Invo Is Chasing the Least Glamorous Problem in Food Distribution

The Macro: The Invoice Graveyard Nobody Talks About

Food distribution runs on credit. A distributor delivers product to a restaurant, sends an invoice, and waits. Sometimes they wait 30 days. Sometimes 60. Sometimes the restaurant closes and they never get paid at all. Restaurant failure rates hover around 60% in the first year, and even the ones that survive tend to treat their distributor invoices as the last bill to pay after rent, payroll, and the liquor supplier.

The result is that food distributors carry enormous receivables relative to their margins. A mid-size distributor might do $50 million in annual revenue on 3-5% net margins, with $8 million sitting in unpaid invoices at any given time. Collections in this industry are still overwhelmingly manual. Someone at the distributor picks up the phone, calls the restaurant, asks for money, gets told to call back next week. Repeat.

The enterprise AR automation market has real players. HighRadius is valued at $3.5 billion and serves large corporations. Tesorio handles collections for SaaS companies and mid-market firms. Billtrust (now part of EVC) went public and focuses on B2B payments. YayPay, acquired by Quadient, targets mid-market. But none of these tools were built for the specific dynamics of food distribution, where the customer base is fragmented, the ticket sizes are small, the payment behavior is erratic, and the relationships are deeply personal. A restaurant owner who’s been buying from the same produce distributor for fifteen years doesn’t respond well to automated dunning emails.

The Micro: Collections for People Who Hate Collecting

Invo is building an AI-powered collections platform specifically for food distributors and manufacturers. The vertical focus is the bet. Rather than trying to be a general-purpose AR tool that sort of works for food, they’re building for the exact workflows, payment patterns, and relationship dynamics of this industry.

Patrick Foster, the co-founder, has an interesting background for this kind of problem. He worked on Netflix’s ML platform, then did checkout optimizations at GoDaddy that increased revenue by $1.8M. Techstars hired him while he was still in high school, which is the kind of detail that either impresses you or makes you feel bad about what you were doing in high school. Either way, he’s been building at the intersection of ML and revenue optimization for a while. The company came through YC’s Winter 2025 batch as a two-person team.

The pitch is straightforward. Food distributors plug Invo into their existing systems, and it handles the collections workflow with AI that understands the nuances of the industry. That means knowing when to push, when to wait, which accounts are actually at risk versus just slow, and how to escalate without torching a relationship. The traditional collection agency model takes 25-50% of recovered funds and often alienates the customer permanently. If Invo can recover more while preserving the relationship, the value proposition is clear.

The Verdict

I like this because it’s boring. Not boring in execution, but boring in subject matter. Nobody at a dinner party is going to get excited about accounts receivable automation for food distributors. That’s a feature, not a bug. The unsexy verticals are where you find real pricing power and low competition, because the talented engineers all want to work on consumer AI or autonomous vehicles.

The risk is market size and expansion path. Food distribution is a big industry in aggregate, but individual distributors are often small businesses. The sales motion could be slow and the contract values modest. The question I’d want answered is: how many distributors are large enough to pay meaningful software fees, and can Invo reach them without an expensive field sales team?

The 30-day outlook is probably about proving the core product works with a handful of distributors. At 60 days, they need retention data showing those distributors are actually collecting more. By 90 days, the question becomes whether the unit economics work. If each distributor pays $500-2,000 a month and the product demonstrably reduces DSO by even a few days, the math should pencil out. The biggest risk is that food distributors are technology-conservative and slow to adopt new software. Getting the first twenty customers will be harder than getting the next two hundred.