In 2015, Allison Ellsworth was mixing prebiotic soda in her kitchen in Dallas, trying to fix her own gut health with something that didn't taste like penance. Ten years later, PepsiCo paid $1.95 billion for the company that grew out of it, with $300 million of that structured as anticipated cash tax benefits, putting the net purchase price at $1.65 billion, according to CNBC. Allison and her husband Stephen owned an estimated 12% of Poppi at the time of sale, a stake worth about $150 million after tax, per Forbes.
The number that matters more, for anyone weighing whether to start a company with a spouse, is the one before the exit. In the first year, the Ellsworths put in roughly $90,000 of their own money, according to Fortune. They sold both of their cars and maxed out credit cards to keep the business running, according to reporting from Yahoo. They did this while raising three young children. There was no car-free household as a lifestyle choice. There was one car, sometimes none, and a bet that a canned soda would work.
The bet was not obviously good
Three weeks after they started selling at farmers markets under the name Mother Beverage in 2016, Whole Foods found them, according to CNBC. That is a fast discovery by any startup standard, and it's also the kind of detail that makes the rest of the climb look smoother in hindsight than it was. Two years later, in 2018, the company was doing about $500,000 in revenue, per Parade. That's a real business, not a hobby, but it's also nowhere near the valuation that would eventually get attached to the name. Half a million dollars in revenue, after three years of two people's savings and credit and one working car between them.
That same year, Allison pitched Poppi on Shark Tank nine months pregnant and walked away with a $400,000 investment from Rohan Oza for 25% equity, according to Wikipedia's entry on Oza. The deal did more than fund the next phase. It came with a rebrand in 2020 that repositioned the product entirely, and the new version generated $100,000 in sales in 24 hours off a single TikTok post, according to Fortune. The line from $500,000 to $1.95 billion runs through that rebrand, and the rebrand runs through a deal negotiated while one founder was, by her own account, close to giving birth.
What the research actually says about married co-founders
The opposing case here is not a straw man. Noam Wasserman, dean of Yeshiva University's business school, has built a body of research on startup founder conflict, and the headline finding is stark: disputes between co-founders are responsible for 65% of failed startups. His data also found that a preexisting relationship between co-founders, which includes marriage, increased the likelihood of a co-founder's departure by 28.6%. That number should sit uncomfortably with anyone romanticizing the founder-couple story. It suggests that the bond you'd assume protects a business partnership is, statistically, associated with a higher chance that one partner walks.
Investors have absorbed this. A married founding team has traditionally been read as a single point of failure wearing two name tags: if the marriage strains, the cap table doesn't care about your custody arrangement. Two people who share a mortgage, a bed, and a board seat have no institutional wall between a bad quarter and a bad Tuesday night. That is the strongest version of the case against building a company with your spouse, and it deserves to be stated plainly rather than waved off with an anecdote about how well one particular couple gets along.
Where the Ellsworth case answers it, and where it doesn't
The honest answer is that Wasserman's data describes a risk, not a verdict. The Ellsworths didn't avoid the conditions that produce founder conflict, they operated inside them for a decade: undercapitalized, underslept, raising three kids, selling assets to make payroll. Those are exactly the pressures that Wasserman's research would predict should strain a marital co-founder relationship past its breaking point. It didn't happen here, and the resulting sale numbers are what they are: $1.95 billion, a $150 million payout to the founding couple, and by July 2026, Allison Ellsworth had signed with WME and collected a TIME 100 Next nod, an Inc. Female Founders 500 spot, and BevNet's Person of the Year, all from 2025, according to Yahoo Finance.
But one outcome is not a refutation of a statistical pattern. Wasserman's 65% figure is drawn from a wide sample of failed startups broadly, and the 28.6% increase in co-founder departure risk applies to preexisting relationships across the board, not exclusively spouses. Poppi's survival doesn't tell you the base rate improved. It tells you that a specific couple, with a specific division of labor and a specific investor who came in with equity and operating leverage at a key moment, cleared the risk. Rohan Oza's $400,000 for 25% wasn't just capital. It brought outside judgment into a two-person system at the exact moment that system needed a tiebreaker who wasn't married to either party.
The condition that makes this work, and the one that doesn't
Operators who've built businesses with a spouse report that the arrangement holds when the roles are genuinely distinct, not duplicated. If both founders are fighting over the same decisions, brand voice, hiring calls, fundraising terms, the marriage supplies no advantage over any other co-founder pairing and inherits all of its downside plus the domestic one. The Ellsworth story, as reported, shows a couple operating through years of resource scarcity toward a single external validator (Whole Foods, then Oza, then TikTok) rather than litigating direction between themselves in a vacuum.
The research shows the risk is real and measurable: 65% of startup failures tied to co-founder conflict, a 28.6% higher departure rate where a personal relationship preceded the business one. Our view is that this doesn't mean don't do it. It means the decision to found a company with your spouse should be underwritten the same way a professional investor underwrites any co-founder team: what happens to equity and operating control if one of you wants out. Couples who skip that conversation because "we're married, it's fine" are the ones who supply Wasserman's numerator.
Who this advice serves, and who it doesn't
The people best served by the "build it with your spouse" model are couples who can genuinely split domains, who have, or can raise, enough capital that the early cash crunch doesn't become the only topic of conversation at home, and who are prepared to bring in outside equity holders early, the way Oza's $400,000 stake did for the Ellsworths. The people worst served are couples treating a joint venture as a way to spend more time together or repair a strained marriage. A cap table is not couples therapy, and a company under financial stress, the kind that requires selling both cars, will amplify whatever is already fragile in a relationship rather than fix it.
The number worth sitting with isn't the $1.95 billion. It's the $90,000 they put in during year one, against $500,000 in revenue three years later, funded by selling the cars and running up the cards, with three kids in the house. That ratio, more sacrifice than return, for years, is what building a company with your spouse actually costs before anyone finds out if it works. The exit is the exception that gets covered. The years before it are the rule anyone considering this should price in first.