Telco vendor case studies/press releases published during 2018/19 were tagged with the main type of A3 (Advanced analytics, AI and Automation) used. The diagram sets out the distribution found, with the x-axis showing total number of case studies.
Main Findings
- There is movement from more simple types of descriptive/diagnostic ML using pattern and anomaly detection towards the next stage of predictive/prescriptive solutions and automations.
- “Pattern and anomaly detection” describes the use of ML to understand data and spot useful patterns. It is implemented as one of the first steps towards more complex predictive or prescriptive capabilities.
- The “Automation – Network” category describes use cases which offer some type of simple prescriptive action in the network, typically the first steps towards self-healing or self-optimization.
- A popular use for ML outside of the network is to understand a customer, segment them and then provide prescriptive actions that offer some type of “personalization”. This is used both in marketing and within channels.
- “Intelligent automation” is a catch-all term used to describe all types of automation which include some type of analytics. It is mostly seen in the more complex telco processes such as trouble ticketing and alarm management.
Methodology
In order to understand the types of advanced analytics, AI and automation (A3) being implemented by telcos, we visited 250 websites of vendors known to sell to telcos – and analysed 130 suitable press releases and case studies, giving a picture of the depth and breadth of actual live implementations.
For more detail on this topic and a view of other upcoming research, please do contact me – charlotte@charlottepatrick.uk
