ECRS Is Flat at $30M and Aiming at $100M. Pete Catoe Explains Why That Isn't a Contradiction
Pete Catoe bootstrapped ECRS from $7,000 in 1989 to $30 million a year selling point of sale to grocers. Growth has been flat for two years, and he says that's how the business has always worked.
On this page
Pete Catoe has been selling software to grocery stores since 1989, which he started with about seven thousand dollars and has never taken a dollar of outside money for. At the time of this conversation ECRS was doing around $30 million a year across roughly 4,000 customers, growing about 10%, and he had just given his team the number he wants them running at instead: a hundred million.
He also had a plan for what happens to his own equity on the way there, and it involves giving most of it away.
The thesis. ECRS is a company that grows in steps, not curves — flat for two or three years while commitments pile up, then a jump. Catoe’s bet is that converting the company to employee ownership is what makes retailers trust a thirty-year software commitment enough to sign the deals that produce the next step.
What ECRS actually sells
Point of sale is the visible part. Underneath it is the rest of the stack: back-office systems, EDI supplier integration, automatic reorder. Catoe describes the customer base as complex, high-volume retail — health-food stores, grocers, food cooperatives, pharmacies — from single-store operators up to regional chains. There are also a couple of thousand vending and employee break-room units around the country.
The company owns its whole stack, most of it built on open source, which Catoe frames as independence rather than economics.
We own our own stack and it’s not really beholden to one of the major players out there.
Pricing runs from about $100 a month for a vending kiosk to an average of $500–$600 a month across the base. About 90% of revenue is software, licences and services; the rest is hardware. The GetLatka profile records $30M for the date of the interview, up from roughly $28 million a year earlier.
The flat years are the product
Latka pushes hard on the growth rate, because 10% does not obviously lead anywhere near $100 million. Catoe does not dispute the number.
It’s been flat for us last two years, kind of like building years for us. We kind of do that over the years. We’ll flatten out for a few years and then we jump up thirty percent and flatten out again and jump up.
Pressed on where the next jump is visible in the data, he points at something that is not in the revenue line at all: signed commitments that have not started rolling out.
A lot of the business you book now, but you’re not going to start rolling into the next two years. We’re already booking a lot of deals that start rolling out in 2019.
That is the shape of complex retail deployment. A regional chain signs, then spends a year or two on the requirements work before a single store cuts over. Revenue reported today is a lagging read on demand booked two years ago, and it is why a company can look flat while its pipeline is not.
Churn that does not behave like SaaS churn
Asked for a churn number, Catoe reaches for a Michael Dell line about acquiring competitors one customer at a time, and then gives an answer that needs unpacking.
We won’t lose the customer. They might not resign for the support agreement, but they’ll probably continue to use the product.
Latka does the translation on air: that is not logo churn, it is revenue churn, and the logo stays. Catoe puts it at two to three percent of customers a year and says the dollar figure is small enough that he does not track it — “it’s just not even on the radar.” A licence-plus-support model produces a different failure mode from a subscription: the customer can go quiet for a year and come back, because the software never stopped working.
The acquisition math of a bootstrapped company
ECRS is picky about who it takes on, which Catoe frames as a community obligation rather than a sales filter. When it does sign someone, the cost is set as a share of what the customer pays in year one.
Fully loaded cost to acquire ≈ 25% of first-year investmentCatoe’s own framing; larger customers cost proportionally less to board.
On a customer paying $500–$600 a month, first-year value is six or seven thousand dollars, so roughly $1,700 to win them, with a payback Catoe puts at about three months. Latka spends several minutes trying to make the cash-flow risk land — add a thousand customers quickly and you are two million dollars out the door before any of it comes back — and Catoe repeatedly declines the premise.
Those numbers don’t affect us at all. That’s nothing to us. We’re not going to add that many customers. We don’t have to add that many to get to a hundred million.
He is right about the arithmetic. Getting from $30 million to $100 million on an ACV around $6,000 would require an implausible number of small customers. The path he is describing runs through regional chains, where a single signature moves the line.
Giving the company to the people who work there
Catoe owns 100% of ECRS. Over the twenty-four months following the interview he planned to move to an ESOP, starting at 30% employee ownership and going further from there. Asked directly whether that makes him less rich, he answers the business case rather than the personal one.
One of the big problems we have in our industry is the constant churn of these service providers and software providers, and a lot of the retailers just want a company they can count on. These are massive investments and they pay off over many years, so they need partners they can count on that are going to be there.
The mechanics are conventional: a third party values the company, employees accrue shares against salary level, and the company must buy those shares back when someone cashes out. Catoe calls it a second retirement plan. What matters strategically is the sentence he attaches to it — that the ownership structure is a sales argument in a market where the buyer is committing for a decade.
- The rallying cry is the number — $100 million over five years, tied to what employee shares would be worth if they get there.
- The mechanism is trust — retailers buying a ten-year system want a vendor that will outlive its founder.
- The constraint is bootstrapping — no outside capital, no plans to raise, so the ESOP is the only liquidity anyone gets.
Boone, North Carolina
The company is headquartered in a college and tourist town next to Appalachian State University, which Catoe describes as having a world-class computer science school along with strong analytics and design programmes. About 80% of the 150-plus staff are there; the rest are spread around the country, and one person is in Canada — not for a function, he says, just one person, alongside a lot of Canadian customers.
Adding someone every week, he says, in a town most software companies would not think to look at.
Asked what he wishes his twenty-year-old self had known, he keeps it to one line.
I wish I wouldn’t have placed any kind of limitations on myself when I was 23.
Sources Pete Catoe’s interview with Nathan Latka, recorded 26 September 2018; revenue and headcount rows from the GetLatka ECRS profile.
