Impact Study

Leading European Telco

Our client is the leading supplier of fixed and mobile networks for telephony, data and television in the Netherlands, with over 10,000 staff and revenues over €5bn.


In 2019, Our client consumed about 700 GWh of electricity – the equivalent of 250,000 households – making it one of the largest electricity consumers in The Netherlands. The company is committed to reducing its carbon footprint through more efficient energy use. They therefore needed to understand its energy consumption, and the drivers underlying it, in granular detail in order to develop a more sustainable approach. Consumption analysis and forecasting were hampered by a shortage of high-quality data, as not all network hardware was equipped with meters. And what data was captured was often inconsistent or incomplete.


Three separate but integrated algorithmic systems were developed to tackle this challenge.  The first used machine learning to model out energy consumption across the entire network. This predicted where data black spots existed, and accurately estimated usage and energy consumption to a granular level. The second is an AI-driven forecasting system that predicts energy consumption across the entire network, based on predicted changes to network traffic and technical capabilities. Combining these two systems allowed the development of the third. This identifies redundant hardware and provides an energy efficiency and sustainability lens through which to interrogate plans for rationalization and expansion of the network.


Gaining detailed and accurate insight into current energy usage provided the solid foundation necessary to improve forecasting accuracy and plan efficiency measures. Armed with this new visibility and insight, KPN has embarked on an energy efficiency program across its business that’s expected to deliver a 44% reduction in energy consumption by 2030 compared to 2010 figures.

Impact stats
Integrated systems
Developed to tackle the challenges of data visibility, usage forecasting and energy reduction
AI algorithms
Developed to predict energy consumption to a high degree of granularity
Lower energy consumption
Estimated company-wide reduction by 2030, compared to 2010.
Enhanced analytics of existing data enabled greater visibility into current energy usage, leading to more accurate forecasting and improved efficiency. While this delivered a huge cost saving to the business, more important was the significant reduction in carbon footprint within a largely unchanged operational framework.
This site is registered on as a development site. Switch to a production site key to remove this banner.