Picking the "right" message

How Dr. Amy Bucher and Lirio learn

One of the themes that has come up over and over again in my conversations with behavioral scientists is learning - that the field continues to learn and how we learn about patients and consumers.

Dr. Amy Bucher highlighted it in our recent conversation when she spoke about how her team at Lirio works with healthcare customers to drive patient behavior change:

We get asked how many messages does it take to make somebody do "the thing."

It's at least bimodal. We typically see a big spike where there's people who take action pretty much right away. I think these are probably more type A people or maybe we really did send the right communication right away. There's another group, another spike where, five or six messages in, they take action.

One of the questions that we also get is what is the most effective message? What is the one that works? For that group that's converting after five or six messages, I wonder if it's the conversation, not the message. Is the fact that they have heard a couple different things now, that they've gotten a couple different angles on this behavior and it's gotten them to the right place where whatever message was the one most proximal conversion, if that was the first one, maybe it wouldn't have worked without all the other ones.

____

This highlights some really important things to consider in creating behavior change interventions - there isn't one "user journey" - as Amy said many times, personalization will matter (and we can have a whole other conversation on what personalization means). There may be a "dose" effect for some people where they need to accumulate a certain understanding before any message works and it is more about the dose than the personalization (or not) of the most proximal message.

Dr. Bucher went on to note how this learning and dosing is a key part of the model at Lirio and newer intervention research study designs:

What I like about the Lirio technology is that it can help support changes over time because part of what we understand about what we're doing, and I think this is true of really any intervention, is that by intervening, we are changing the state. When I say something to you and I provide you with information or I combat one of your barriers to action, the next time I communicate with you, you are at least slightly changed from the first time.

I think another weakness of personas is that they're not dynamic to capture some of the changes that happen, not just naturally with time, but through this behavior change process. If you think of something like a condition management program, like let's say diabetes management program, people are going to be learning new skills. They're going to get more comfortable with testing their blood sugar. They're going to eventually not struggle as much with the calculations about their insulin dosage. They might start to be able to eyeball like what's a high carb and a low carb food. For each of those things we have to account for the change.

I really love the just in time adaptive intervention or JITAI approach, too, because I think it's brilliant at right sizing the support that we offer to people. It's not just knowledge either. It's habit formation. Some of these things might just become ingrained over time. For example, I take a couple medications in the morning and they're right next to my alarm clock and I never forget them unless I'm not here. If I'm on vacation, that's where I forget. A really good JITAI intervention would be able to track my data and say, "Oh, Amy is going to need more reminders because we just noticed she's probably not where she usually is." Anything that can evolve with the person is really what we need for the sustainable longitudinal behavior change, especially when it is lots of little things that have to be done over and over and over.