Telco AI investment is expected to ramp up in 2021, but there is not as much clarity as there should be around best practice and business outcomes.

Straight Talk Service Provider

Omdia view

Telco AI investment intentions are finally catching up with the hype, as service providers realize they need to invest in AI-enabled tools in order to automate network and service management and improve customer engagement.

According to Omdia’s ICT-Enterprise Insights survey, nearly 80% of CSPs see the use of AI/analytics to automate network activities as an “important” or “very important” IT project for 2021, with nearly 60% of CSPs planning to increase investment in AI tools. Top AI use cases are expected to include network fault prediction and prevention, automation of end-to-end life-cycle management, and the management of network slicing. AI will also support a variety of non-network use cases, including using AI to support new business models such as contextual offer management as well as automating and personalizing customer engagement and delivering customer insights.

In truth this growing service provider interest in AI-driven capabilities is only playing catch-up with the transformation that has long been happening to vendors’ software and services portfolios.

But what will this AI-focused investment support, and how will service providers recognize a positive outcome? If we take automation as an example, it is often said that it is crucial to consider the potential ROI before initiating an automation project. ROI is certainly a good starting point for sorting out “must have” from the “nice to have” automation project, whether looking at it from the perspective of five-year cost savings, annual operating cost, time to value, or some other indicator.

However, measuring automation outcomes is more complex than it may at first seem. It is of course useful to directly compare operations costs before and after adoption of an AI-based solution, but it is not the full story. It is also important to consider the business outcomes that require prioritization. These can include improving the accuracy of a process, increasing consistency and predictability, including ensuring compliance with specific SLAs, delivering greater reliability, boosting productivity, or reducing turnaround times. The list is extensive, but to make a success of an automation project it is important not to lose sight of the end goal, and to identify those KPIs which specifically support the business outcomes you are seeking to achieve.

The successful implementation of an automation project is more than the sum of the parts. It’s all very well targeting a specific KPI—such as a 30% improvement in diagnostic capabilities—but to implement a successful automation it is necessary to reduce process time from the point of view of the complete process cycle and avoid potentially creating more work elsewhere. Nor does it end there, an AI-driven automation also needs to be sustainable. It’s not just about having the capabilities in place to address incidents as they occur; a process automation also needs to continue to be relevant even when a network/IT element is upgraded, or a vendor swapped out. 

Last but not least, visibility is essential, because to improve anything you need to be able to measure it. But how does a service provider know if they are automating more successfully than their peers? There are plenty of sources of AI-linked training and support, as well as best practice guidance and models provided by industry bodies like the TM Forum. But there is not as yet a commonly agreed methodology to assess automation in the telco space. Some vendors have internal measures, such as internal process automation indexes, but this is not the same as having an industry-wide measure.

As things stand, any assessment of a vendor’s or service provider’s automation success has to be tied to achieving specific business goals and the KPIs that directly support those goals. This will suffice in the short term. But as the market becomes more mature and AI capabilities more pervasive, there will be an increasing need to have the means to measure the effectiveness of automation initiatives.

Straight Talk is a weekly briefing from the desk of the Chief Research Officer. To receive this newsletter by email, please contact us.