Loom Case Study: From a Failed User-Testing Tool to a $975M Atlassian Exit

Loom exists because two other products died first. Shahed Khan and his co-founders — sharing a San Mateo apartment, one of them sleeping in Khan’s room — built a user-testing tool called OpenTest, found no market, pivoted, found no market again, and were running out of cash when they noticed something odd in the wreckage: people were using their screen-recording Chrome extension to send work updates to colleagues, a use case they’d never intended. “We had to come up with some hail mary,” Khan told Nathan — so they followed the misuse. Loom the product launched in June 2016.

By the February 2019 interview — recorded the very week Loom put up its first paywall — this is where the hail mary stood:

Revenue to that point: essentially nothing, on purpose.

Engineering the loop instead of buying growth

The growth was a designed system, and Khan itemized it:

• Every shared video was an advertisement with a recipient — “If you send someone a Loom video, they’ll watch it, they’re curious, and they’ll sign up.”
• Post-watch prompts — pushed viewers to reply with a Loom of their own.
• A referral program — paid $5 in credits per invited co-worker back when the base was small.
• A dedicated growth engineer — shipped one experiment per week: ship the winners, drop the rest.
• Even Gmail became a surface — compose a video where you’d have typed a paragraph, thumbnail embedded for everyone else.

The underlying thesis was Khan’s tidiest line:

Of 1.1M registered users, about 200,000 were monthly active — an 18% ratio Nathan called out as genuinely strong for a free product.

The $60K/month leap of faith

Loom Pro launched at $10 per user per month, and Khan — with unusual candor — shared the launch model:

3–4% of active users convert → about 6,000 payers → roughly $60K of MRRBy Khan’s own description, “more on the conservative side.”

The financing had bought the patience: ~$14M raised across pre-seed, seed, and a Series A led by Kleiner Perkins’ Ilya Fushman, announced the same week. Sixteen people, remote-first (devops in Omaha), burning low-six-figures monthly with eight months of runway when the A came together — a raise that took two weeks of meetings because the relationships predated the need. Khan was 23.

Nathan’s on-air valuation math — $11M for maybe 20% implies $40–50M pre — drew no correction, just the argument that cohort engagement depth was what let a pre-revenue product raise on B2B SaaS terms rather than consumer-app terms.

What the dataset recorded next

Year	Revenue	Team	Event
2019 (Feb)	~$720K	16	Paywall launches; $11M Series A
2020	$5.9M	94	$28M Series B at $350M
2021	$35M	222	$130M Series C at $1.53B
2022	$42.5M	266	—
2023	$50M	271	Atlassian acquires Loom for $975M

The pandemic did for Loom what it did for every remote-work bet — compressed years of adoption into quarters: revenue 8x’d in 2020 and 6x’d again in 2021, and the valuation sprinted to $1.53B. Then the honest rows: growth cooled hard in 2022–23 ($35M → $50M over two years) as offices reopened and the 2021 price became unearnable.

$975MAtlassian’s October 2023 price — roughly 20x revenue, and a 36% markdown from the peak round

The acquisition was simultaneously a landmark exit — a life-changing outcome engineered from a dead user-testing tool — and a repricing, the same arc Gong’s valuation history shows without the exit. Both things are true, and founders should hold both.

Current data lives on Loom’s GetLatka profile; the full February 2019 conversation is here. Khan’s answer on what he wished he’d known at 20 — recorded at 23, two failed products behind him and a $975M exit invisibly ahead — holds up: “Being a founder is very, very tough. You only understand how tough once you jump into the deep end.”

SourcesNathan’s February 2019 interview with Shahed Khan; the GetLatka dataset’s Loom rows.