In finance, healthcare, and insurance, launch strategy is UX
Generated with Midjourney.
One common question we are asked is if organizations in highly regulated industries should phase their rollouts? The answer is if you can, yes, and there’s a specific way to do it for best results. Let’s dive in.
Soft launch, firm advantages for highly regulated industries
Phased rollouts (where we deploy to a certain subset of users first instead of enabling for all users at once) enables us to learn about how users will perceive changes, using a smaller group. This avoids overwhelming support channels and allows us to tailor the experience for the next group.
This isn’t a replacement for testing (automated and live), preliminary UX research, usability testing, or any of the phases that go before it. What it does that nothing else can fully replicate is get a real user, with real tasks to accomplish, to use your product in the context of their normal day. In other words, the mindset is well, real. And you can’t reproduce that in a focus group or observation session.
Let’s say your app is designed to be used by people who are going through an intense medical event, such as navigating a cancer journey. Interviewing people with exposure to a cancer journey in a focus group during initial UX research is valuable to get some general impressions and get things most of the way there, but there’s no way it can reproduce the emotional conditions they will have in real life when opening the app after a chemotherapy treatment has left them wiped out and questioning if they can continue.
Yet a phased rollout can get you this kind of deep insight.
Let’s go back to our principles of building an emotionally connected product experience:
“I see you” → Get the basics right → Introduce new pathways
Let’s break down each component of that framework and how it applies here.
“I see you”: To build trust, we must first establish to the user that we know and understand them (1). To truly see someone, we need to meet them where they are at the moment of entry into the feature or even the platform as a whole. Getting that kind of live, in-the-moment feedback can confer a critical edge. One time this kind of mindset feedback told us that caregivers didn’t want to be told how noble their service was, because they have complex feelings about what they’re doing. It turned out that emphasizing it as a virtuous act was at odds with how people saw themselves, and immediately invalidated the attempts to connect — in other words, it pushed people away. That insight was crucial and only happens if you get people in the moment. You can’t get that kind of insight from a survey.
Get the basics right: This is another area where people just want a smooth experience at the point of entry, especially in highly regulated industries. They’re completing a critical task such as dealing with their insurance or money, and they need to be reassured that you’re running a tight ship. It’s important to remember that for every edge case, while it might be an outlier to the organization statistically, to that user, it’s 100% of their experience at that moment. The small sample size allows you to diagnose early issues. For example, in one product, we found that a segment of users were put off by Meta’s description of the permissions users needed to grant in order to use a Facebook integration and were less likely to convert. This didn’t show up until real users were granting access to their real social media accounts. That feedback allowed us to tailor the ‘bookends’ that prompted the user for the permissions to make sure they understood what was happening, thus building trust and increasing the conversion rate.
Introduce new pathways: This is where it gets interesting. Once we’ve built trust by showing the user that we see them and we’ll take care of their basic needs for security and privacy, we can introduce them to new pathways, such as the new features we want them to love. They’re primed to be more receptive because we followed a proven sequence to earn their trust.
Like a restaurant soft opening, the best phased rollout strategies employ a specially curated segment of the user base who are willing participants and can be relied upon to provide detailed feedback on their experience. If you’re a card-carrying Green Shirt visionary, this might feel like a delay in the glory of your big launch — but rest assured, the payoff is worth it because it helps avoid a major launch pitfall.
The Two Weeks of Ire
To understand why this is a good idea, we first need to acknowledge an important truth: users hate change. Or more specifically, they like the idea of change when it’s an abstract concept, but they hate being forced into the moment of change without their consent. This is present in every industry, but magnified in highly regulated industries, where the emotional stakes feel higher to the user.
Forcing users into change tends to trigger a phenomenon we refer to as the Two Weeks of Ire.
You can do all of your research, build a feature that helps people in every possible way, validate it with stakeholders, run your focus groups, test it every which way, and still, if users are forced into accepting the change, it’s likely they will hate it, at least at first.
The good news is that if it truly is a well-designed and developed feature, most of the users will eventually get over this. And in our experience, the shift from “What fresh hell is this?” to “Oh, that’s kind of cool.” takes about two weeks.
The bad news is, some users don’t really ever get over it. They may leave a negative review, or call customer service in frustration or close the app in the heat of the moment — all actions that cement the negative memory in their mind where it will sharpen over time (2).
Timeline of the Two Weeks of Ire.
Most of the users’ frustration here comes from the lack of choice and feelings of being out of control. They had their corner of the world all figured out, and something came in and changed it on them. These feelings are magnified in settings where emotions are already elevated, like insurance, healthcare and finance.
The near-magic solution for a better launch experience
There is a near-magic solution to this problem, and it involves a spectacularly useful quirk of humans which is that we perceive almost everything to be better if we chose it ourselves (3).
Your neighbor who will stand out in the driveway for 20 minutes telling you about how he selected his new truck is the classic example of this: his pride is partially in the truck itself, but largely in the wisdom of his own selection of that truck, which then reinforces further in his mind why that truck is the best (4). It’s the perfect positive spiral, and you can use this to improve the release experience for your users.
If we want a phased rollout, and choice is the antidote to the Two Weeks of Ire, then at the hallowed intersection of the two is the Opt-In Release.
What’s an opt-in release?
An opt-in release is when you let users elect to upgrade to your new experience. This generally happens over a period of time, with increasingly assertive messaging until your chosen cutoff date. Ideally, even then, there’s a gate of some kind but the user still hits a button to choose the new experience.
In one example of this, we worked with a B2C platform that wanted users to upgrade to a new experience, and it was going to involve a somewhat involved process: integrating multiple third-party accounts. The integrations were critical to the success of the platform, but early research showed user apprehension around the process; they’d had negative experiences with similar apps in the past and were inclined to avoid the process.
We created a 60-day runway, giving users ample communication centered around them and how it would improve their experience. We created touchpoints within flows they were already engaging with, offering them the opportunity to upgrade if they wanted, and advising them of the approaching deadline. About 35% of monthly active users upgraded on their own in the first month. As the deadline drew near, they were able to capture over 50% of engaged users prior to the deadline.
Once the deadline arrived, the upgrade wasn’t treated as a barrier to engaging with the platform — it was used as a carrot. When a user initiated engagement with a chosen high-value feature, they were told that the feature required the upgrade, and led through the process at that time.
This provided a smoother experience for the client organization and the users. And best of all, no Two Weeks of Ire.
When not to use an opt-in release in highly regulated industries
This isn’t feasible for every release. If there’s a high priority security fix, that needs to be a mass, immediate release. You can still use the principles of “I see you” → Get the basics right → Introduce new pathways to soften the communication around the experience.
Some people are tempted to try to flip this and do a forced upgrade with an opt-out, i.e. “Switch back to the old experience.” This is not the same experience, and generally not advised. What you gain in inflated adoption metrics, you more than make up for with increased rage. A forced experience with an escape hatch is still a forced experience.
There are times when it’s just not practical to maintain the old and the new for a period of time due to the nature of the upgrade. However, the emotional connection framework still holds.
You are skewing your early data a bit by doing an opt-in release. You tend to get the people who enjoy the cutting edge and are a little more risk-tolerant to opt in first. This is a worthwhile tradeoff as long as you are aware of the inherent bias so you can account for it.
Just like Sade: soft and smooth
Opt-in soft releases are a great way to build user trust in highly regulated environments like insurance and finance. They give users something they often lack in these systems: a sense of agency. In complex, emotionally charged experiences, that feeling — the belief that I get to choose this — can be the difference between resistance and adoption, between anxiety and confidence, between distrust and loyalty.
Notes
(1) Morelli, S. A., Torre, J. B., & Eisenberger, N. I. (2014). The neural bases of feeling understood and not understood. Social Cognitive and Affective Neuroscience. https://pmc.ncbi.nlm.nih.gov/articles/PMC4249470/
(2) https://pmc.ncbi.nlm.nih.gov/articles/PMC2676782/
(3) https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2017.02062/full
(4) Randy, you know we all love you. And that glorious Honda Ridgeline.