The Value of A3 In Telco Marketing

This article summarizes calculations made on the use of new A³ (Analytics, AI and Automation) capabilities within telco marketing. The y-axis of the chart below is an estimate of yearly financial upside (revenue uplift + cost decrease). The x-axis is a count of the number of uses seen for more “complex” A³ (e.g. using machine learning for predictive/prescriptive tasks and AI capabilities such as text analytics).

The uses are divided into seven areas:

Customer Insight tracking and behavioral analysis, segmentation/modelling and voice of the customer/customer journey capabilities

Market Savvy  improvements to competitive planning and market share calculations

Marketing Programs churn management, social marketing, paid channels, loyalty programs, as well as search and website optimization techniques

Owned Channels  post-purchase campaigns, up-sell/X-sell, next-best action, stimulation and retention programs

Product Management product portfolio management, device portfolio tracking, pricing analysis and wholesale contract negotiation capabilities

Brand and Media  public relations and brand reputation, event management/planning, and content management strategies

Marketing Operations planning and optimization, financial planning and management as well as market research

Findings

• Most A³ used in marketing is in the form of rules-based analytics and machine learning. However, there are some uses of AI including sentiment, text and image analysis.
• Marketing Programs and campaign management in a company’s “Owned Channels” have shown the best returns from the addition of analytics and machine learning in the last few years. Although many telcos have already taken some of the easier benefits, there are still new improvements that could yield additional value.
• Marketing Operations, Brand and Media and Product Management feature moderate complexity demands with more modest returns. Most use cases for machine learning are in the collection and manipulation of data. Some AI techniques such as speech and text analytics are used within market research.
• Customer Insight offers new opportunities to add machine learning for segmentation and use of sentiment and text analytics; but has a limited financial value against it on the assumption that much of the value is taken by other downstream systems that then use the insight