Intuit Payroll: How We Increased Emergency Savings by 5X

You’ve likely heard the stat: Americans can’t cover a $400 emergency. This figure has been circulating in nonprofit and policy circles for over a decade. And this is important because Vanguard researchers found that emergency savings are the strongest predictor of financial well-being.

Irrational Labs took a behavioral design lens to the problem to see if we could increase emergency savings.

When we looked deeper, we realized there was already a playbook: Retirement savings.

Before automatic 401(k) enrollment, only 47% of Americans saved for retirement. Once it became standard in the 1990s, participation shot over 90%. Without automatic enrollment, most Americans would not be retiring anytime soon.

Automatic enrollment works because of two key behavioral factors:

Timing: Enrollment in retirement savings happens when someone starts a new job and is already making payroll decisions.
Default effect: It requires people to take action not to save vs. to save. Inertia works in favor of the desired behavior.

Together, these insights created one of the most successful financial interventions in modern history. Could we copy-and-paste them into emergency savings?

Not exactly.

There’s a key constraint: it’s illegal to automatically open a new bank account on someone’s behalf. Unlike with retirement savings, people will therefore always need to take an action on their own behalf.

But that didn’t mean we had no levers. The payroll moment was still a chance to make saving feel like the natural next step.

Enter: Payroll Providers

Most payroll providers allow employees to split direct deposit into two accounts, so, in theory, a portion of their regular earnings could be sent straight into savings.

But behaviorally, this approach falls short because it lacks:

Any recommendation to use as a second account
No guidance on how much to allocate
Framing for what the second account is for, such as savings or emergencies

What it looks like:

Most payroll platforms look the same. Rippling, for instance, has you connect a checking account. Then, in light gray text, it offers an option to “Add a Bank Account.”
Intuit Payroll works similarly — you can add a second account, but only if you know to expand the dropdown.

Imagine if retirement savings had been tucked away in a dropdown menu during payroll setup. How many Americans would be saving for retirement today? Not many.

When we spoke with Intuit Payroll at the start of this project, they told us that only 2.9% of new employees set up automatic transfers to a second account during payroll enrollment.

Technically, people could save at the payroll moment. But almost no one did. Was that because they didn’t want to save, or because the UX made it too hard?

That’s the definition of latent demand: people didn’t ask for e-bikes before they appeared in cities, but once they were everywhere, people used them.

We set out to test whether this was really a latent demand problem. To answer that question, we partnered with Intuit Payroll to design a test intervention that intercepted employees during payroll setup — the exact moment they were already deciding how to allocate their money.

The Opportunity: How Would a Behavioral Scientist Design the Payroll Moment?

We couldn’t default people into savings, but we could make savings feel like the normal path.

This approach has worked in healthcare. For example, doctors who assume patients want a flu shot and ask when they’d like it (rather than if they want it) achieve higher vaccination rates (Chapman et al., 2016). Patients can always say no but the default assumption is they would like a shot vs they wouldn’t like a shot.

We applied the same principle to payroll design. Instead of asking, “Do you want to save?” we built a “continue” mindset in the UX, framing “saving” as the natural next step. Employees could always opt out, but the flow made it feel like saving was the expected choice.

The Design Questions

We faced two critical design questions:

1) What numerical saving options should we present?

Anchors matter. If we set anchors too high, people might reject every option: think of exaggerated tipping defaults that push people to opt out entirely. Too high, and people could also overdraft their accounts—a risk we’ve seen in retirement savings when people can over-allocate and then accumulate debt elsewhere.

So we worked backwards from the goal: how much would help people avoid debt and what did they consider a reasonable emergency cushion?

StepChange research shows that £1,000 (around $1,300) of savings reduces the likelihood of falling into problem debt by 44%.
Duke’s Common Cents Lab found that 48% of savings goals cluster around multiples of $500 or $1,000.
LendingClub reported the average emergency expense in 2023 was $1,400–$1,700.
Pop financial expert Dave Ramsey advises a $1,000 buffer before tackling debt.

Taken together, these benchmarks pointed us toward a $1,000 annual cushion. Looking at median take-home pay, we backed into an anchor of $50 per paycheck and 3% of pay.

2) Should we frame savings in dollars or percentages?

We quickly ruled out “set a goal” prompts. That’s common in financial wellness apps, but it often backfires by anchoring people on unrealistic targets. Research in charitable donations shows small periodic amounts (“$1 a day”) drive higher opt-in rates than large lump sums (“$365 a year”) (Atlas & Bartels, 2018; Gourville, 1998–2003). Behavioral scientist Hal Hershfield has also shown the “small-number effect” is real: framing deposits in daily amounts as opposed to monthly amounts quadruples the number of consumers who enroll. So we stripped it down: Just ask people how much they wanted to save each paycheck. But this also revealed the real debate — should we ask in dollars ($50) or percentages (3%)?

Experts in low-income decision-making argued for dollars. A concrete figure like $20 or $50 feels manageable and reduces fear of overdrafting.
Experts in retirement savings argued for percentages. Workers are familiar with the 3% frame from retirement plans, so why not use the same logic here?

With no consensus, we decided to test both.

The Experiment

In December, 2024, we randomly assigned nearly 90,000 employees across 69,000 companies to one of three groups:

Control: the standard Intuit Payroll experience with the drop down
Monetary condition: dollar-framed choices ($20, $50, $75)
Percentage condition: percent-framed choices (1%, 3%, 5%)

What is the population?

Our sample looks more like an hourly workforce—and a lower-income one—than the U.S. workforce overall. Seventy-four percent of participants were hourly (versus ~55.6% nationally), and 90% were paid weekly or bi-weekly. The median annual income across the full sample was $26,400, reflecting that hourly workers dominate the composition. Within the sample, hourly workers earned an average of $22,568 per year, while salaried workers earned an average of $52,988.

Results: Do People Want to Save?

Yes. When prompted, 30.6% of employees chose a savings amount — clear evidence of latent demand. People want to save.

But friction mattered. To finish setup, employees had to enter savings account and routing information. Many dropped off at this step. 38.4% of those who chose an amount never entered their savings account and routing information.

Even with the nearly 40% drop off, the impact was striking.

On average 19.9% of employees across the treatment groups successfully set up savings accounts, compared with just 3.2% in the control.

Which Approach Worked Best?

The percentage framing outperformed monetary by 13% and drove a 5.6x increase over control. Treatment condition was the strongest predictor of savings behavior—even more than income, pay type, or pay frequency.

Why did the percentage condition work? Our hypothesis is the small number effect. “3%” just feels like less than “$50.” People also know exactly what $50 buys, while 3% of pay is abstract and therefore feels more manageable. Framing mattered.

Control: 3.2%
Monetary: 18.6%
Percentage: 21.1%

Did People Stick With It?

Most did. From enrollment through six months, participation dropped by an average of 12% across conditions. Once people set up savings, they kept saving.

How Much Did People Save?

Anchors shaped behavior. 75% of Percentage participants and 91% of Monetary participants chose one of the predefined options.
The middle “recommended” option was most popular, selected by over half of Percentage participants and 43% of Dollars participants.

On average, employees in the Percentage condition contributed 3.79% of pay. In the Monetary condition, the average was $54.80 per paycheck.

Do High Income Participants Save More?

Income was a factor in the control group. In the control group, lower-income workers were far less likely to set up a second account. But in both treatment conditions, that gap narrowed significantly. The Percentage condition especially helped equalize participation across income levels.

Contribution amounts told a similar story. Higher-income workers saved more in the Dollars condition. But in the Percentage condition, both low- and high-income workers contributed at similar rates relative to their paychecks.

Why This Matters

This experiment demonstrates the power of system design. A small change at the payroll moment led to nearly one in five workers setting up automatic savings — a six-fold increase over the current baseline. People not only enrolled at higher rates, they stayed enrolled, and the effect held across income levels. These results suggest that emergency savings can be scaled when the right defaults and prompts are built into everyday financial systems.

The takeaways:

Payroll providers: Enrollment flows are a powerful lever. Embedding savings prompts at payroll setup could move millions of workers into healthier financial habits. Intuit payroll is planning to weave savings habits into their employee experiences.
Banks: Banks live off their deposits. One way to increase deposits is to help people build up a savings buffer. Prompting customers to set up automatic transfers that allow them to “save when you get paid” at the moment of account origination is a low-friction way to drive deposits.
Employers: Companies influence the design of payroll systems. By asking providers to add savings features, or paying for them directly, employers can deliver a measurable benefit to workers.
Policymakers: Sidecar savings tied to retirement plans have struggled to gain traction. More effective policies could include promoting liquid, no-fee savings accounts, regulating overdraft fees, and lifting outdated withdrawal caps.

Together, these shifts point to a simple truth: small design changes at routine financial moments can unlock savings that traditional approaches never reached. If scaled, this kind of program could reshape how millions of workers build financial security.