Moonlight: 10,000 Engineers, $10,000/Month in Subscription Revenue — Choosing “Sell, Not Raise” in the Pandemic

(Dollar amounts in this article include a rough conversion at ¥150 to $1.)

This is the story of a marketplace built from nothing more than the founders’ own desire to work while traveling the world, and how it got acquired three years later. Emma Lawler and Philip Thomas started Moonlight in 2017 to connect software engineers who wanted to work remotely with companies that needed development talent. By 2020, the platform had roughly 10,000 registered engineers, thousands of client companies, $10,000 a month in subscription revenue, and $55,000 a month passing through the platform to developers. Then COVID hit, and the two chose to sell rather than raise more capital.

Three years, in sequence

A market built to solve their own problem

The founding motivation was their own life design, not market research. ‘A lot of people were moving to more livable places, like Austin. We wanted to do that too, but we didn’t want to give up access to high-quality work,’ Lawler explains. Keep landing good work while moving around. That need turned into something they could sell to others, and that became Moonlight.

Their approach to building the supply side is distinctive. The two targeted cities outside Silicon Valley. ‘We picked cities that already had software development and startup communities but weren’t yet known as tech hubs.’ Those places already had skilled engineers, but no route to access high-quality remote work. They bet on that mismatch, demand concentrated in San Francisco, supply scattered elsewhere.

And the means were remarkably unglamorous. In each city they moved to, the two ran weekly meetups for tech workers and published a newsletter. ‘One of the big problems with remote work is the loss of community. It can feel like you’re throwing away money, but doing things that don’t scale is something people really remember.’ Rather than buying sign-ups with ads, they moved to a place and showed up in person. The relocation every three months was simultaneously a lifestyle and a sales activity.

Shifting the revenue model, and its cost

Early on, Moonlight took a 10–15% cut of what companies paid. It later switched this to a flat subscription fee, dropping the take from hourly rates entirely. By the time of the acquisition, the structure was $10,000/month staying with the company and $55,000/month passing through to developers.

This shift changes the nature of the marketplace. Under a commission model, growth in transaction volume flows directly into revenue; under a flat fee, revenue only grows if the number of companies grows. On the other hand, companies benefit more the more they use the service, making heavy users less likely to churn. Applying the old 10–15% to a $55,000 pass-through would yield $5,500–8,250/month, meaning revenue was actually higher after the switch, in raw terms, though that figure depends on the state of ‘heavy-usage companies paying a flat rate.’ It reads as a decision to shift the upside from an uncapped commission structure toward accumulating more companies.

The decisive moment was March 2020

The moment the trajectory shifted can be pinned down: the arrival of COVID. ‘It was a weird time to raise money,’ Lawler recalls. Their options had narrowed to two: raise the next round, or find an exit. After talking it through with investors, they chose the latter. And the decision criterion wasn’t purely their own return, as she puts it, making sure members on the platform didn’t lose access to work during an uncertain time was the priority.

The buyer was code review service PullRequest, and talks started through an investor introduction. PullRequest founder Lyal Avery valued Moonlight’s network of 10,000 developers both as a human resource for correcting code errors and as an asset useful for machine-learning training. This was less a valuation of the marketplace’s own revenue than a price on the supply network it could feed directly into the buyer’s core business. Consideration was a mix of cash and stock. The amount wasn’t disclosed.

Why the negotiation went smoothly

The factor Lawler herself cites is unglamorous: ‘We had real revenue, we could see exactly what was happening, and all our data was organized.’ Being able to explain the $10,000 subscription figure separately from the $55,000 pass-through figure (that is, keeping company revenue and transaction volume from getting conflated in the books) directly became low explanation cost. A small marketplace has plenty of ways to inflate its numbers, they simply didn’t.

Two more conditions applied. One is that Lawler had watched an acquisition process from the inside as an early employee at FitStar (later acquired by FitBit). A seller who knows how an acquisition unfolds is less likely to get swept along at the other side’s pace. The other is the investor’s role, someone to talk through ‘raise or exit’ with, and simultaneously the route that brought the buyer in the first place. A bootstrapped business would have to fill both of those roles itself.

What can’t be glossed over

After the acquisition, the two stayed on as employees for a while. Lawler doesn’t evaluate that period kindly. ‘Going from founder to employee at another company is really strange. I spent three years building this business, had complete autonomy, and then went back to having a boss and caring about someone else’s milestones.’ On being co-founders as well as a couple, she’s said, ‘Would I do it again? Probably not,’ and, ‘It was too much to handle at once’ (the two reportedly remain a couple).

The business terms carry caveats too. The sale price was undisclosed, and since the consideration included stock, the actual take-home depends on how the buyer performs afterward. The trigger for the sale was a deteriorating funding environment, not growth, and $10,000/month in subscription revenue is small for a marketplace run for three years. This isn’t a story of ‘cashing out at a high price’. It’s more accurately read as a case where ‘when the external environment closed off, having well-organized revenue and data kept options open.’

What transfers, what doesn’t

What’s easiest to reproduce is, first, how the numbers were built: recording company revenue separately from transaction volume, and keeping it explainable to anyone at any time. That’s useful even when a sale isn’t planned, especially at moments when external factors suddenly change whether raising is possible. Next, the approach to building supply: find the mismatch between where demand is concentrated and where supply is dormant, and go there physically. In this case, the value of doing things that don’t scale was recovered in the form of 10,000 registered users.

The non-replicable conditions are equally clear: Bay Area connections that got them to their first raise at all (she herself admits, ‘The first fundraise was really hard’), an investor with acquisition experience who functioned as an introduction channel, and Lawler’s own prior experience with a sale. On top of that, a pandemic acting as an external shock that created a ‘decide now’ deadline isn’t something you can engineer on purpose. Her words land, in the end, on this: ‘The only thing that moves a business forward is putting one foot in front of the other and continuing to move. No one else will do it for you.’

Related reading

• MENTA — a domestic case of an individual-built matching service passing to a major player in the same industry
• ScrapingBee — the exit for a small-team SaaS built without outside capital