Mobile Data Intelligence: The Next-Generation BI
Mobile service providers struggle to match data usage to revenues. A new kind of BI can help.
by François de Repentigny
In recent weeks, there have been several high-profile pricing structure changes by major mobile operators (such as AT&T in the U.S. and O2 in the UK) in an effort to move their data subscriptions away from costly “all-you-can-eat” models. As application usage grows at an exponential rate, many operators are finding that data revenues are not keeping pace with demand. Although business intelligence tools typically come into play to address this issue, available solutions do not provide a clear understanding of where pricing structures need to be changed, and where they can drive the greatest revenues.
According to AT&T, mobile data traffic grew 50-fold between 2006 and 2009. Nokia Siemens estimates this number will increase 300-fold over the next five years; data revenues for operators are only projected to triple in the same time frame.
With limited insight into usage beyond volumes, however, operators’ options to date have been equally limited. Some are either increasing rates across the board or restricting how much data subscribers can use. Neither solution sits well with customers. In fact, these approaches could easily alienate their existing high-value subscribers as well as potential new users. While it is becoming increasingly apparent that pricing plans need to change with the times and the customers that pay for them, it is time to consider another approach to business intelligence.
The biggest challenge for operators as they struggle to match data usage to revenues is the lack of insight into subscriber usage patterns. For the most part, existing business intelligence tools only provide operators with a general understanding of the data traveling back and forth on their networks without the context needed to drive smarter marketing decisions.
One of the reasons for this blind spot is the fact that subscribers increasingly access data services provided by third parties. Until recently, mobile operators were able to control the customer experience via a dedicated portal, which meant ready access to subscriber usage patterns. The advent of vendor-agnostic application stores and readily-available navigation tools however, means that fewer customers are using mobile operator portals as their gateway to information and services. As a result, the information gathered through dedicated operator portals is far less representative of behaviour and, therefore, of less value.
A New Perspective
Mobile data intelligence (MDI) is a new breed of business intelligence that enables operators to move away from the walled-garden approach and capture usage patterns across boundaries. With MDI, operators can see beyond the “front-line” cause of increased data traffic (e.g., smart phones vs. dongles vs. feature phones) to understand the percentage breakdown of each in relation to overall data volumes, as well as breakdowns relating to services (e.g., browsing, Skype, messaging, music downloads, applications, etc.). Further analysis can help determine if there is a particular type of application or device causing the increase (or if it’s across all applications or devices) as well as how much data is consumed on a per-user basis to understand usage clusters.
With this information in hand, operators can then adopt a “smart pricing” approach by developing multi-tiered plans that address different users’ requirements while monetizing the portion of the traffic that was once unaccounted for without compromising customer satisfaction. They can build more accurate segmentation, refine their promotions to match the segmentation, and track performance. With better information, operators can develop optimized subscriber packages that combine device, voice, and data plans. The end result is greater returns on investment for every marketing dollar spent.
Mobile Data Intelligence
Mobile data intelligence is focused specifically on mobile data and enables operators to gain detailed insight into consumer behavior and data usage. It does this by enabling mobile operators to collect and link device, subscriber, and service-usage information through a single point of capture, without interfering with network performance in order to engage in more accurate segmentation.
MDI includes a wide range of behavioural variables such as service, application, content, time and location, and device preferences for all data usage – and not solely activities that terminate on the operator’s own network. Rather than using log-based systems across local architectures that focus on a limited spectrum of data, new generation MDI solutions can perform a “transparent” analysis of all transactions that flow on the mobile data network. This information is used to build powerful KPIs (key performance indicators). For example:
- For each service: number of users, data volume, number of sessions
- For each service: most used devices by number of users, data volume, number of sessions
- For each service: peak usage times and locations by user, data volume, number of sessions
- For each device: most used services by number of users, data volume, number of sessions
These “next-generation KPIs” can be used to build segmentation models that reflect the richness of the mobile data environment, instead of just a few services coming from a single provider. In addition, they capture detailed information about any service and application moving across an operator’s network, regardless of the source.
In addition to tracking fundamental trends such as traffic characterization per service, marketers can also gain insight into difficult-to-obtain trends such as device fleet distribution, including inbound roamers and grey market devices. Finally, these KPIs enable the discovery of hidden segments (for example, most suited devices for social networking, top application downloads per age group, etc.).
Consider this real-world example of how MDI has been leveraged to drive new revenue streams. An operator has seen an increase in mobile data traffic that is not matched by an increase in revenues and wants to determine if the increase is driven by feature phones, smart phones, or dongles (broadband wireless adapters). The operator determines if there is a particular type of application that is causing the increase, or if it is equally distributed across applications. In this case, the operator finds that P2P traffic accounts for the majority of the increase in dongle traffic.
The analysis is taken one step further to determine that the majority of non-P2P users usually consume less than 200 MB per month, while the P2P users typically consume a minimum of 5 GB per month. The operator introduces a two-tier pricing structure for dongle users – a 200 MB plan at $35, and 5 GB plan at $60 per month. This allows the operator to monetize the portion of its traffic that is straining the network without alienating its customer base.
The ability to track data usage and apply detailed segmentation is a key to mobile operator success. The more operators can better understand the link between incremental revenues and incremental marketing dollars, the greater the opportunities to generate new revenues and improved return on investment.
François de Repentigny is director of product marketing at Neuralitic Systems, a mobile data intelligence solutions provider that enable operators and communication service providers to manage and enhance their data services initiatives. You can contact the author at firstname.lastname@example.org.