When adoption stalls, the first instinct is to add a feature. In my experience the features are usually there. Customers have not yet changed a habit.
At a global bank's consumer business I owned digital and mobile adoption. Annual digital adoption reached the 85%+ range across 2022 and 2023. Very little of that came from new features. It came from changing behaviour, one habit at a time.
What moved it
- Teach the app in small pieces. We started a "Did You Know" education series to show customers features they already had. Short, repeated and specific beat one big launch campaign.
- Move one transaction at a time. Mobile education campaigns shifted telegraphic transfers from the branch to mobile. That grew mobile transaction share and cut cost-to-serve.
- Retire the paper default. We migrated customers to eStatements, moved alerts from SMS to push, and automated notifications. Each change removed a reason to stay offline.
- Reward the moment, in the moment. Auto-triggered FX nudges and real-time voucher fulfilment for wallet spend reached customers while the action was still fresh.
Why this matters for AI
Every item above is a behaviour, matched to a trigger, measured by whether the customer did it again. That is also the job of an AI engagement engine. A bank that has learned to change one habit with a well-timed nudge has already built the muscle AI needs.
So before asking for a new feature, ask which habit you want to change, and what would prompt a customer to change it today.