Factors shaping overall satisfaction and customer service satisfaction with energy suppliers: 2025 to 2026

Main points

• Customer service measures explain 38% of variation in overall satisfaction, with satisfaction with billing and ease of contact emerging as the strongest predictors in this model, followed by satisfaction with smart meters.
• A separate demographic and energy characteristics model explains 14% of variation in overall satisfaction, with financial circumstances and supplier emerging as the strongest predictors.
• Demographics, energy characteristics, contact method and reason for contact explains 25% of variation in customer service satisfaction amongst those who have had contact with or tried to contact their supplier.
• In this model, financial vulnerability was the strongest predictor, with age and supplier as joint second strongest predictors.
• A separate model focused on contact perceptions explains 54% of variation in customer service satisfaction among those who had contact or tried to contact their supplier.
• In this model, several contact perception factors are of comparable importance, with empathy and overall handling of the interaction emerging as the strongest predictors, closely followed by issue resolution.

Methodology

BMG Research have been running the Energy Consumer Satisfaction Survey since 2023 on behalf of Ofgem and Citizens Advice. 

Following record-high satisfaction in January 2026 (81%), we carried out further analysis to understand which experiences and consumer characteristics were most strongly associated with satisfaction. We also conducted new analysis to understand what impacts customer service satisfaction amongst those who have had or attempted contact with their supplier.

The analysis of overall satisfaction combines data from two survey waves (July-August 2025 and January 2026), representing 7,701 respondents. The dependent variable is the ‘overall satisfaction’ measure from the Energy Consumer Satisfaction Survey.

The analysis of customer service satisfaction amongst those who have had or attempted contact with their supplier combines data from three survey waves (January 2025 to January 2026), representing 4,262 respondents. The dependent variable is the ‘satisfaction with customer service’.

The models use logistic regression – a statistical method used to predict the likelihood of a binary outcome (in this case, either being satisfied with the energy supplier, or not being satisfied or feeling neutral). Regression modelling allows us to hold other factors so we can pinpoint which ones matter for satisfaction.

For more information about sampling methods please refer to the Energy Consumer Satisfaction Survey technical report published alongside the main satisfaction report.