Astronomer Hit $600K ARR With 20 Customers and a $600 Marketing Budget

Nathan Latka asked Ry Walker what Astronomer did in revenue in 2015. The answer was one word: “Zero.” Then he asked about 2016. “We ended the year at a 600K ARR run rate,” Walker said. “So we went from zero to 600.” What makes that first revenue year worth reading is what Astronomer was actually selling to get there — because in January 2017, when this interview was taped, the data platform the company existed to build was not finished, and it was not the thing anyone was paying for.

Two figures circulate for this interview and both are true at different moments. The $600K is the annual run rate Astronomer exited December 2016 on — $50,000 of monthly recurring revenue, which Latka checked on tape and Walker confirmed. By the January 2017 taping it had moved to “somewhere around 55, 60,000 bucks in MRR,” and GetLatka recorded the run rate as $720,000 on 22 January 2017. Same company, six weeks apart.

Walker was selling his engineers. Astronomer billed a flat monthly fee for access to its team and its half-built technology — “data engineering as a service,” as he called it — and then grew each account from the inside. None of that first $600K came from a self-serve funnel. Almost none of it came from paid acquisition either: the company’s first-ever marketing budget, switched on weeks before this interview, was $600.

The pivot that started the clock

The company Walker describes did not begin as a data-pipeline business. It began as a product-analytics company — the same category as Mixpanel and Amplitude — and it died at a conference booth. “We were actually at the Collision conference in Las Vegas,” Walker said, “and decided to pivot it after the first day of the conference.” The reason was blunter than most pivot stories: “We just realized we didn’t quite have the passion for the problem.”

What replaced it came straight out of the old business’s failure mode. Astronomer could close analytics deals; it could not get the data. “We could get customers to say yes, we could get them to give us money, and it took us usually like four to six weeks of nagging to get the data,” Walker said. “So we basically decided, let’s help these companies get data to services like the one we had. And that’s a bigger, more global problem.” The onboarding step that killed the first company became the second company. Walker dates Astronomer from that day in May 2015 — not from incorporation. “That’s the day of the pivot,” he said.

A second pivot followed at AngelPad, the accelerator Astronomer went through. The plan going in was bottom-up: freemium, maybe a hundred dollars a month from small companies. “He kind of like knocked our plan off the table and said go after big companies,” Walker said of the man running the program. The insight that made it work was that enterprises had the identical problem in different clothing — not SaaS silos but SAP and Oracle. AngelPad’s own terms, which Walker disclosed on tape: 7% common, plus roughly $50,000 on a safe note at a $4M cap.

Services priced like software

The pricing was two numbers, and Walker gave both without being asked twice.

We basically offer access to our team and our technology for one flat price — six grand a month if you wanted to run in our shared cloud, or 10 grand a month if you want us to put up a private instance of the platform.

Ry Walker, co-founder and CEO, Astronomer

List price and realised price were not the same thing. Twenty customers against a $720K run rate works out to roughly $3,000 each per month — half the shared-cloud rate — and Walker confirmed it directly: “Average revenue historically is around 3K a month.” That gap between the small first cheque and the standard deal is the whole strategy, and he named it.

We basically have a land-and-expand sort of a strategy. So if a customer’s like, hey, I just need this little thing done, we’ll say sure, how much money you want to give us for it — and we take that money, and then we eventually try to work them into our standard deal as their use cases expand.

Ry Walker, co-founder and CEO, Astronomer

The number that made it a venture business

Astronomer had lost exactly one customer, ever. What it had gained inside the others is the figure Walker led with when Latka asked about churn: net negative churn of 48%. Latka made him show the arithmetic, and the definition Walker gave is the standard one — a third of current revenue coming from expansion inside contracts the company had already signed.

148%of original contract value, from the same customers — January 2017

Latka restated it as a bucket: take all of Walker’s original customers, put them in one, and that bucket grows even after the leaks from the original signups. “That’s exactly right,” Walker said — and then pointed out that new logos were growth stacked on top of that, with a backlog of deals that had slipped past December. He expected to “probably double our revenue in the next six or eight weeks.”

For a company doing $60,000 a month, that expansion rate was the entire argument to an investor. It is why a services-shaped business could raise venture money at all: the revenue behaved like software revenue — sticky, compounding inside the account — even though the delivery mechanism was people. The mechanics behind it are revenue churn and its negative-churn variant; the convention Walker used to state the $600K is annual run rate.

A $600 marketing budget and a nine-month payback

Astronomer spent nothing on customer acquisition for the whole of its first revenue year. “Everything’s been referral and outbound up till now,” Walker said. The first paid dollar went out in January 2017, the month of this interview, and Latka pushed him on how a founder should size that first budget. “Start small,” Walker said. “Like, we’re starting very small.” Pressed for the actual number: “I think our first month’s budget was 600 bucks” — just enough, he said, “to get retargeting up and running for us.”

With no paid history there was no clean CAC, so Walker reported the number he did have: a fully weighted payback period with sales and marketing salaries counted in, not just ad spend. He walked through how it had moved.

• Started around 12 months — the first fully weighted payback period Walker modelled, labour included.
• Down to about six months — the efficiency peak, before the team grew.
• Back up to 12 — they doubled the growth team, which “threw more cost in there.”
• Back down to nine — where it sat on the tape, and “creeping back down again.”

Latka did the multiplication live: nine months of payback on a $3,000-a-month customer implies an acquisition cost somewhere around $27,000. Walker’s reply is the reason this tape is worth quoting at all. “That might be true. I don’t know, you’d have to check that.” He was not defending a CAC; he was reporting a payback period and saying out loud that it was modelled on thin data — “our math on that is based on not a ton of data.” Anyone comparing that $27,000 to their own numbers should note it is Latka’s arithmetic on air, not a company-reported figure. That distinction is exactly what muddies most CAC benchmarks.

Four notes, and a pile to burn

The $2 million behind all of this had not arrived in one round. It arrived in four. Latka’s introduction named the backers: 500 Startups, CincyTech, Router Ventures and Social Starts.

“We’ve actually done like four rounds of notes,” Walker said — a KISS, then a SAFE, then “a real KISS,” then one more, stitched together with an intercreditor agreement to keep the terms congruent across all four. The largest was $1.1 million, closed in August 2016. Latka, interviewing a founder a day at the time, flagged it as a pattern he was seeing everywhere: note after note after note, into the millions. Walker’s explanation was structural. The investors who like to price rounds had all moved up-market. “The old B is an A now,” he said, “and so a lot of the earlier-stage investors don’t like to price things.”

Then came the part most founders would not say into a microphone. Latka assumed a company that had raised recently had runway. “No, we don’t actually,” Walker said.

We decided to burn the pile as fast as we can, in a responsible way. But yeah, when you get to 25 people, the burn — I think our gross burn is around 200. So we’re burning net 150 right now. So yeah, a million doesn’t last too long when you’re doing that.

Gross burn of $200,000 a month against roughly $50,000 of monthly revenue nets out to $150,000 leaving the bank every month. On a $1.1 million round that is eight or nine months of life, which is why Walker was raising while he talked. The round he described was a “post-seed” of $3 million to $5 million, explicitly not a seed extension: “It’s essentially the old A.” On valuation he refused to posture — “whatever we can get it at, to be honest” — reasoning that share price matters more than percentage ownership, and that leaving headroom for the next round beat pricing this one high.

What the $600K became

From a $3.5M seed to $377.8M raised

He got the round. GetLatka’s funding table records a $3.5 million seed dated June 2017 — five months after this interview, inside the $3M–$5M band Walker named on tape. (The four convertible notes he describes here predate GetLatka’s record of the company; the table starts with that seed.) What came after is a ladder almost nobody would have drawn from a 25-person Cincinnati services company running on nine months of runway.

• Jun 2017 · Seed $3.5M — the post-seed round Walker was out raising on this tape.
• Jul 2017 · Seed $1.9M.
• Nov 2018 · Seed $3.5M.
• Sep 2019 · Seed $5.7M — four seed-stage rounds before a priced one.
• June 2020 · Series A $13.6M, five years after the pivot.
• Jan 2021 · Series B $43.6M.
• Mar 2022 · Series C $213M, announced alongside the acquisition of Datakin.
• May 2025 · Series D $93M, led by Bain Capital Ventures.

That totals the $377.8 million GetLatka records as Astronomer’s funding to date. Revenue followed the capital, though only the first of these figures is a reported number.

The $720,000 recorded at the January 2017 interview is Walker’s own $55K–$60K MRR, annualised. The three later points are GetLatka estimates and are flagged as estimates in the dataset: roughly $12.1M in December 2023, $24.5M in October 2024, and $39.5M as of June 2025. Headcount is not an estimate — 25 people on the tape, 322 recorded in October 2024.

The half-built platform got built. In its own May 2025 Series D announcement, Astronomer described itself as “the company behind Astro, the leading unified orchestration platform powered by Apache Airflow,” sold to more than 700 enterprises. The data-engineering-as-a-service arrangement Walker was billing $6,000 a month for was the bridge, not the destination.

One line in that release is a direct echo of this tape. Alongside “150%+ YoY Astro ARR growth,” the company claimed “world-class 130% Net Revenue Retention.” Expansion inside accounts it already had is still the engine — the same mechanism Walker described in January 2017 at 148% across twenty customers, now running two orders of magnitude further down the same track. Astronomer’s current figures sit on its GetLatka profile.

Walker was 44 at the taping, married with three children, and sleeping badly on purpose — to bed at midnight or one or two, up “as late as I possibly can each day before someone forces me to get up.” Latka closes every interview the same way: what do you wish your 20-year-old self knew? Walker had started early, so the answer was never going to be about starting.

To be bold, to be confident. I think confidence is pretty tough to do when you’re that young — but if you can figure that out early enough, it works wonders.

SourcesNathan Latka’s January 2017 interview with Ry Walker, co-founder and CEO of Astronomer, on the Latka podcast; the GetLatka company record for Astronomer, its funding table and dated metrics_history rows, with estimate flags carried through to the prose; Astronomer’s press release “Astronomer Secures $93 Million Series D Funding,” dated 1 May 2025, for the Series D amount, its lead investor, the 700-enterprise count, the 150%+ ARR growth and the 130% net revenue retention figures; PR Newswire, March 2022, for the $213M Series C and the Datakin acquisition.