Every day smart meters, don't actually provide insights into
exactly what is going on in the home and what can be done to reduce
energy consumption. Kantar wanted to try to make smart meters
smarter, by partnering with ONZO, a utility data science software
company, to detect the 'signature' of specific appliances.
The project involved analysing the results of smart-meters to
identify not just the different appliances being used, but also the
modes that those appliances were being used in. For example, the
team were able to detect the 'signatures' of different washing
machine cycles. Looking at three households with the same profile,
the team found that one family was consuming three times more
energy than the other two. The culprit for high energy consumption
was the washing machine set permanently to the 'cotton' setting,
and never being run on the 'eco' setting. Such findings help in
targeting relevant households with personalised messaging to drive
behavioural change and identifying ways to drive sustainability,
high on agenda of Government bodies and corporates.
The team found they could infer consumer lifestyles based on
appliance usage patterns. They could also get a clear understanding
of particular segments of society. For example, people who cook
more from scratch on weekdays tend to cook between specific
time-slots, allowing for better 'in the moment' targeting.
Unsurprisingly, detection using the smart meters was more
accurate that the results of the diaries or surveys that the
households were asked to complete, but it's when the energy meter
data is integrated with other data sources, such as search, social
or purchase data, that there is the potential to transform consumer
There is lot of speculation around how Internet of Things (IoT)
will transform marketing of the future. Energy data is one form of
IoT enabled passive data with potential to transform consumer
insights and is available today for us to harness.
Source : Kantar