How to make better use of internally available customer data has been a hot topic during the past few years. We are now seeing a variety of offerings from vendors that propose new ways to monitor, manage and monetize customers using existing data, including data previously visible only to network and IT departments. This document is intended to provide communications service providers' (CSPs') marketers and strategists with a summary of use cases and examples of successful implementations where deployment has advanced beyond the trial phase.

• CSP marketers are starting to access a variety of new types of data from network and IT systems — for example, mobile broadband usage, more in-depth views of content, application and Web usage, and maps of consumer calling circles.

• This new data is being used partly for traditional purposes — such as segmentation, calculating churn propensity and making tactical marketing decisions — and partly for newer activities such as:

  • Monetizing CSPs' core functionality of network provision — for example, devising pricing and other offers in relation to quality of service.

  • Creating more complete views of consumer usage in order to negotiate with other participants in the ecosystem, such as content providers.

  • Providing more personalized offers.

• CSPs are keen to analyze customer usage patterns in order to improve pricing and product decisions with a view to monetizing their data networks and improve their time-to-market in the face of new competition.

• CSPs' transformation initiatives should take into account the shift from network-and-product-focused activities to entirely service-and-customer-focused initiatives. Organizational barriers need to be removed to achieve this. Technical and commercial customer data must be shared across network, operational and service-marketing departments.

• Vendors should present clear use cases to help marketers navigate the wealth of insight. Offering a list of all available data and uses of that data is less helpful than providing a shortlist of the activities that yield the highest return on investment.

Customer insight is held up as a key differentiator for CSPs fighting new entrants coming from outside traditional telecom markets.

The solutions discussed above are being developed by established players in the world of customer insight, such as billing, business intelligence and network equipment vendors. But new types of niche vendor are also addressing this area by developing new techniques for gaining customer insight, such as personalization and SNA, or providing deep packet inspection, which gives CSPs access to new types of customer data.

The document concentrates on the newest uses of customer insight data. It leaves out the more widely used solutions that, we believe, most clients will already be familiar with. Also, the list of vendors providing solutions shown in each table is not intended to be comprehensive, but to show those companies that we have spoken to recently about each type of solution. Lastly, to contain the document's length, it does not include a small number of specialist data types, such as data on customer usage of channels.

Gartner's "Dataquest Insight: Superior BI is the Key to Communications Service Providers' Future Competitiveness" gives a good overview of the importance of customer insight to CSPs. It identifies their marketing and strategy teams as being especially in need of this insight.

For the past 18 months Gartner has collected examples of new ways to use customer insight. This document brings them together in a single document designed to:

• Summarize what customer data is available to marketers and strategists, and identify how this data can be used by CSPs' marketing and strategy teams.

• Help these teams find vendors of solutions that interest them, so that they can investigate further.

Table 1 shows the categories of customer data examined in this document. This data can be captured in real time or near real time for immediate use. Alternatively, it can be stored to provide a historical view of usage for a particular period.

There are multiple reasons for marketers and strategists to use customer insight. See Figure 1, which breaks down the marketing and planning processes into various areas. In addition, the chart includes some of the other main uses of customer insight to highlight where delineations are made; for example, customer insight data already generates significant benefits for the network operations and finance teams — however, analysis of these is beyond the scope of this document.

Below we look at how data is used in these different areas of marketing and strategy, and give examples of the newest or most interesting use cases.

Since CSPs lack knowledge of individual customers until they are acquired, there are a limited number of examples in this category. Most uses of customer insight relate to segmentation of potential subscriber groups using information about the devices, services and consumption patterns of other customers who are thought similar. Table 2 gives a few examples of the use of customer data for acquisition activities.

Example from LGR Telecommunications: Dialed-number (or "B-number") analysis collects a list of all the numbers dialed or messaged by the CSP's customers, along with information already available about the size of the CSP's customer base, to calculate the CSP's market share. LGR Telecommunications provide this to a large U.S. CSP — enabling it to calculate its competitive position with a confidence level of about 99% — a level significantly higher than is achieved through customer research. Also, the information — market share, share of gross or net additions or churn, share of disconnections and other measures — is available at a granular regional level for detailed insight into what effect marketing campaigns are having. All CSPs operating in very competitive markets would benefit from this sort of tactical information, but it is most viable for those with a large-enough market share to obtain a full picture from dialed-number information. Smaller CSPs will have less confidence in this type of analysis, as their customers do not interact with a large-enough proportion of the market's consumers.

Example from Xtract: Xtract is looking to incorporate location data into acquisition targeting. Location tracking can improve a CSP's predictions of attributes such as age, gender, ethnic group, family type (single or family), income level and the areas where subscribers live. This is useful in the prepay market, where little is known about customers, and it enables more tightly targeted acquisition messages to be created. For example, if a store is in an area with a high concentration of a particular ethnic minority, there are opportunities to create local-language advertising for that ethnic group or to offer appropriate international-calling deals.

There are more examples in the area of customer retention (see Table 3). Customer insight data has been used for many years to increase the accuracy of churn propensity scoring, best-plan advice and loyalty programs. But there have also been moves in the past 18 months to make personalized offers and incorporate social network analysis (SNA).

Example from Scorecard Systems: Rotational churn can be caused by dealers cancelling customers and then re-signing them to create additional acquisition payments, and by customers not paying an old account and then taking a service as a new customer. Traditionally, CSPs have relied on seeing paper identifications from customers and other investigative techniques to mitigate this problem. Scorecard Systems offers CSPs a way of using SNA and location information to examine calling patterns with a view to identifying rotational churn. Its product looks to see if a new customer is calling the same numbers from the same location as an older customer — and then flags the new customer as a potential rotational churner. The CSP might then deny the customer service (if legal), ensure that this issue is reflected in future retention-related decisions for this customer, or take action against the dealer.

CSPs are experimenting with a variety of ways to personalize offers. Retention of prepay customers (as mentioned above) and increasing these customers' top-up frequency are the two main drivers of these activities as prepay offers are relatively simple to create (they do not require complex datasets to be collected as for postpaid offers) and seem to yield good results. However, some vendors are now moving into postpaid cross-selling and upselling, as Table 4 shows.

Example from Pontis: Pontis has recently branched into postpaid cross-selling and upselling. Products enable upselling of minute bundles and targeting of customers without data packages or those with low data usage who have handsets capable of supporting data services, and stimulation of data usage when a mobile broadband customer has a pay-per-day data plan. Pontis claims to have seen increases of between 15% and 40% in average revenue per unit (ARPU) for customers responding to its offers (for which the takeup rate is about 5.5%).

Customer insight data can be used to provide customers with information that will educate or otherwise help them (see Table 5). A CSP might, for example, send a brief help guide on how to use the Multimedia Messaging Service (MMS) to customers who have just taken a photo for the first time on a new phone but have not yet sent an MMS message.

Example from BroadHop: Given the pressure placed on bandwidth by data services, BroadHop offers CSPs a solution that educates customers about bandwidth availability and gives them the opportunity to take particular courses of action. For example, CSPs could create a type of loyalty program whereby consumers in highly congested areas can upgrade their network connection temporarily by using accumulated loyalty points or downgrade it voluntarily (freeing resources for other customers) in exchange for loyalty credits.

During 2010 we have seen a large number of vendor products designed to use customer insight data with a view to moving customers away from "all you can eat" mobile data plans, especially in countries where data consumption has reached network-affecting proportions. Table 6 shows a variety of pricing options that use customer insight data, according to the vendors we spoke to.

Example from Informatica: An international CSP is aggregating call detail records including information on calling and receiving numbers, call times and durations, results of calls, call routes and call types (voice, SMS, etc.). The ability to analyze customer usage with call detail record granularity enables it to devise competitive packages and other offerings for targeted audiences. It has also reduced the cycle time for defining, marketing and pricing packages by 80%, from several weeks to a matter of days.

Example from Allot Communications: This provider of deep packet inspection offers CSPs a form of time-based pricing for mobile data tariffs. It provides a "happy hour" each day during which consumers receive upgraded bandwidth or free downloading. This encourages customers — particularly heavy users using peer-to-peer applications — to move away from the busy hours for mobile data traffic.

In the last two sections on use cases for customer insight data, we look at activities in which marketers are involved, although responsibilities are also shared with other departments.

This first area looks at the planning of user experiences (see Table 7). What services and applications should be delivered? How are those already in use performing? What more can be done to improve users' experience of devices, services and applications?

We talked to a variety of vendors in this area, some of which provided information derived from network equipment and some from clients installed on the devices.

Example from Clarabridge: A large U.S. CSP is provided with a detailed look into the activities and interactions of its customers. Clarabridge processes structured and unstructured data from post-transactional surveys taken 24 to 48 hours after a customer interacts with the CSP via call center, retail (face to face) or the Web. Information from these sources are processed and insights are delivered via a portal to call center operations, product managers, the marketing department and senior management. These insights are used for enhancing product and service offerings and other non-marketing uses, such as improving call center efficiencies/training and providing the executive team with a better view of customers.

Finally, customer insight data is used to aid negotiations with third parties, including content providers, application developers, advertisers and media agencies. Closely related to this area is the provision of customer insight data to these third parties for their own purposes (audience monitoring, advertising and so on), but we do not include these uses in Table 8 as we believe them to be too far removed from CSPs' core marketing uses.

Example from Openet: Openet gathers viewership data, including all linear TV, video-on-demand, pay-per-view and advertisement consumption. This is used by Time Warner Cable in the U.S. to negotiate its billion-dollar content purchases with providers each year. Gathering viewership data directly from the network is preferable to sampling because samples are too small to be statistically relevant for all time slots and channels For example, with traditional sampling methods, only about 100 of 400 channels are measured and about 30% of primetime viewing goes unmeasured. Increased accuracy enables Time Warner Cable to negotiate content costs with content providers more accurately — and to achieve cost savings.

These solutions should have several impacts on CSPs' marketing teams:

• They should be able to make tactical decisions quicker and using better insight. However, traditional bottlenecks, such as IT's capacity to implementing new pricing plans and supporting new products remain an issue.

• Marketers will have a dilemma — new types of personalized offers will be possible, but exposing what they know about customers risks attracting bad publicity if they step over individuals' "privacy thresholds."

• By using these new solutions for their own internal purposes, marketers will gain a better understanding of the uses of customer insight data and of how they might package it for sale to third parties (e.g., advertisers, application developers, enterprises and content providers).

• The market for customer insight solutions will show differences by region and country, depending on the maturity of the market, the needs of customers and regulations regarding the use of customer data.

These solutions will also affect CSPs' network and IT requirements in a number of ways:

• As there was little discussion from the vendors of bringing the two main streams of insight — more traditional billing insight and newer types of insight from networks — together in one solution, those involved in providing customer insight to marketers will have to work out which solutions provide the best datasets, given the overlap of data fields. This may well be a complex decision, as the return on investment will depend on the benefits of each individual marketing activity that could be undertaken if particular pieces of data are available.

• One solution will be provision of master data management functionality both in the traditional business intelligence environment and to enable real-time analytical decisions to be made.

• The amount of database storage capacity and computing power required by marketers looks set to rise significantly.

Our investigations illustrate that CSPs' marketers remain focused on the more traditional uses for customer insight, although this insight now includes data that they have not had access to before.

From vendors providing information from billing records, CSPs are looking for new types of insight on usage to improve segmentation, reduce churn and make smarter tactical marketing moves.

From vendors providing insight from network equipment, marketers are looking for insight into how the network is being used — insight to which they have not previously had access. They also want the ability to monitor and implement fair-usage policies.

Early-adopter CSPs are also deploying new products for marketing and strategy that use customer insight data to:

• Conduct more involved analysis of customer usage patterns for improved pricing and product decisions.

• Examine ways in which their core functionality for providing networks can be monetized — e.g., through pricing of quality of service, richer service-level agreements and ways to reward customers with improved network quality.

• Negotiate with others in the ecosystem, aided by new knowledge of network usage.

• Provide more personalized offers.

Lastly, some of the vendors mentioned the use of contextual information such as location data, but as yet there are relatively few concrete examples of this usage.