If you work in health behavior change, you have probably seen the pattern: an app teaches people what they “should” do, and then… nothing changes. Users drop off. Habits stay the same. Outcomes barely move.
In this conference talk, Sam Liberty (a designer and researcher known for rigorous, evidence-driven work with organizations like UNICEF and the Red Cross) makes a compelling case for a simple truth that too many health products still ignore:
Knowledge does not equal behavior.
And if you design as though it does, your intervention might look impressive on paper but fail in the real world, sometimes with serious consequences. For health and gamification professionals who want to design experiences that actually improve lives (not just engagement metrics), this talk is a must-watch.
Why you should be excited to watch this talk
Sam’s talk is practical, opinionated, and grounded in real measurement. He is not interested in “gamification theater” or feel-good design stories that stop at anecdotes. He is focused on efficacy, especially when the stakes are high.
Early on, he shares a vivid example most of us recognize: at a holiday party, everyone knows the vegetable plate is healthier than brownies, but the brownies disappear first. Not because people are uninformed, but because behavior is driven by far more than information.
That framing sets the tone for the rest of the talk: this is about designing for what people actually do, not what we wish they would do.
A powerful case study: Hand Washing with Ananse (and why it worked)
The heart of the talk is a case study Sam helped lead: Hand Washing with Ananse, a behavior-change intervention in Ghana developed with partners including UNICEF, the Red Cross, Right to Play Ghana, and the Engagement Lab.
The challenge was not ignorance. People already knew hand washing mattered. The real work was understanding why the behavior wasn’t happening anyway.
Sam describes how the team started by bringing together diverse stakeholders (parents, teachers, water and sanitation experts, doctors, healthcare workers). Predictably, each group had a different “true” answer:
- Teachers pointed to curriculum gaps
- Water and sanitation experts pointed to infrastructure and placement of stations
- Parents pointed to affordability of soap
One of the most valuable lessons here is methodological: behavior problems are rarely one-dimensional. They are often a mix of logistics, culture, habit, and misalignment between systems.
Then the team asked a deceptively important question: “How do you play?”
That question revealed something many designers still get wrong: play is not universal. What feels like a “game” in one culture may not land at all in another. In Ghana, the team found that play often showed up through singing, dancing, acting, and folk storytelling, forms that traditional “game design frameworks” do not always treat as “games,” but which were essential to creating a meaningful intervention.
The result was a co-designed experience rooted in local stories and familiar forms of play, including the folkloric trickster spider Ananse.
And importantly, the outcome was not just “more awareness”:
- Yes, knowledge increased
- Yes, intention improved
- But most notably, they measured an 8.25% decrease in stomach illness compared to control schools, with continued year-over-year improvement after the intervention
For anyone serious about health gamification, this is the kind of result that makes you lean in.
Digital therapeutics: why most retention collapses
Sam then turns the lens toward digital therapeutics and health apps. The promise is huge: interventions in your pocket, constant access, reminders at the right moment, rapid A/B testing, scalable iteration.
So why do so many of them fail to keep people engaged?
He points to an uncomfortable statistic: average retention for health-related apps can be around 10%.
His diagnosis is blunt and useful:
- The content often “sucks”
- Many apps use a scolding, finger-wagging tone that makes users feel judged. Even when the information is correct, the experience is emotionally misaligned with how change actually happens.
- “Designed by doctors” is not the same as designed for engagement
- Sam is careful not to dismiss medical expertise, but he challenges the default hierarchy. Clinical experts should inform the intervention; designers should lead the user experience. If the product feels like a lecture, it should not be surprising when people quit.
- Teams don’t really talk to users (especially churned users)
- Many products claim to be user-centered, but skip the hardest part: reaching people who left and hearing the brutal reasons why. Data can show where drop-off happens; qualitative work explains why.
If you build behavior change products, these points land as both critique and checklist.
A memorable story (and a key design lesson)
One of the most engaging moments is the Ananse story Sam shares, set during the Yam Festival. Ananse wants more than everyone else, so he contaminates others by shaking hands after not washing his hands, hoping they will be too sick to eat, leaving more food for him. The twist is perfect: even the cooks get sick, and Ananse ends up hungry and alone.
It is gross, funny, and unforgettable. It also demonstrates something important: sometimes a story can carry the behavior-change lesson more effectively than a complicated game mechanic.
Sam reinforces this when he explains a prototype “infection handshake” game that seems clever and persuasive, but ultimately failed:
- It modeled the problem (disease spread) more than the solution (hand washing behavior)
- Kids in Ghana did not find it fun or intuitive (hidden-role play and “double squeeze” mechanics were unfamiliar)
- The team cut it, despite ego, and replaced it with something the kids already loved and understood (a familiar playground tag-style game)
For professionals, this is a great reminder: co-design is not just about inclusion. It helps you avoid shipping “smart” mechanics that miss in real contexts.
The most actionable takeaway: people change when they feel good
Toward the end, Sam shares what may be the talk’s most practical “secret,” drawing on BJ Fogg’s behavior model and the concept of celebration.
The key idea:
- It is often easier to increase ability (reduce obstacles, make the action easier) than to increase motivation
- After someone performs the desired behavior, celebration matters, because “we change when we feel good, not when we feel bad”
Games are naturally good at this because they provide feedback, reward, and that burst of emotional success (Sam references the powerful feeling of “fiero,” the joy of overcoming a challenge).
In the Ghana intervention, kids literally experienced joy while doing the correct hand washing steps in a competitive play format. Sam’s interpretation is striking:
When it feels good to do the behavior, the habit sticks more naturally the next time the trigger appears in real life (like seeing soap by a sink).
For health and gamification designers, this point is gold: the goal is not just to inform; it is to create positive emotional reinforcement at the moment of action.
What you will learn (and apply) from this talk
If you are designing health gamification projects, this talk can sharpen your approach in several ways:
- How to stop over-indexing on education and start designing for real behavior
- How to treat measurement as a core design responsibility, not an afterthought
- How to structure early discovery so you uncover cultural, logistical, and institutional constraints
- How to validate mechanics with real users before you fall in love with your own cleverness
- How to build interventions that celebrate progress instead of shaming users
- How to think about retention as a design problem, not a user problem
Most importantly, it will help you ask the right question at every stage:
Does this experience make the desired behavior easier, more likely, and emotionally rewarding to repeat?
If not, you might be building something people admire, but do not use.
