Behaviourlab Comms Testing – Telecoms Case Study

Synthesis

Multivariate statistical models were applied to each outcome metric – comprehension, annoyance, perceived fairness, likelihood of complaint – to quantify the degree to which variations in outcome were attributable to the email design changes. We included telecoms profiling and demographics in all models to control for individual differences and identify the customers at most risk of a negative outcome.

This analysis identified email variants with consistently better performance across outcome metrics – fewer negative emotions and a reduction in undesirable behaviours such as complaints. It also identified high risk groups, enabling us to devise targeted mitigation strategies in advance of distribution.

Predictive eye-tracking, which utilises a deep learning algorithm trained with data from more than 70,000 real eye-tracking studies on consumer comms, was run on the email design, to provide further insights into its attention economy and to highlight important information that might otherwise have been missed.