One of the NHS apps I work on at the moment supports people to cut down the amount of alcohol they consume.

The primary mechanism for achieving this encourages users to commit to at least three drink free days per week.

The data shows us that 48% of our user base find their pledged number of drink free days difficult to stick to, citing reasons such as social events, stress and not readily seeing the benefits of changing an established habit.

In a recent discovery process, we tested various ways we might alleviate some of the friction users experience when cutting back, the most promising (and Policy approved) of which was in demonstrating the health impacts of reduced alcohol consumption.

We created two prototypes to test with users which each took quite a different approach. The first enabled users to navigate the human body and interact with various aspects to review the impact alcohol had on e.g. the liver, pancreas, skin, heart, etc.

Users were interested in this and it provoked a number of thoughtful conversations, but overall we wondered how many times people would realistically return to the feature as coming down from ‘high risk’ into lower risk categories would be a sustained and gradual progress without meaningful short term change to core messaging in the meantime.

The second prototype we tested addressed the ongoing relevance concern to a far greater extent and in fact proved infinitely scalable as the more it was used, the more useful it became to the user. This feature offers a self reflection feature which aims over time to demonstrate to the user that feelings of wellness e.g. improved sleep, mood, energy, weight etc are enhanced by reduced alcohol consumption.

Analysing user experiences of the reflection prototype in turn raised a number of interesting questions including:

• How do we equate feature usage with pledge adherence gain?
• How do we powerfully reflect data back to the user to encourage behaviour change?
• How do we handle adverse correlations between drinking and mood?
• How do we balance freeform feedback with meaningful health insights?
• How do we balance a preference for weekly checkin with regular accurate results?

The next round of design and testing will seek to address these questions and identify the level of confidence we can expect in progressively rolling out the feature to a wider pool of users in the app.