‘Context Mining’ To Go Beyond Customer Needs in Telecommunications

As telecom product and service offerings are increasingly becoming commoditized, telecom companies need to offer their subscriber bases more than just communications solutions. This homogenization of offerings necessitates telecoms to differentiate themselves, which can be done through understanding the lifestyles of their subscribers and adapting offerings to cater to these lifestyles – which brings us to the fairly new concept of ‘Context Mining’.

You can download PDF version of this whitepaper here.


Conventional (and even relatively innovative) customer analytics applications in telecommunications focus on how customers use their products and services, rather than in which context they use them. Although extremely beneficial, such a limited look into customer behavior keeps operators from understanding who their customers truly are, and how they can relate better to their lifestyles. ‘Context mining’ provides telecoms operators with a tool that lets them put these lifestyles under a microscope, by focusing on the context of telecommunications service used by customers (such as who they call, where they travel, what they watch, what they communicate about, etc.). Staying within the boundaries of customer privacy, operators now have to take this next step in customer analytics to create an environment in which they can become truly customer-centric.


In recent years, some leading telecom companies have begun catering to their customers in ways beyond just offering communications services; NTT Docomo, for example, provides life and behavior assistance to its subscribers. Qwest similarly has adopted a vision that states it goal is to bring more life into the lives of its customers. Such shifts in strategy require a better understanding of the lives of subscribers by telecom companies, beyond just knowing how subscribers use telecom products; rather, it necessitates an understanding of how subscribers live their lives – knowing their hobbies, if they like to go clubbing, if they prefer staying in, their vacation habits, etc.

The ultimate purpose of learning so much about subscribers is so that a telecom can:

  • Develop new products or services, which are most relevant to customers’ lifestyles, expanding the realm of telecommunications offerings (i.e. concierge services which are now provided by some of the leading operators)
  • Customize marketing communications for each customer, relating to their day-to-day lives in the most effective manner (i.e. positioning data bundle as a means for staying online during vacations vs. being accessible for business 24/7)
  • Partner with the most relevant brands in campaigns and loyalty propositions (i.e. with certain sports clubs in case there exists a large group of customers who regularly go to games)
  • Provide unmatchable opportunities to companies via mobile marketing, with detailed data on customers’ lives that is not available elsewhere (i.e. providing a service whereby tour operators for a fee can target subscribers who travel to specific destinations each year during the holiday season)

To be able to get at such information about its subscribers, operators need to take the next step in customer analytics. Through analyzing the “context” in which its subscribers use its products and services, an operator can make many inferences about the lifestyles of these subscribers.

Getting at context related data doesn’t require the collection of subscriber surveys or even demographic data; by conducting customer analytics in a different manner, telecoms can get at such information through the data already in their hands. For example…

  • A customer’s home-cell location can be used to predict the SES group, based on the neighborhood profile
  • The phone numbers called by the customer can indicate frequent reservations to concerts and events, based on matching the destination with phone numbers in yellow pages
  • A frequent use of the word ‘meeting’ in SMS messages can indicate the business nature of communications, as does frequent connection to a company’s e-mail servers from a Blackberry

…all of which are accessible data providing clues into the lifestyle of the customer.


Understanding the lifestyle of a customer can be achieved by analysis of the context of communications, which can be performed looking into some or all of four main dimensions of communications:

1. The Location

Using GPRS or call location data, operators can track home and work location, as well as frequently visited special cells, and derive an understanding of the profile and type of places a customer is accustomed to frequenting. From this location data, it is possible to gain an understanding of certain bits of the customers’ lives, such as:

  • Home & Work-cell Locations: Predicting the socio-economic status of the customer based on the other customers living in same area.
  • Specific Locations Visited: Understanding additional characteristics, such as whether the customer is a student (school cells), likes to attend sporting events (stadium cells), likes to travel frequently (holiday destinations),  travels abroad (roaming or airport cells), likes to go out on weekends (popular club neighborhoods), or enjoys shopping (mall cells).

All it takes to derive context intelligence from the location data is classification and interpretation of geographic cells / locations from a customer lifestyle perspective.

2. The Destination

Similar to location, destination of calls (i.e. the owners of numbers called by a customer) can tell a lot about the general interests and needs of a customer, especially when focused on the calls made to enterprise call centers, such as:

  • Regular Calls to Specific Destinations: Predicting the company a customer is working for, which brands he/she prefers, the life stage of the customer (i.e. calls to a school) or understanding regular activities such as managing money (bank call centers), travelling (airlines), or ordering food (restaurants).
  • Recent Calls to Specific Destinations: Understanding recent changes and interests of a customer, such as purchasing electronics (technology companies), having medical needs (hospitals), looking for a new house (estate agencies).

Again, similar to location, deriving context intelligence from the destination requires a mapping of numbers called by customers to company phone lists, such as yellow pages.

3. The Content

The content of a customer’s communications can be used to describe almost anything about the lifestyle of a customer, which requires looking at the content of four different communication types:

  • Data Content: Understanding the web pages visited such as news pages, online stores, sites of leading brands, and different applications downloaded and used such as instant messaging and games can give useful hints about the overall lifestyle of customers.
  • TV Content: Channels watched on mobile or regular TV can provide insights into the interests of a customer (i.e. fashion channels, sports channels, etc.), as well as whether the customer is likely to have children or not (i.e. watching kids channels).
  • SMS Content: Performing text mining on the contents of SMS messages sent and received by a customer, it is possible to identify key words such as ‘meeting’, ‘party’, ‘school’, ‘wedding’, each of which gives information in regards to the lifestyles of customers.
  • VAS Content: Similar to data and TV content, type of VAS preferred by a customer, such as games, sports, or more niche applications such as diet programs can show the special interests of a customer.

Product mapping into lifestyle segments is nowadays a common practice among leading retailers. What it translates to in telecommunications is the mapping of content into customer lifestyle segments, which should be performed for all types of content listed above.

4. The Relation

Last, but not least, many telecoms operators partner with various other retailers and service providers, for the sake of promotions, loyalty program offerings, etc., where the preferences of customers for specific partners can indicate their special interests and preferences (i.e. a customer using loyalty points mostly for movie tickets, fine-dining or sports apparel brands).

Putting these four pieces of the context puzzle together can facilitate a much deeper understanding of the profiles of a telecom operator’s customers, which, then, can be put into use for internal purposes as well as mobile marketing for enterprise customers.

What Next?

As operators move from traditional data mining into context mining, they will need to expand their thinking beyond standard telecommunications needs of customers and into how they can capitalize on newly discovered areas of opportunities. This change will also bring in new challenges, such as customer privacy management in using data.

To make this shift into context mining as smooth as possible, companies should start with pilot activities, analyzing small groups of customers focusing on specific lifestyles or dimensions first, testing actions on them and expanding into a full-fledged implementation only after the pilots have proven results without invading customer privacy.

To receive more information about our recommendations, and learn about related Forte Consultancy service offerings, please contact us at info@forteconsultancy.com.


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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