Unleashing the Potential of Mobile Analytics

The ever-evolving technology has transformed the phone from a mere calling device to a smart device that can perform multiple tasks in a jiffy. In the process, volumes of data are generated. Effective use of Mobile Analytics has the potential to open up a unique set of business opportunities for communication service providers in the Indian Telecom Industry

– Compiled by Preeti Gaur

There has been a phenomenal change in the way we use technology. The phone, for instance has transformed form a mere calling device to a smart device to take pictures, browse the internet, listen to songs, access email, consume media like photos and videos, and much more. Such transactions have led to the generation of volumes of data banks. Brands and communications service providers are increasingly keen to understand what their subscribers are doing, when and where and how they can provide offerings that can meet their individual needs.

Effective use of Mobile Analytics has the potential to open up a unique set of business opportunities in the Indian Telecom Industry. Knowing how customers interact with your mobile channel is vital to the success of a mobile strategy. Mobile Analytics provides deep insights into customer engagement, behavior and loyalty, revealing the content and media they find most compelling. From measuring key customer engagement indicators to tracking specific conversion activity, the payback of Mobile Analytics is manifold.

Mobile Analytics enables monitoring and monetizing subscriber engagement through sophisticated Data Analytical techniques. It allows for ongoing monitoring of mobile subscriber data across all channels and touch-points and monetizing them effectively into usable, meaningful and actionable intelligence through two sided business models.

Elucidating Mobile Analytics

What is Mobile Analytics?

Mobile Analytics refers to a set of processes, methodologies and tools that:

•    Provide a 360° profile for every subscriber across all access points, channels, devices and services

•    Enable segmentation and micro-segmentation based on profiling attributes

•    Allow automated and human assisted actions to respond to behavior exhibited by micro-segments (or) individuals.

Who are the user groups?

The target user groups for Mobile Analytics processes and tools include:

•    Marketing groups mandated with retention, up-sell and cross-sell

•    Quality of Service and Customer Insight Groups

•    Front-Office Personnel and Customer Service Agents

•    3rd Party Eco-System that wants to leverage and target micro-segments.

Key players in this category would include mobile advertisers, content publishers, content aggregators, application developers and so on.

What it can accomplish?

Following is what Mobile Analytics can accomplish for Service Providers

•    Personalized products, bundles and applications to meet subscriber preferences

•    Generate revenue stream from third parties who access mobile subscriber data

•    Better customer service through 360 view

•    Understand outliers in terms of negative customer experience

•    Target retention offers for subscribers on the cusp of churning

•    Target up-sell and cross-sell offers to increase top-up and deepen service adoption

Following is what Mobile Analytics can accomplish for the Partner Ecosystem:

•    Understand Brand Engagement on mobile

•    Understand market trend, competitors and its penetration in digital world

•    Advertise Uptake Analytics (Impressions, Clicks, Engagement)

•    Optimize portals for Engagement and Conversion

•    Include Mobile Marketing as a part of the online marketing budget

•    Understand Application Download and Consumption Patterns

How it will be effective?

The following diagram outlines the key information sources and the building blocks to an effective Mobile Analytics Strategy:

Mobile Analytics Process Flow

Step 1: Get the building blocks right

Mobile Analytics Schema – The Information Model for Mobile Analytics, viz., the Customer, Products, Offers and Event schema at the summary and detail levels.

The first step is to design the right data structure for successful Mobile Data Analytics.

It would capture the customer demographics, network events, transaction information, product, services and offers, and cross channel customer interactions, in a way that fits the needs of the service provider’s business. It might prove to be a good idea to partner with specialist consultants and solution providers for this critical design phase since this will essentially determine the boundary conditions of the capability of the Mobile Analytics initiative.

Customer Profile: The most important aspect of the information model is the customer profile. A wide array of attributes and counters have to be defined to capture the customer’s interaction across various touch-points, his consumption behavior and so on. It is not uncommon for the list of unique attributes to record a truly 360° view.

While some of these attributes belong completely within the Service Provider’s domain, others will be ‘virtual attributes’ which deal with interactions with 3rd Party sources and services. Accurate definition of the binding (mapping to the right ‘raw’ data source) and the calculation logic of each attribute are equally critical.

Step 2: Ensure Quality of Data

Once the data structure of the information to be captured is defined, the next step would be to actually collect and manage the data. Not surprisingly, this is the most challenging aspect. Even with an overwhelming number of IT (and non-IT) systems and solutions collecting, transforming and analyzing every Telco’s data, the concept of the ‘golden data store’ still remains largely elusive. While complete, timely and accurate data is a pre-requisite to get the ‘right’ customer insight, more often than not one will have to make do with acceptable margins of error.

The key data acquisition steps are:

•    Collection of events across all customer touch points

•    Accumulation of customer event data at every step along the customer life cycle on an ongoing basis (from activation through churn)

Data Collection may have different frequencies and embedded business logic depending on the type of data and based on specific conditions (for example, high value customers vs. low ARPU customers, new customers vs. old timers, and so on).

Moreover, as we are dealing with millions of transactions and event records on a daily basis, effective high volume data management is critical. It might make sense to leverage existing data marts within the Telco’s business, coupled with customized but scalable ETL solutions to collect new information.

Step 3: Transform and Enrich the Data

The raw data collected from the OSS / BSS by itself will not be suitable for gaining the kind of customer insight that service providers are looking for. The collected data needs to be evaluated, aggregated, correlated and filtered carefully.

Available ETL tools and Data Analytics Solutions may be used to do this efficiently.

The objective is to build a customer profile that can be leveraged to make intelligent decisions on retention and sales, hence it may not be necessary to store all detailed event data. A key aspect of the design phase would be to define the most appropriate enrichment and normalization rules to ensure that the transformed data not only contains all the relevant insights, but is also easy to analyze.

Step 4: Generate Customer Profile

Once the customer profile attributes are populated for the first time, a view of the customer can be obtained.

Subsequently, it needs to be updated on an ongoing basis based on input events in near real time. The update can either happen periodically or through some external trigger.

Not only the usage (consumption) data needs to be updated periodically, non-usage information e.g. Service Requests, Payments, Preferences, etc. needs to be refreshed from time to time to ensure that we always have a relevant and real-time view of the customer.

Moreover, if the ‘customer’ is defined at the right level, the profile would be able to integrate multi-device and multi-SIM views of the same customer as well.

Step 5: Segment the Customers

Once the customer attributes are available, dynamic segments can be defined based on specific value ranges of the attributes (or metrics derived from them) to identify customer clusters, i.e. groups of subscribers with specific behavioral patterns and similar preferences.

Service Providers today have a choice of wide array of sophisticated tools for advanced data mining and dynamic segmentation at macro and micro levels. The customer attributes are further used to derive specific metrics with a particular focus area in mind.

Examples of such pre-packaged customer analytics

would be:

•    Customer Experience Metrics

•    Customer Journey Analytics (Cross Channel Behavior)

•    Portal usage

•    Business Monitors by Service (Music, MMS etc.)

•    Off Portal Usage

•    Customer Care Metrics

•    Quality of Service

Step 6: Identify the Treatment Strategy

Once the segments are defined, each of them needs to be analyzed to identify the appropriate treatment strategy. Simulation Tools can help the Service Provider to perform what-if analysis before an action is taken. The typical boundary conditions that need to be carefully considered prior to devising an action strategy for each segment are as follows:

•    Control Group Attributes

•    Eligibility Criteria for Action Execution

•    Adoption and Uptake Analysis

•    Approval Workflows

Step 7: Action execution and Closed Loop Analytics

The key variables that must be considered while defining actions for a particular segment are:

•    What (Offer) – Pluggable Text, Email, Image

•    Who - Initial list of MSISDNs / Subscribers

•    When - Date / Time; Frequency Capping / Scope

•    Channel - SMS / Email / Tele-sales

•    Condition – Eligibility Criteria

Action Execution will take place based on an internal or external trigger or a pre-defined schedule for all subscribers who meet the eligibility criteria.

Closed loop Analytics – A well thought through Mobile Analytics Strategy does not stop at defining and executing the actions. It is equally important to study the response to executed actions and perform uptake analytics in order to be able to dynamically change action parameters subsequently to further optimize the response.

Source: Mobile Analytics – CII, Protiviti