Can energy-use profiles help homeowners make decisions to decarbonise their homes? Recent work from Nesta has focused on developing energy-use profiles, data-based representations of how households in Great Britain use energy. We wanted to understand how these profiles could be used to change homeowner behaviour, such as by encouraging them to shift energy usage or switch electricity tariff. To do so, we adopted a test-and-learn approach, prototyping a recommendation tool for households based on their energy use. 

We found that telling people they are receiving information based on their energy-use profile is less effective than saying, "Here’s your personalised advice based on your input."However, energy-use profiles could play a valuable behind-the-scenes role in building products. Future work will focus on these behind-the-scenes technical applications of energy-use profiles.

Smart meters provide real-time data that can help consumers understand their energy use from moment to moment. But they’re also capturing increasing amounts of potentially useful data with applications for consumers, energy companies, policymakers and regulators.

Now that we have developed our own energy-use profiles from smart meter data, we want to understand how they could be used to support households to decarbonise their homes.

What if understanding how they use energy over longer periods could inform how consumers make bigger decisions? For instance, which tariff should they be on? Or could getting a heat pump lower their energy bills? This project is exploring how consumers engage with the concept of an ‘energy-use profile’. Does it make sense? Is it exciting or boring? Could it increase their confidence to make big decisions?

Unlike most design-led projects, this one originated from a technical feature – energy-use profiles – rather than a clear user need, such as knowing the cost of a heat pump before purchasing one. As a result, our goal was not to validate a product but to test ideas, or "provocations," such as: Does seeing how similar households use energy assist in decision-making? Is it helpful to compare actions taken by similar households? And would an interactive tool facilitate this process effectively?

Given the exploratory nature of the project, we developed prototypes that were not intended to serve as final products. Instead, we introduced a concept – a tool offering personalised advice based on energy consumption – as a "sacrificial idea". This was designed to help us gather insights and build foundational knowledge within a specific context. These prototypes functioned as interactive interview guides, created to explore specific research questions (eg, Are participants willing to share their smart meter data?) rather than to deliver a polished user experience.

We began the project with desk research and behavioural mapping, followed by rapid prototyping to test our assumptions. This process resulted in a tool offering tailored recommendations based on user-provided information. These included switching to a time-of-use tariff, installing a heat pump, or combination of both.

Over the course of three months, we conducted two rounds of user testing–first with 16 participants in February and then with 11 in March. The participants included homeowners with and without heat pumps, with varying levels of environmental motivation. This diversity allowed us to compare responses across different user groups.

In recent months, we structured our research around three core hypotheses. Key questions we explored included: How personalised does a recommendation tool need to be to effectively persuade users? Or can comparing energy use with other households encourage behaviour change? It is important to emphasise that our findings represent “signals” rather than concrete evidence; they should be interpreted within their specific context and not considered definitive truths.

Hypothesis 1: Households are more willing to switch to low-carbon heating when provided with recommendations tailored to their current energy consumption and situation. 

Answer: Yes, but not necessarily through the vehicle of energy profiles.

Personalisation and feeling like recommendations are relevant matters to people. It increases their confidence and trust. This feeling can be created with low-fidelity qualitative playback (see prototype below) but could potentially be strengthened through the use of energy data.

Hypothesis 2: Households find the concept of an energy profile – a group of people who use energy in a particular way – a useful framing to think about their own energy use or how it could change. 

Answer: No strong evidence to support this hypothesis.

The more personal, the better! We observed that an energy profile introduces an additional layer of abstraction, which hinders personalisation and creates confusion among participants.

Hypothesis 3: Households are willing to share their smart meter data in order to get personalised recommendations or matching with an energy profile. 

Answer: Yes, in theory.

In theory, households are likely to be willing to share smart meter data. Whether this is true in practice was beyond the scope of testing, as they can face physical barriers that would discourage them (eg, look for a long code on your device).

Through conversations with homeowners, we discovered that saving money remains the primary motivation for their actions, even when they have strong environmental concerns. Most participants expressed a preference for "energy-saving tips"– simple, practical steps that deliver immediate, albeit modest, savings. This finding reinforced our initial assumption that tools based on energy profiles are not aligned with the type of recommendations households seek. For example, homeowners don’t require a complex tool or smart meter data input to understand that doing laundry outside peak hours can lower costs. 

Additionally, many participants revealed that they had already taken measures to reduce their energy consumption or conducted their own research on the topic. Although our sample size was small, the results suggest that the energy crisis has likely increased public awareness of energy use.

This project provided valuable insights into how households interact with energy-use profiles. While these profiles did not significantly influence participants, social comparison – illustrating how others approach similar decisions – proved effective in helping people position themselves and identify changes to optimise their heating systems and save money. Personalisation emerged as a critical factor, making recommendations more relevant and establishing clearer expectations.

Although energy-use profiles were less impactful as a framing tool, they hold potential as a behind-the-scenes feature in product development. They are better suited to supporting technical functions rather than serving as a direct information delivery mechanism for users. Moreover, the type of information households seek does not necessarily justify the resource-intensive creation of personalised tools. Consequently, our future work will focus on leveraging energy-use profiles for technical applications targeted at non-household users.