picture of profitability. Perhaps the provider could plug in a constant to account for this leakage to subtract from overall profitability—at least until the leakage problem is fixed, or until the provider can cross-sell or up-sell a customer off an old billing system, if the old billing system turns out to be the reason for leakage.
Accurate analysis also becomes a challenge, to say the least, when information about costs and revenues for each customer comes from multiple billing and OSS systems. Miriam Deasy, mobile business marketing manager for Amdocs, says having multiple billing systems is a drain on profitability, as they increase the cost of maintenance and hinder the gathering and analysis of information housed in disparate systems. “Given the composition of many billing environments today,” she explains, “that data extraction is a tricky job.”
Service Profitability
If a new service has high uptake, that by itself does not mean the service will be highly profitable. Different services offer different revenues, and the cost to deliver and support the service may cut sharply into its profitability. And bundles of services can yield yet other profit margins, Morehead notes (see “Customer Profitability Analysis by Service” p.20).
A text message, for example, may have a 90 percent profit margin for the carrier, versus voice, which may yield 50 to 60 percent for the carrier, says Morehead. A user is willing to pay a higher price for a shorter usage of the network to send a text message based on the value of message to the customer. Likewise, the customer may be able to say in a text message what might take a five-minute voice call to relate at an equal or higher cost, explains Morehead. Yet network costs are minimized.
Morehead notes that 23 percent of Telenor Mobile’s revenue that flows through Infranet comes from data and messaging services. These data services have a 90 to 95 percent margin for text messages sent, given that the price of SMS is pretty high compared to the cost of delivering it. Because of this, Morehead says, the profitability of Telenor’s prepay users has increased.
Several providers have even hosted contests to drive SMS traffic volumes up. Telenor held a contest to see who could type the most text messages in a certain time period. The end result was that usage increased, in particular among teenagers. In fact, some of Telenor’s prepaid messaging users are as profitable as high-end postpaid users. A prepay customer, for example, who produces $15 per month in revenues actually could be more profitable than a postpay customer who pays $40 per month for services that cost more to deliver.
Another highly profitable mobile service might include video clips. As Morehead explains, 15 seconds of a video clip might require 100 KB, which is roughly equal to one minute of voice at a 14.4 connection. One minute of voice may cost a customer 10 to 15 cents, and cost about 5 to 8 cents to provide. Customers may be willing to pay 25 to 50 cents to send a video clip, which is double to triple the profit of voice. Plus, if the end user creates the content, there is no third party to cut into the profit margin.
At the other end of the profitability spectrum are services like video calls, which require a lot of network capacity to deliver. For example, a video call may cost $2 per minute to deliver. The service requires a continuous network feed and still shows highly noticeable degradation of quality of service. A customer may only be willing to pay 50 cents per minute for such a call, whereas the cost to offer the service exceeds the revenue generated. Thus, providers must carefully weigh the necessity of offering the service, especially at such a low cost to the customer.
Morehead believes voice is really an anchor service. It should be the lowest margin service that a provider offers.
How to Treat Customers
Getting information about customer profitability is one thing, but using it to an advantage is another. CSG uses its system to help a provider look at a subscriber’s data history. A provider might look at 250 to 300 fields to determine which could be linked to the acceptance of a plan by a customer. Perhaps 50 fields can be statistically linked to whether a customer accepts a plan or not. Every week, the system could take a dump of the billing system data to understand how the customer relationship has changed and whether a customer might accept a plan or might need to be turned on to a new plan.
CSG also uses utility models to predict the expected network revenue if a provider does or does not take a certain action. Say rate plan A reveals a 10 percent chance for customer acceptance and rate plan B offers a 20 percent chance. If the customers who sign on to plan A actually stay with the plan longer than those who sign on to plan B, Wolniewicz explains, plan A might be more profitable over time.
Upgrades to high-speed data are a big push right now in broadband, Wolniewicz explains. In CSG’s work with Time Warner Cable of Desert Cities, CSG measured the propensity of customers to churn and identified the ones most likely to purchase additional services. Based on this project, almost 83 percent of the identified customers were saved from churning through retention strategies, and more than 95 percent of customers identified with a high probability to buy new services did. For every customer who purchased new services when offered, this strategy ultimately increased ARPU by $287.
Amdocs’ Deasy says providers can absolutely use profit metrics to shape how they treat their customers. She points to some providers, including T-Mobile and Virgin Mobile overseas, that charge customers to call the call center. She says some have a premium-rate number to call customer care, or else deduct minutes from a customer’s plan for calling the call center. Whether transferring the cost of customer care to subscribers will work remains to be seen; however, providers can save money by automating a lot of their customer communications or directing them to a Web site for self-care.
Coming Soon
PeopleSoft will announce its new customer profitability product in mid-year, prompted in large part by the financial industry, which asked for risk-weighted capital and activities-based management. Kenyon says the ability to measure profitability is really a marriage between the CRM and financial products. The telco industry is seeing this happen at a grassroots level now, he says, but it will become a greater priority for providers. “I think by the end of this year, you’ll see a lot of people talking about this,” he says. Two wireless companies have already expressed interest in speaking with PeopleSoft to address their interest in customer profitability.
CSG’s Wolniewicz predicts, “In two to three years, carriers that weren’t adopting this technology will find themselves at a disadvantage.”
| Provider Case Study: AT&T Measures Customer-Level Profitability While some say providers cannot measure profit down to the customer level, AT&T is doing just that. Working with Denver-based CRM software provider Eyeris, AT&T matches costs against revenues to determine profit margins as one element of customer value scores. What it set out to do was develop a tool for the desktop that was “simple, fast and actionable,” says Drew Glassman, consumer services financial director. AT&T began working with Eyeris’ EyeProfit software in 2001 and rolled out the product internally in late spring/early summer 2002. According to Glassman, the company already had a number of databases that serve CRM and customer targeting. But it wanted to take data from various internal sources to create a central repository of information from which it could measure such customer metrics as profitability and value. Although the system was implemented in the financial department, Glassman says it partnered heavily with the company’s internal information technology department for the project’s success. In total, 63 different data feeds run into the EyeProfit software. Each month, the financial and operational data crunch results in about 1 billion rows of information generated, consisting of about 200 columns of costs and 30 columns of revenues and drivers. The system tracks nearly 2,000 different products gathering information from about 40,000 central offices. Despite the actual amount of data gathered, Glassman says, this is a manageable amount of information. It took AT&T about six months to develop the methodology tracked in this product. AT&T constantly tweaks the information collected, sometimes finding different information feeds to cue into the system. “We have the flexibility to make releases on a real-time basis,” Glassman states. The output of the EyeProfit system is read-only browser-based, to protect the integrity of the information in the database. What Is Measured? The EyeProfit software serves as both a data warehouse and analytical engine. AT&T regularly provides information to Eyeris, which processes and stores the data. AT&T maximizes the capabilities of the multidimensional database to give three major views based on the customer information, the products used and the geography involved. To perform this analysis it gathers information from the billing and network traffic systems, as well as from customer care. In matching the revenues and the cost of providing a service to a customer, AT&T measures the number of minutes used by a customer, whether the customer called the call center and whether the customer has a change order in place. Glassman says there are 50 unique costs of goods sold. These incorporate the cost of delivering the service, such as access or connectivity with another provider’s network, and include services such as international calls, which would have a higher cost to deliver, for example. These costs also include network transport and operator services. Selling, general and administration (SGA) costs would include billing, database marketing, direct marketing communications, product management, telemarketing and general overhead. Dynamic Scoring Customer profitability is certainly not a static number. “We know the single-month customer value fluctuates,” Glassman says. Although he sees customer profitability as a bottom-line metric within customer value, and value being a multi-month metric, he notes that customer profitability and value can change from month to month based on usage costs for a subscriber. For any given month, AT&T sees in its analysis these changes, according to Glassman. The company adds up its revenues and compares them to the costs of goods sold and SGA costs, and pushes that information down to the customer. It can add in the acquisition cost and track a customer over time to see if what may be at first a loss because of acquisition costs turns into a profit as revenues generated from the customer catch up with and surpass acquisition costs. “ Until you get down to a customer-level profit, you are stuck [measuring profits] at a service level,” Glassman says. Having this data down to the individual customer gives the provider an even greater ability to alter its actions to affect the bottom line. While the company performed some modeling internally in the past, Glassman says the information gathering for this one database for analysis is vital. “There’s simply no substitute for the facts,” he says. Many in the industry say that providers have yet to employ this kind of data repository and profit analysis technology to actually effect change within their organization down to how the provider treats the customer. Yet Glassman says AT&T is implementing a coding system for its UNE-P business, so that when customers call in, they get prioritized in the call channel based on their profit score with the provider. This would involve scoring the millions of UNE-P customers over time, he says. The score could be used in conjunction with predictive modeling of how the customer acts in a given situation to determine how to treat the customer. Who’s Using It? Within AT&T consumer, various departments benefit from having this customer, product and geographic information coming from one repository. AT&T’s finance, marketing and CRM departments all use this information, generating various different views and reports from the one database to suit their needs. AT&T employs the product to assess product profitability and review overall programs, such as the benefits from loyalty programs. It also uses EyeProfit to help assess how best to treat its customer base in their call servicing. “ There are too many different ways to slice this information to have one metric that’s useful to everybody,” Glassman says. The different tiered reports can provide information to heavy users, such as analysts; to the product managers to have an accurate, customer-level view of how their products and services are performing each month; and on up to executive-level users. Those within the financial department can only benefit from understanding how the operations department is affecting costs. While every person who uses EyeProfit at AT&T may put a different spin on the information they receive, Glassman says it is important that they all get the information about the customer from the same data. |
