Coralogix Spent Three Years at $0, Then Hit $24M by Flipping the Log Pipeline

In September 2017, Coralogix’s board met to decide whether to close the company and hand what was left back to its investors. Three years in, the log-analytics startup had no revenue, no customers, five employees and roughly $100,000 in the bank against about $2M raised. What saved it was not a pitch. It was arithmetic: “there was so little that they decided not to,” co-founder Ariel Assaraf told Nathan Latka — the remaining cash was too small to be worth returning.

Four years later, at the October 2021 taping, Coralogix was past $24M in run-rate revenue with 2,000 paying customers, net retention above 130%, and a $55M round behind it. In between, the team did not simply work harder. It reversed the order of two verbs.

Data grows faster than revenue

Assaraf dates the modern observability market to the moment Splunk moved logging off an appliance and into software, then into SaaS, on a simple promise: ingest everything, index it, store it, and “become some sort of Google for your organization.” That worked when the data was small enough to treat that way. It stopped working when the data stopped being small.

“Data and observability tools are part of your margins… they’re part of the unit economics of an organization, and as data grows faster than revenue it becomes hard to manage, super expensive, hard to scale. Performance issues start to rise.”

That is a claim about somebody else’s P&L, and it is also Coralogix’s entire sales motion. If a customer’s observability spend rises faster than the business it monitors, the vendor is charging a growth tax. Coralogix’s answer, the product Assaraf calls Streama, is to move the analysis in front of the storage.

The second-order effect matters more than the cost line. Analyzing before storing decouples Coralogix from the store, so the customer keeps their own front end — Kibana, Grafana, a SQL client, Tableau. “So now it’s a data platform play,” Assaraf said, “it’s not just a product.” It also let one product line become four: logs, metrics, security and, next, tracing.

Three and a half lost years

None of that existed at the start. Coralogix was founded at the end of 2014 — Latka’s intro said 2015, and Assaraf corrected him on tape — and the first version was, by his own description, “a lighter version of Splunk… a cloud cheaper Splunk.” It sold nothing for three years. The first dollar arrived in October 2017, a month after the board meeting that nearly ended it.

The money behind those years was smaller than the record suggested. Latka had a $5.2M seed from 2016 on file; Assaraf rejected it flatly: “it was a million, then another million.” The GetLatka dataset agrees — two $1M rounds, one recorded August 2014 and one November 2016.

By late 2017 the company had five people, the CEO had left, and the CTO followed him out. Assaraf, then a co-founder rather than the chief executive, brought in Yoni — a friend who had run a group alongside him at Verint — as CTO and co-founder, and rebuilt the company around the new pipeline. Within four or five months they were at $20,000 to $25,000 in monthly recurring revenue. The board, which had been preparing to write the thing off, put in another $2.5M as a convertible loan at a valuation of roughly $10M.

A dating discrepancy worth flagging: the GetLatka dataset files that $2.5M at a $10M valuation as a “Series C” dated June 2017. On tape it is unambiguously a convertible note that followed the September 2017 board meeting and funded the work that started “beginning of 2018,” and it converted later inside the 2019 round. Latka’s own closing summary put total pre-revenue funding at $4.5M, which only reconciles if you count that bridge.

“This company basically is a company that started in January 2018… we had three and a half lost years, and then in January 2018 the company started. That’s just a different life. You see the graphs of growth — it’s like this, and then just boom.”

Ariel Assaraf, co-founder and CEO, Coralogix

2018 ended a bit over $1M.

The offer that got to a signature page

At the beginning of 2019, a buyer came for Coralogix. Not an approach, not a term sheet — the two sides got all the way to a share purchase agreement, the definitive document. The price was in the tens of millions: “like 30, 40,” Assaraf said. “Not 90.” He was 29, the company was seven or eight people, and those people were tired.

“When I thought about starting the company, for me that was like an ultimate outcome,” he said of the payout. Then he and Yoni decided they did not want it: “we have a lot more to do.”

Killing the deal was harder than declining it, because there was no money in the bank to decline it with. One board member, who sat on the committee running the sale, resigned from that committee so he could back the founders. Then Assaraf WhatsApped the numbers — about $1.4M a year, mid-2019 — to Aaron at Aleph, an Israeli VC, with the board meeting set for the following Monday. Thirty minutes on a bad line into a room of Aleph partners in New York produced a term sheet, also over WhatsApp, at the same valuation as the acquisition offer, so the board could compare the two like for like.

“Tell the board that we are 100% behind it. We’re not backing off no matter what we discover in the due diligence.”

Assaraf’s verdict on that message: “which was very ballsy.” The round signed in the summer of 2019 — $10M, recorded in the GetLatka dataset in November 2019 at a $40M valuation. On $1.4M of annual run rate, Latka noted, that was a steep multiple for a company that had been near death eighteen months earlier. Assaraf shrugged: “back then it was a nice leap of faith.”

Then the graph turned. December 2019 closed at $2.8M — double, in four or five months. (The dataset’s November 2019 revenue row reads $3.5M; the $2.8M is the figure Assaraf confirmed on tape for December.) One employee thought enough of the trade to put $100,000 of his own money into that round.

COVID froze it at $3 million

Everything until then had been inbound and word of mouth — no salespeople at all. Coralogix hired its first sales team in New York and Assaraf flew out to onboard them on 10 March 2020. He went home, and the world shut. “No one would take our calls, no one answered emails, everyone were shocked. New York became hell.” For four or five months the company sat at $3M and did not move.

What he did with that dead quarter is the tell. In June 2020, with the numbers visibly stalled, he went to two Israeli growth funds — Red Dot and O.G. Tech — and pitched not the traction but the pipeline inversion: “I know the numbers don’t look really well, but we’re going to launch Streama.” They wrote a $25M check. The dataset records the round at a $100M valuation dated September 2020; Assaraf placed the close at “roughly November.” By the end of December 2020, revenue was $8M.

He also declined to optimize the round the obvious way. “One thing that we’ve done in both that round and this one after was not go after the highest valuation, even when we got it, but to optimize to the structure of round.” Structure meant who put in how much, what the board looked like, how many board meetings, which advisers came, and how much option pool stayed clean — 13% after the Series A, 8% after the Series B, none of it taken by the founders. The reason is scar tissue, and it is the sharpest line in the interview about who you take money from:

“If our seed round was with like an American top VC, we would have shut down for sure. They’d shut us down. They wouldn’t even ask.”

Ariel Assaraf, co-founder and CEO, Coralogix

The funding ladder

Seven rounds, $236.5M, and a secondary nobody took

• Aug 2014 · Seed $1M.
• Nov 2016 · Seed $1M — “a million, then another million,” not the $5.2M Latka had on file.
• 2018 · Bridge $2.5M convertible note at roughly $10M, after the board had considered closing the company. Filed in the dataset as a June 2017 “Series C.”
• Nov 2019 · Series A $10M at $40M, signed summer 2019 at $1.4M of revenue.
• Sep 2020 · Series B $25M at $100M — “roughly November,” per Assaraf, sold on Streama rather than on growth.
• Jul 2021 · Series C $55M at $400M. Coralogix disclosed no valuation; TechCrunch guessed one, and Assaraf would say only that it was “roughly around the number.”
• June 2022 · Series D $142M at $1B, after the tape.

The creative part of the Series C was not in the round. Coralogix ran a secondary on top of the $55M, outside the round — $5M to $10M, a range Latka put to him and Assaraf did not dispute — to clean early investors off the cap table — people who had assumed the money was gone and “suddenly found out they’re at 20, 25x.” Many sold. Employees were offered the same window. Not one took it.

What 2,000 customers actually pay

Coralogix was fully SaaS at the interview with a single exception: one semi-on-premise client whose data volume “is not something that any cloud can handle.” Beneath the headline customer count sits a very long tail — Assaraf described “so many zero-touch customers” that an average contract value across all 2,000 would be meaningless. The revenue is concentrated at the top:

• Top 100 accounts — roughly $80,000–$85,000 annual contract value.
• 25 accounts — over $100,000 a year.
• 5 accounts — over $500,000 a year.
• 1 account — over $1M a year, the largest customer on the book.

Those accounts get bigger for two reasons, and only one of them is a sales motion. The first is that the customer’s data grows on its own; even with Coralogix actively helping them cut cost by prioritizing data per use case, volume compounds. The second is multi-product: a customer starts on logs, adds metrics, adds security, and will add tracing. “I think this is a game that Datadog played really well,” Assaraf said. “They have nine products now and they just upsell and cross-sell all the time.”

The result is a revenue churn profile most SaaS companies would take without reading the rest of the deal:

130%+ net retention = 100% base + 32% expansion − under 2% gross churnAssaraf’s own breakdown, year to date at the October 2021 interview.

The newest line was security, and it came with something most investors flinch at: bundled consulting. Not billable hours, and not sold outside the Coralogix ecosystem — a customer buying a given data volume gets an expert to help onboard, define alerting criteria and build an incident response plan. Assaraf’s justification came from an adviser, a former ServiceNow security chief, who had polled LinkedIn on how much security spend had risen in two years (answers: 4x to 10x) and how much safer the respondents felt (answers: barely). The gap he is selling into is a staffing one. “Everyone talk about DevSecOps, but there aren’t really many DevSecOps engineers — many of them are just DevOps engineers that got a new responsibility to secure the cloud.” The security product had five customers at the taping, two of them among the largest companies in the world.

$240,000 per employee, and why that number was so high

$240,000revenue per employee at $24M and 100 people — “almost triple” the average private VC-backed SaaS company, by Latka’s on-air math

Coralogix had just crossed 100 people, about 70 of them engineers. Two and a half years earlier it had eight. Assaraf refused to take credit for the efficiency, attributing it to conditioning rather than discipline: “you tend to not be a huge spender when you experience four years having zero budget and almost shutting down a company.” He expected to cross 200 people within a year, and was far too conservative: the dataset records 199 by 1 January 2022, roughly ten weeks after the interview.

Equity stayed unusually wide. Every hire — “from office admin to anyone in the company” — gets options, and employees collectively held a little under the 15–20% Latka guessed at.

Where the numbers went after the tape

The Series D closed seven months later at a $1B valuation on $36M of revenue, recorded June 2022. After that the GetLatka dataset switches to estimates, and they should be read as estimates: about $62.9M by November 2023 and about $90.8M by October 2024, against $236.5M raised in total. Headcount is the firmer series — 240 at the end of 2022, 362 by October 2024, and 572 as of November 2025, from the eight people who were left when the CEO walked out. Current figures live on the Coralogix profile.

The billion-dollar hypothetical

Latka closed by putting the 2019 decision back on the table at scale: if Snowflake or Salesforce arrived tomorrow with a billion dollars, all cash, does he sell? Assaraf gave the answer of someone who has already been through a signature page once. “This is a question to the board. It’s not my company.” He had already said the operative part earlier — a decent runway, no urge to sell into the valuations of the day, and the observability space about to get crowded by the people who already own the data: Snowflake, Salesforce, Confluent.

The last question of every Latka interview is what the founder wishes they had known at 20. Assaraf’s answer is the same trade he made at 29, seen from the other side of it.

“You’d imagine at 20 years old that if you get to a certain point you can lay back and chill and have more time — friends, family, relax. It’s actually the opposite. The higher you get, the harder they work… I always thought that there’s gonna be a race, and everyone says I’m gonna be a millionaire by 30 and basically retire — and when I had that chance at 29, I gave that up. And I know that now I work much, much, much harder than I did.”

Sources Nathan Latka’s October 2021 interview with Ariel Assaraf; GetLatka dataset rows on Coralogix funding, revenue and headcount through November 2025, with revenue after June 2022 flagged as estimated.