Atlassian reported fiscal first-quarter 2027 results on November 5 and cleared every bar it was measured against: revenue of $1.772 billion, up 24%, cloud revenue up 31% and a 32% non-GAAP operating margin. Then it described the rest of the year, and a stock that had nearly doubled since August gave back a large part of the run. The quarter was not the problem. The slope was, and underneath the slope sits a question every seat-priced developer tool now has to answer.
A clean beat, line by line
Revenue landed against a consensus of $1.71 billion and adjusted EPS of $1.65 against $1.43, per Investing.com's tally, which expected a 28.6% adjusted operating margin. The earnings release fills in the rest: subscription revenue of $1.698 billion, Data Center revenue of $390.7 million (about 5% growth, the slowest line), and subscription ARR of $6.812 billion, up 23%. SiliconANGLE puts ARR growth at 22.5%. Remaining performance obligations rose 44% to $4.746 billion. GAAP operating income was $188.8 million, an 11% margin, against a 7% operating loss a year earlier; free cash flow was $462.6 million, a 26% margin.
The call added colour: cRPO grew 28% and gross margin reached 85.2%, up three points, even as agentic automation on the platform rose ninefold.
Why a raised outlook still sank the stock
Here is the irony: the full-year guide went up. FY27 revenue growth of about 13.75% is three-quarters of a point above the August forecast, cloud growth moved to about 26% from 25.5%, and the non-GAAP operating margin target rose to 25.5% from 25%. What the market heard was deceleration. The release guides Q2 revenue to $1.825 to $1.835 billion, cloud growth to about 27%, Data Center to a 14% decline, and subscription ARR growth to about 18.5% for the year. If cloud grows 31% in Q1 and about 27% in Q2, a 26% year leaves the second half running slower than both.
Mizuho's Gregg Moskowitz said on the call that the implied four-point cloud slowdown was larger than guidance had suggested 90 days earlier. Truist cut its price target to $200 from $250, writing that "Strong execution in F1Q27 was offset by a cautious outlook."
The setup did the rest. Shares traded at $106.90 when Atlassian reported in August and $204.10 on October 29, then fell from a $183.29 close on November 5 to $154.00 after hours. Friday's tape came in pieces: down 15.8% premarket, 10.9% at 9:50 a.m., and about 16.2% in later quote data on Investing.com and Finviz.
The Data Center hangover, read both ways
Much of the slowdown is mechanical, and the bulls lean on that. Data Center is being retired in March 2029, and customers pulled purchases into fiscal 2026 after the announcement, which toughens this year's comparisons. Q2 also begins to lap the DX acquisition, and hybrid deals split between Data Center and cloud moved some revenue out of the cloud line. CFO James Chuong called the guide a "thoughtful and prudent approach" amid macro uncertainty, and Atlassian beat its own Q1 cloud guide of 28.5% by two and a half points. Migrations added mid to high single digits to cloud growth, and the company says migrated customers spend 1.75 times more three years on.
The bear reading needs fewer words. ARR growth falling from 23% to 18.5% is a run-rate signal, and Truist, while calling collections, pricing and seat expansion healthy, still reads the guide as moderation for the rest of the year. TIKR's Michael Douglass states the hinge plainly: if cloud slides further after Q2, the market was right to worry.
Agents, seats, and the meter
The theory that has hung over Atlassian all year, as diginomica summarised it in February, is that AI-driven developer productivity shrinks seat counts, tooling budgets and vendor revenue. Mike Cannon-Brookes's rebuttal then: customers using AI coding tools expand Jira seats 5% faster, because more software means "more management, more overhead, more collaboration." He also argued customers prefer predictable seat pricing.
On seats, this quarter still points his way. Customers kept adding Jira and Confluence seats, including business teams outside engineering. Rovo is used by 87% of enterprise cloud customers, and more than 10 million human-agent collaborations happen on the platform monthly, double six months ago. The Teamwork Collection bundle, which carries 10 times the Rovo credits of standalone tiers, grew sevenfold to more than 3,000 customers.
The product direction, though, says where pricing is heading. The release describes agents that can be @mentioned and assigned Jira work items, with distinct identities, scoped authority and audit trails under a new Agentic Multiplayer Protocol. A rebuilt MCP server serves over 2 million monthly users and 15 million daily tool calls, and an expanded OpenAI partnership connects ChatGPT and Codex to Atlassian's Teamwork Graph. Usage-based pricing enforcement starts in early December, though Chuong said it is not a near-term revenue driver and is not meaningfully in FY27 guidance. An agent holding a Jira ticket is a worker the per-seat model was not built to bill; Atlassian is keeping the seat and installing a meter beside it.
Peers offer no verdict yet. TIKR points to Salesforce, whose Agentforce ARR passed $1.5 billion while the company guides to 11 to 12% revenue growth: agent revenue can grow fast without moving the total much.
The 30,000-foot read
Atlassian did little wrong on November 5 except arrive after a near-doubling. It is GAAP-profitable where a year ago it lost money, and much of the deceleration follows from its own decision to retire Data Center. What it cannot control is the multiple the market grants seat-priced software while nobody knows whether agents add seats or subtract them. What it can control is whether December's meter becomes a second engine or merely a second invoice.
For teams on Jira and Confluence, the near-term change is administrative, not existential. Data Center shops have a dated exit and a vendor whose plan assumes they take it. Anyone letting agents into Jira should read the usage terms before the meter starts, and use the identity and audit trail that come with it. Watch Q2 cloud against 27%, ARR against 18.5%, and the first disclosed usage revenue. The seat is not dead. It is simply no longer the only thing being counted.
Sources: Atlassian Q1 FY27 earnings release, 8-K Exhibit 99.1 (all quarterly results, Q2 and FY27 guidance, Data Center dollar figures, agent, AMP, MCP and OpenAI details); Investing.com (consensus estimates, operating margin guide raise, Truist target and quote, 9:50 a.m. move); Yahoo Finance, via Investing.com (premarket move); SiliconANGLE (22.5% ARR figure, revenue guide versus August, Data Center retirement and pull-forward, seat additions); 24/7 Wall St. (cloud guide change, Q1 cloud guide, Mizuho comment, DX and hybrid-deal effects, CFO quote, share prices); diginomica, Q1 FY27 (cRPO, gross margin, Rovo and collaboration figures, migration contribution, usage-based pricing); diginomica, February 2026 (seat-erosion thesis, Cannon-Brookes on AI and seats); TIKR (bull and bear framing, Salesforce comparison); Finviz/InvestorsHub (Friday quote data). All results are company-reported; consensus figures are analyst estimates compiled by Investing.com. The exhibit cited is Atlassian's earnings press release; the separate shareholder letter was not reviewed. Data Center Q1 growth is computed from the release's dollar figures (SiliconANGLE also reports about 5%). Call commentary (Mizuho, CFO, usage-based pricing, cRPO) comes from secondary coverage, not a transcript. CAVEAT: I could not confirm an official November 6 closing print; the roughly 16.2% figure comes from quote widgets on Investing.com and Finviz, while intraday reports ranged from 10.9% to 15.8%. The Salesforce figures are TIKR's, not checked against Salesforce filings.