The business challenge
What needed to change
Our client, a leading European bank, wanted to move personalisation beyond demographics and event triggers to build a genuinely individual experience across its digital banking platform.
Three connected problems were holding this back:
Personalisation relied solely on event triggers, with no deeper behavioural or attitudinal understanding of customers
Existing models were siloed, non-standardised across channels, and lacked the explainability needed for activation at scale
Demographic segmentation had limited reach, it described who a customer was, but not why they acted
This was compounded at product level, sales of insurance products sitting outside the bank's core offering were stagnating, particularly on its digital platforms.
The human problem
The behaviour behind it
Our behavioural diagnosis found that psychographic factors, not demographics, are the strongest predictors of online banking behaviour, because they are the only signal that directly informs attitudinal preferences, motivations and capability. For insurance specifically, three barriers stood out:
Delayed reward: insurance offers a reward customers can't appreciate in the present, making them reluctant to pay for it
Complexity of perceived value: insurers' offers emphasised functional and emotional value, but ignored the individual and social value that also motivates or discourages purchase
Lack of psychographic personalisation: attitudes to insurance are driven primarily by risk tolerance, so customers at different points on that scale need different products and different messaging
Our breakthrough was mapping psychographic profiles directly from the client's own banking transaction data, no surveys, no self-reported inputs. How people spend, save and engage with their bank reveals who they are. Psychographic dimensions were derived from transactional proxies, for example:
Goal orientation, inferred from automated savings, deposit patterns, credit card behaviour
For each segment, we built full activation profiles, including tone-of-voice guidance, product recommendations and communication-framing guidelines, so marketing and design teams could act on the insight immediately.
Our impact
The difference we made
We personalised promotional messaging within in-app communications based on the psychographic segments, then validated the model in live campaigns:
+125% increase in click-through rate on a personalised insurance promotion
+50% increase in online sales of the same product
6x higher home insurance sales from personalised versus control messaging in the affluent segment, generating 927 applications over the campaign period
