🎬 Every week I share a short (fun!) product video teardown and some hot takes on behavior change. Pro tip: The video only takes 〜5.5 minutes at 2x speed. And it will be more memorable. ;) ⏩
Excited about this teardown because it closes a loop. Two years ago, I went on record with a teardown and told Intuit that payroll onboarding was the single biggest lever for driving short-term savings in America, and that the whole industry was leaving it on the table.
Intuit saw the teardown and leaned in, and now the results are in: two questions, added at the right moment in Intuit’s payroll onboarding, increased savings 5x. In the winning condition, people will save an incremental $1156 annually.
Behavioral scientists build hypotheses from existing research on what’s worked and what hasn’t, and the core idea behind this project came from retirement savings. When people are automatically enrolled in a retirement plan, over 90% save; when they’re not, less than 50% do. That gap has nothing to do with how much anyone likes saving. It’s the design of the form that decides if they save or if they don’t save.
What if short-term savings could work the same way? 36% of Americans struggle to cover an unexpected $400 expense. We need better solutions to support building up an emergency fund. In a perfect world, we could replicate the retirement playbook and just auto-enroll everyone in a short-term savings account. However, it’s not that easy. Legally, we can’t open a short-term savings account for someone automatically. We need to ask people if they want to do this.
How should we ask people if they want to save? Over the years our team has tried and failed to get people to save. We have worked with banks that prompt users to open a savings account, and single-digit percentages say yes. You get a pop-up and quickly dismiss it because you’re actually logging in to pay bills. It’s the wrong moment. The right moment is when people are actually thinking about their money. One of those moments is when they are setting up payroll for the first time. So we paired the ask with that moment: the onboarding of payroll.
When I did the initial teardown of Intuit’s payroll system, it was clear they already had the technology for this. In settings, there’s a way to split your paycheck across two accounts, and around 3% of people were already doing it. Our job was to move it into the onboarding flow, and two design decisions shaped how.
We made savings the path of least resistance by putting it in the flow.
Saving isn’t a visible social norm in the U.S.; I don’t know what my friends save or what’s in their accounts, so a new employee looking at this screen has no anchor for what people like them do. The research we pulled from was flu shots: when the shot feels like the natural next step (a doctor schedules it for you), more people get one than when they’re asked to go schedule it themselves. In our design, that translated to a button that simply says Next. There are multiple ways to opt out; anyone can go back or decline to set up savings, so no one is forced in. The design just removes the pause to wonder whether it’s the right thing to do.We gave them options on how much to save and a recommendation.
To do this, we had to figure out a key question: should we ask people to save in dollars or percentages? The experts were split. A dollar amount is concrete, so a worker knows exactly what’s leaving each check; a percentage scales down automatically when a paycheck is small. When smart people disagree, we run the experiment. So the study had a control and both conditions. We showed three amounts with a recommended middle option, and the question shifted from “do you want savings?” to “which of these amounts do you want to save from each paycheck?”
It’s the toddler principle: you never ask a toddler to put on a coat, you ask which coat.
Okay, drum roll: around 20% of people saved, relative to a 3% control. Percentage beat dollar amount. Most people chose the recommended middle option (3.7% in the percentage condition, an average of $54 in the dollar condition), and income had little effect on who opted in. We let the study run for six months, and people aren’t going back and changing it, so we’re fairly confident people want this.
The 20% who saved actually understates the demand: roughly 30% of people wanted to save. 30% of people completed the first two steps but dropped off at the last step of entering a routing number. That’s a 40% dropoff. Friction matters.
The lesson here is, just ask. When the McDonald’s drive-thru asks if I want fries with that, I say yeah, sure, great. I like fries; I just wasn’t thinking about them until someone asked at exactly the right moment. Savings works the same way. People do want to save, and when nobody asks at the right time, it looks like they don’t.
Put the ask where people are already thinking about money: payroll setup, not a banking session spent paying bills.
Make saving the natural next step, with a Next button and easy ways to decline.
Ask “which amount?” with three options and a recommended middle, rather than “do you want savings?”
Remove friction at the finish line: syncing an account beats typing a routing number.
Thank you, Intuit, for listening and for building this with us. If you’re a financial company and want to roll this template out, or if your product has been included in a teardown here (Monarch, Eight Sleep, Calm, I’m looking at you!), let’s do something together: kristen@irrationallabs.com.
🎬 The post above is just a preview. Only takes 〜5.5 minutes at 2x speed. ⏩
Have a friend who would enjoy these teardowns? Click the button below to refer them (& earn some great rewards).👇
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.