Estimating customer defection in personal retail banking

Using Juster’s probability scale, this study examines a sample of 881 customers’ self‐reported estimates of probable defection from their main bank. One year later these customers’ statuses with their main bank were reviewed, allowing comparison of intended with actual defection. This methodology allows not only investigation of the Juster scale’s performance in a “subscription” type market but also identification of those customers with a predisposition to switch banks. Bank management can then decide if and how they might address their switch‐prone customers. Respondents were asked to allocate their probability of closing all accounts at their main bank in the next 12 months. The sample estimate was 10 percent while the actual defection result was 5 percent. While this may seem somewhat inaccurate, similar studies in the literature report the tendency of the Juster scale to over‐estimate switching behaviour, but its estimates have consistently proved superior previously to those derived from attitudinally based intention‐to‐buy scales and at least the equal of the Conversion Model.