The Increasing Importance of Analytics and Machine Learning For Telco Marketing
The sea of data within telcos, if correctly interpreted and utilized, can lead to greater monetization opportunities and understanding of customers.
Telco data hubs bring new insight to a wide range of functions across the telco and beyond to their enterprise customers through creation of data monetization products. The diagram below illustrates future monetary value from the addition of analytics and machine learning to data stored in a telco’s data hub and then used in various internal and external use cases.
This data has a wide range of applications, typically adding customer insight or customer usage patterns into a process. But roughly a third of the total calculated value sits within two of the boxes on the left – Marketing Campaigns and Marketing Programs.
The Untapped Data Potential in Marketing Campaigns
Although marketing teams were one of the earliest users of data and analytics, there remains additional value that they will be able to make use of over the next few years. This is primarily related to the addition of machine learning for personalization and event-driven targeting, but also includes more intelligence and automation in the underlying data preparation and model building processes. This underpinning of campaigns and other marketing programs with machine learning automates and improves the quality of various processes including data discovery, data integration, improvements to data quality and dynamic data engineering.
Both Marketing Campaigns and Marketing Programs are positively affected by the addition of more automated insight. Indeed, the diagram above assume that the majority of the additional financial value for telcos will not come from new types of programs or campaigns but from the discovery of previously unknown insight and patterns around customer behavior that improve understanding for the marketer and hone/speed up campaign. This improved targeting and more nuanced and accurate prediction of future behavior results in better recommendations and optimization of the time and place of offer delivery.
These capabilities have been appearing in the market over the last couple of years, and the most noticeable feature is their shift towards delivering a “data story”. The underlying automation reruns analysis on the customer both at regular intervals (to capture subtle changes in profile/behavior) and when required by the marketer – it then delivers insight in contextual forms that highlight the most important insights. For example, those who stream movies have the highest propensity to churn but are most likely to respond to campaigns for additional data. These capabilities allow citizen data scientists – who typically have strong intuition about the customer but lack the skill set of a data scientist – to improve their understanding of customers and run marketing programs and campaigns more effectively.
Barriers to implementing these capabilities include organizational immaturity, low levels of data literacy and lack of trust in the perceived” black box” approaches to running marketing programs and campaigns. Telcos and vendors will need to work on creating solutions that offer good explanations of the insight they provide and also on comprehensive training programs so that users feel confident in using the capabilities and also in understanding the underlying data.