In the UK, the domestic sector, excluding transportation, accounts for 11% of the national energy usage (CarbonBrief, 2015) and it saw its share increase by 3.6% from 2015 (BEIS, 2016). On average a 2-person British household pays around £790 in bills for electricity and gas (UKPower, 2017), with an additional £380 for water (TheTelegraph, 2017) per year. While prices of dual fuel deals vary across suppliers, it remains that the average person in the UK consumes both electricity and water excessively. Reducing our home energy and water usage would have a direct impact on national figures, thus participating in the national efforts towards achieving carbon mitigation targets. Among various carbon reduction strategies, the smart meter rollout aims to equip each of the 26 million British households with smart meters by 2020 (DECC, 2013). This will enable government and companies to implement large scale Energy Demand Side Management (EDSM) measures to help consumers reduce their energy and water usage using more appropriate feedback and billing systems than existing ones. This PhD project revolves around a very simple question: how are consumers in the residential sector incentivized to reduce their energy consumption? Existing research in the literature, that will be presented in more detail in the next section, has considered different levels of interventions aiming at triggering behavioral change, still findings are far from being homogeneous across studies since both contexts, methods and achievements vary significantly between experimental setups.

Technological progress now allows personalised intervention: smart meters collect data in real time, smart application dispatch user friendly feedback and machine learning enables precise profiling. Research is needed to understand underlying decision mechanisms which lead consumers to adapt their behaviours in response to feedback mechanisms. Comparing the performance of each behavioural intervention is necessary to identify the best mechanism in the relevant context, alongside with psychological research on discrepancies across individuals towards designing personalised interventions which suit each consumer.

Still if these reductions cannot be stretched any further, what solutions are left to control domestic appliances and better manage consumers' assets? While overloading consumers with information exhausts their interest in campaigns (Pugh, 2016), technologies which avoid consumers to even think of their consumption by managing it on their behalf are expected to hatch and quickly develop in British homes (SmartEnergy GB, 2016).

This work could help tailor feedback messages to individuals according to predictive algorithms based on near-real time data collection. To do so, the intervention would identify and exploit the mechanism which is the most likely to succeed in sustaining reductions in energy and water consumption. This should provide evidence that feedback intervention best shines when coupled with systems which support consumers in managing their appliances. From a research perspective, given a certain home configuration this work should provide a deeper understanding of the psychology of energy consumption and help towards identifying which feedback mechanism best fits the targeted consumer. The results of this research could be of relevance, and therefore impact both policy makers, because it will give the understanding of how to achieve efficient reductions needed for society, and utility companies, because although they do not particularly want reductions in consumption, a better understanding and service to their customers could lead to customer retention (important when competition) and more efficient utilisation and control of their system and resources.