The group of technologies known as the internet of things can capture behavioural data to enable the personalisation of products and services. Energy Smart Meters are an example of this application.
Smart meters: need to become even smarter in order to detect granular energy data
Personalised marketing is now becoming a reality through a broad technology known as the internet of things, social media and mobile data applications. Right now there are three kinds of IoT marketing applications:
1. Harnessing passive household data
Household energy data can be harnessed from Smart meters. More than 60% of households in Europe and north America are projected to have smart energy meters installed by 2017 and this may increase to over 70% by 2020. This provides a rich passive data source of appliance usage with a range of marketing applications: which appliance is being used, for how long, and in what mode for every day of the year without soliciting any data from the consumer. Kantar undertook a pilot study in the UK in 2014– 2015 that illustrated the following applications.
Personalised behavioural change and sustainability.
The pilot found three groups of households with similar characteristics – age of household members, type of house, socioeconomic strata – suggesting similar energy consumption. however, one of the groups had three times the energy consumption. Deeper analysis for this group revealed higher energy consumption for washing machine usage, on account of using cotton cycles 100% of the time against 40–50% of the time by the other groups. this was despite the economy cycle facility existing in all the washing machines. this highlights the role of passive energy data to drive personalised behavioural change by creating educative messages – in this case, use of economy cycles.
Utility companies can identify who to target today but not how to drive energy conservation for each of the relevant consumers. one can imagine how energy bills of tomorrow will be a new source of programmatic targeting to drive behavioural change towards energy conservation/ sustainability. this will be relevant to utility providers, corporates and government organisations. Driven by regulatory mandate, Smart meter installation is on the increase. could such applications become a reality today or are there challenges in commercialising these applications at scale?
Smart energy meters provide energy signatures every 30 minutes, which can neither detect usage of short-duration appliances (such as electric kettles or microwaves) nor track usage mode accurately for long-duration appliances. In order to detect granular energy data, one needs smarter energy meters recording energy signatures every second or every ten seconds. Such granular energy meters are available but differ from those installed by energy providers. this means additional commercial considerations are required if we are to scale such applications today, or we wait for technology to mature further to allow utility providers to render Smart energy meters even smarter.
The appliance usage pattern of ovens and microwaves uncovered a segment with a higher incidence of cooking from scratch. this segment were cooking on alternate weekdays, mostly between 3pm and 6pm. Based on this, one can target in-the-usage moment with cooking, seasoning and baking products. In the future, such targeting can be done via connected devices (e.g. connected hobs) but, today, such personalised targeting can be done via mobile devices. to make this implementable, companies will need a mobile database of customers and prospects – something that has already started. Such applications will call for cross-sector partnership between companies to harness relevant data assets – for example, between consumer goods and mobile service providers.
Better enrichment of brand strategic segments and trend detection.
Longitudinal behavioural data will help brands better understand their strategic segments. Additionally, this will help detect the time frame of consumer evolution on a living standard continuum to identify early opportunities for premiumisation. This is a key area which many consumer goods companies have failed to detect well in the recent past. There are many other such data sources that fall under this category of harnessing the existing household data – for example, set-top box data for programmatic TV.
2. Augmenting appliances with sensors
First, let's consider RFID-enabled shopping trolleys which allowed TNS to map the path to purchase for individual in-store shoppers. This mapping was based not only on the sequence of aisles visited but also on the 'time x day'/mission, speed, the aislespecific value of goods and shopper-specific distance travelled. RFID-enabled shopping trolleys identified nano-segments that can help retailers target personalised promotional offers based on the shopper's previous paths to
Sunando Das is head of marketing analytics and data science at TNS UK, bringing 18 years of experience across data science, brand strategy and digital insights. His recent work on artificial intelligence algorithms won the Research-Live 2015 best innovation of the year award.