Since its founding, Even’s been known for “evening”* income. As of this week, we won’t be anymore.
As I’ll describe in this post, that’s because the story of how even income was going to help Even’s members (our term for users), though compelling in the abstract, ultimately did not survive its encounter with the story of members’ real lives.
The basic goal of Even’s signature feature, Pay Protection, was to turn paychecks that go up and down (usually due to hourly pay), like this—
—into a steady stream of average income, like this:
Looking forward to at least their average pay every payday was meant to make it much easier for members to plan out their expenses. With a solid plan in place, they’d be able to pay bills on time and have enough for regular discretionary spending, too, without relying on overdrafts, credit cards, and payday loans. The money they used to waste on expensive forms of short-term credit could go toward building liquid wealth instead. All this would be done automatically, through deposits and withdrawals to and from members’ accounts, so members wouldn’t have to do any work. Pretty neat, right?
As Ryan puts it, Pay Protection was a really good story. It was novel, rational, and simple. It was hopeful.
But as we started learning pretty early on, the story was flawed. First, lots of potential members weren’t as enamored of the concept of even pay as we were. Second, income-evening alone didn’t make money-management easy enough, because most members’ expenses were still extremely volatile.
In the face of this evidence, instead of doubling back, we doubled down. We decided to even not just income, but expenses as well, by embarking on the hugely ambitious and fiendishly difficult project of using neural networks to model members’ recurring expenses.
The decision fit with our story: that we were going to automate away all the stress and busywork of managing money. We started referring to what we were building as the “self-driving car of finance.” As we got to work on the car, we continued to see evidence that Pay Protection wasn’t working great for members, but dismissed it as irrelevant given our new plan.
Thanks to our incredible engineers (I can brag about them since I’m not one of them), we made significant progress on modeling bills, which you can experience in the app that’s available today. But, it took much longer than anticipated to understand all the factors around expenses; and, we realized there were some factors we might never understand well enough, since the way people pay bills is so idiosyncratic.
The reality was that even as our app got better and better at automatically driving finances, there was still going to be at least some period during which people would continue to drive themselves. That meant it was time to turn back and look at the hard facts of what Pay Protection was like in a world of human drivers.
Since people in the paycheck-to-paycheck cycle tend to be “zoomed-in” on their finances, budgeting paycheck by paycheck, the best way to understand how Pay Protection affected people is by zooming in on how its mechanics played out in individual pay periods.
As background, the typical shape of a pay period for an Even member living paycheck-to-paycheck looks like this: a sharp spike up when you get paid, a quick dip back down as you pay bills (and sometimes save), and then a slow downward slope as you spend the rest.
There was one version of this story in which Pay Protection worked really, really well: when a low paycheck directly preceded high expenses. That story looked like this:
In this situation, since the member would have run out of money without a boost, Pay Protection was really valuable.
In other iterations of the story, however, Pay Protection did not come out like such a hero. For example, when a member got a low paycheck, but also had low expenses during that pay period, the extra money from Even was unnecessary and (members told us) a temptation to overspend:
Another example: If a member received an average-sized paycheck, but had extra-high expenses to pay out of it, this could happen:
In the circumstances above, though the member really needed a cash infusion, Pay Protection didn’t send one, since the paycheck wasn’t lower than average. That was not very helpful.
Though we’ll never forget the many stories about how Pay Protection let people keep the lights on, buy food for their families, recover after emergencies, and just sleep better at night, there were too many stories in which Pay Protection didn’t solve members’ problems well enough. So, we made the tough decision to retire Pay Protection for good.
After studying all the stories about how members used Pay Protection, two things were very clear:
Putting those two points together, Even decided to continue offering an income-smoothing feature —but a much more user-driven one.
The new feature, called Instapay, lets a member choose how much of the pay they’ve already earned in a pay period (up to 50% of net pay) they’d like sent to them before payday. The amount they chose to get early then comes back out of their very next paycheck, automatically. Instapay works via employer payroll integration and is available to members who work for Even’s partner employers, such as Walmart.
With the switch to Instapay, Even members have taken the wheel of income-smoothing. And with their upcoming bills listed in the app, they’ve got better visibility on expenses. Our members are driving. And meanwhile, we’re improving the features that make driving easier. Somewhere down the road, we’ll reach that point where members do only the minimum amount of work to keep things headed in the right direction. But we won’t cross that bridge till we come to it. This time, we won’t let our product story get too far ahead of the story our members are living.
*Little-known fact: The name “Even” actually comes not from the idea of even pay, but from our mission of evening the playing field by providing quality, affordable tools for money management, so that more Americans can realize their potential.
Get updates around new research and findings in your email.