Data-based customer service strategy
Using mathematical techniques to identify key problem areas for a network provider's clients
Challenge
Energy measuring company as part of a large network provider (Stedin). Their basic question was how can we explore our data and instead of re-actively measuring energy use by our clients, start pro-actively advising our clients e.g. in terms of energy saving measures.
Approach
We collected energy consumption information (gas & electricity) (per every 5 minutes past 20 years) for some 100 (wholesale) customers. The data was reduced using Fourier transformation keeping its properties. We than analysed the energy consumption profile per type of object by defining smart metrics.
Result
A dynamic dashboard showed the metrics and energy profile for individual objects such as a particular swimming pool or office building (plotted on a map). This allowed our client to identify for example the worst isolated objects and engage into a discussion with the respective client. It also led to the identification of unusual profiles (e.g. school buildings with continued energy consumption during the weekend).
