Business Intelligence 2.0 – The Hidden Treasures in Cell Tower Data

A new and unconventional means for learning about one’s customers, cell towers capture unique information that can be used in groundbreaking manners by any mobile operator to better serve, market, and sell to its customer base.

You can download PDF version of this whitepaper here.

Customer-related business intelligence efforts in telecoms have traditionally relied on a standard set of information about the customer base, be it a BI effort around predicting churn, segmenting customers, identifying sales opportunities, etc. Normally analyzed categories of information have included (on a customer-level) those that are usage related (VAS, SMS, international calls, etc.), volume related (frequency, length, etc.), value related (ARPU, profitability, etc.), information that is then analyzed in light of the demographical uniqueness of each customer in an effort to identify patterns.

Most telecoms in developed countries have by now tapped in to their business intelligence to segment their customers, and to develop churn prediction models. Recently some have gone a step further and defined the social networks formed within their customer base, among other models they have developed with business intelligence.

What almost no telecom has done is to use the wealth of data captured by cell towers regarding its customer base, data that can not only be used to strengthen existing models, but to launch new ones. From helping telecoms decide where to open a new dealer to how to allocate its marketing communications budget, cell towers are the next great source of business intelligence for mobile phone operators.

Cell towers (also known as cell sites), for those who may be new to the term, are essentially antennas that are placed on top of buildings or are standalone structures that allow mobile phone subscribers to receive a signal when they try to make or receive a call. Each cell tower is able to cover a certain range and thus allow subscribers within that range to benefit from its signal. To cover the USA, for example, mobile phone operators have put up more than 250,000 cell towers, spread strategically so as to maximize their individual network coverage ranges.

Being within close proximity of a cell tower is in and of itself not enough to ensure receiving a signal however, as these towers have a number of channels available for usage at any given time – one cell tower along cannot cover an entire city block of Manhattan, for example, if half the subscribers on that block try to make a call at the same time (as is the case in emergencies when one has a difficult time getting a signal).

One final note about cell towers – when subscribers talk on the phone while they are driving, they will switch between them as the signal of one fades and the other strengthens. Thus, during the course of a 10 minute drive a subscriber may be switched between 5 – 20 towers, depending on how rural / urban the area they are driving through is.

When it comes to business intelligence, it’s not about how cell towers work, but more about what information they capture about subscribers who connect to them at any given time. Some of the key facts gathered in each and every single connection to the cell tower by a subscriber include:

  • Subcriber Phone Number / ID
  • Type of Connection (Voice, SMS, Data)
  • Time of Connection
  • Duration of Connection
  • Direction of Connection

In and of itself the above doesn’t seem to add up to much in terms of business intelligence. It’s when this information is combined with other key facts about subscribers that the true value can be obtained. Add in to the mix…

  • Subscriber Nationality
  • Subscriber Age
  • Subscriber Gender
  • Subscriber Value-Added-Service Usage
  • Subscriber ARPU
  • Subscriber Handset
  • Subscriber Churn Risk (if conducted)
  • Subscriber Customer Segment (if conducted)
  • Neighborhood by neighborhood population data (if available)

…and now numerous findings will be obtained that can be used by sales and marketing teams in telecoms to take action on. Five examples to start with:

1. Neighborhood-by-Neighborhood Market Share: As most telecoms have no idea where their subscribers actually live (especially in prepaid dominated markets), they have no way of knowing their penetration rates in a given city, let alone a neighborhood. By examining each subscriber’s key cell tower connection in the evenings and on weekends, a telecom can assign a general address to each customer, down to that given cell tower’s location. Doing this for every customer will give a telecom a general home address for each one, thus allowing it for possibly the first time to know its subscriber base city-by-city, neighborhood-by-neighborhood.

Combined with neighborhood level population data, a telecom can estimate its market share and penetration rate across the country. This data can be used by sales teams in innumerous ways, from determining where to possibly open a new dealer store, to deciding where to send door-to-door sales representatives (especially for the SOHO / SME segment). Information that would normally cost thousands to obtain via market research is already available thanks to cell towers.

2. Neighborhood-by-Neighborhood Connection Types / Time / Duration / Direction: Using the same principles above, a telecom can know down to a neighborhood level the usage behavior of its customers. In some parts of a city a telecom will find that local SMS is popular on weekends, in others, short voice calls abroad. Not only can such information be used to help design new tariffs and campaigns, but also to determine how to market down to a neighborhood level. The telecom can be positioned as the best when it comes to its SMS bundles in a neighborhood that cares for SMS, whereas in another the low off-peak international calling rates will be put in the forefront – this done in the form of billboards, flyers, in the dealers located there, etc.

3. Value / Churn / Location: Combining consumer churn likelihood and value, telecoms can identify which cell towers are critical in terms of retention – those towers with customers assigned to them that have high value / high churn likelihood are of the utmost importance. Studies have historically found poor network coverage plays a critical role in churn – the capacity of these mentioned cell towers must be examined to ensure consumers do not experience dropped calls in their home zones, or, even in their workday zones. Cell tower boosters or additional cell towers may need to be considered in cases where capacity issues are observed.

4. Neighborhood-by-Neighborhood Value-Added-Service Usage: Akin to connection types, analysis of the home zone of consumers will identify in which neighborhoods certain value-added-services are popular, have succeeded, or have failed. Such an analysis would thus again show the potential in any given area of the city for any value-added-service, and would determine marketing strategies down to a neighborhood level. For example, the neighborhoods that have the highest mobile phone data usage would be the best bets for putting up billboards for a new GPS Maps or Online Gaming application to be or recently launched.

5. Cell Tower Segmentation: Taking into consideration all of the above mentioned pieces of data and different observations, telecoms can segment their cell towers into actionable clusters, assigning them each a value based on connected subscribers, their churn risks and segments, their tenure, etc. Such an effort would help a telecom know which cell towers need to be prioritized when they go down, which ones should be upgraded to 3.5 technology based on those connecting, which ones are below-zero (have the lowest ARPU subscribers connecting to them) and can be maintained / consolidated, etc.

The above is really just the tip of the iceberg. Once a telecom gets the basics set up, the ways in which cell tower data can be used will increase, findings it way into the core of sales, marketing, and operations teams, becoming an indispensable part of their intelligence-driven day-to-day practices.

Cell towers also capture significant information regarding international customers roaming in the country, information that can be used in numerous ways to devise separate strategies aimed at getting the most out of each of them.

To learn more about starting your own cell tower data mining effort, please contact


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s