Linear Hits $100M ARR Without SDRs: Agents Now Create Over 50% of Issues, Up From 3%
Linear, the issue tracker for developers, has passed $100 million in annual recurring revenue (ARR) and has never hired an SDR, the sales reps who screen prospects by phone. A year ago, AI agents created 3% of issues in the product. Now they create more than 50%, filing and sorting tickets on their own. Linear's COO argues that if an agent works no better inside your SaaS than outside it, people will simply never open your app.
AI-processed from Latent Space (swyx); edited by Hamidun News
Linear is an issue tracker popular with IT teams, where developers file bugs and tasks, assign them to colleagues and close them. The company has now passed $100 million in ARR (annual revenue from subscriptions) without ever hiring an SDR, the sales reps who call prospects to find out whether they have a budget. The biggest change is in the product itself: a year ago, AI agents created 3% of all work in Linear, and now they create more than half.
For example, an agent that finds a bug in the code files the issue itself, adds a label and assigns a team, so a human doesn't have to do it by hand. Linear COO Cristina Cordova described this on the Latent Space podcast. The lesson is useful to anyone building SaaS products: it shows how to avoid becoming a "data warehouse" for other people's agents, and how to sell without an army of sales reps.
Half of issues are created by agents
Linear released its MCP server in May 2025. MCP is a standard way to connect an AI agent to a service so that it can read and change data. According to Cordova, Linear was practically the first in its category to do this. The company then added a platform for embedding agents, integrations with Codex, Cursor and GitHub Copilot, and its own Linear Agent.
Cordova considers it a serious mistake for a SaaS company to settle for being a database with an MCP server on top. If the agent workflows inside your product are no better than the ones outside it, people have no reason to open your app. So she tracks two lines: how often agents use the product, and how often people open the app itself. If the first grows while the second stays flat, the product becomes an interchangeable service. At Linear, both are growing: people come back to the app to see and coordinate the work.
CEO Karri Saarinen's line "Issue tracking is dead," Cordova says, does not mean the issue disappears as a unit of work. It means the end of manual labor: agents, not people, should now file, label, assign and close issues. Linear is small enough to afford to "eat" its own business. Google has it harder: if it put an AI answer above its search results, fewer people would look at ads.
Sales without SDRs or forward deployed engineers
When Cordova joined in 2023, Linear had about 40 people and one salesperson. Today the company has about 170 employees, and the only executives are the founders and her. Besides $100 million in ARR, Linear has more than 40 000 paying customers and net revenue retention (NRR) of 177%: that is, customers who were already paying now spend 1.77 times more in total than before. In August, employees sold $99 million worth of shares at a company valuation of $2.5 billion.
Linear has never had SDRs. Cordova is convinced that a qualifying call wastes the customer's time: they want to see a demo, not answer questions about their budget. So account managers run outbound contact themselves, using personalization tools. Linear also has no forward deployed engineers, the engineers who go to a customer's site and deploy the product inside its systems. Linear should connect so easily that nobody needs access to your infrastructure. On large deals, solutions engineers work alongside the sales reps. Sometimes they build a working ServiceNow integration overnight to show that it's possible.
The company keeps support in-house and does not let an LLM answer tickets on its own: the AI drafts a reply and a human decides. Customers sometimes think the AI is answering because replies arrive so quickly. In three and a half years, Cordova says, there has not been a single escalation over poor support.
From Stripe she brought the 40% rule: at least 40% of employees should be engineers, product managers, designers and analysts. If sales asks for more people, it first helps hire engineers. By her estimate, Linear runs on about half the headcount typical for companies with this level of revenue.
AI budgets and pricing
What sells fastest right now is whatever falls under a separate budget for AI tools: that money doesn't have to be fought for against other line items. But Cordova warns that buyers increasingly sign one-year contracts and paid pilots instead of three-year deals, because they don't know what they will need in two years. Some of the "$100 million ARR in a few months" stories are really a handful of $30–40 million contracts with defense companies or large law firms. That is a completely different business from Linear's 40 000 customers. When renewals come, it will become clear whose ARR was real.
Until June, Linear charged only per seat. Cordova would like a model like Intercom Fin's ($0.99 per resolved conversation), but it is hard for Linear to price what it is worth to automatically route a bug to the right team: at one company that is a product manager's job, at another it isn't.
So the company charges where it has direct costs. Seats remain, and the Linear Agent chat is free even though it runs on frontier models. In coding sessions, the user picks OpenAI or Anthropic models, and Linear charges the published API price plus a small markup.
Loops, recurring tasks for agents, are paid for with AI credits. New features launch as a free beta with a warning about the future price, and the price starts high and then comes down.
The practical takeaway for founders: measure agent activity and human activity in your product separately, keep the share of engineers high, and charge for AI where it costs you money. A caveat: this is an executive's account of her own company, the figures have not been independently verified, and the podcast transcript is automatic and may contain inaccuracies.
- Nov 19, 2026MLPerf Training: 10,240-GPU clusters and 49% faster training in six months
- Nov 19, 2026Amazon and Google will hold 28.4% of Anthropic after its IPO, but without voting rights
- Nov 19, 2026AI token prices fell 21.2% in six weeks, while corporate API spending dropped 6.3%
- Nov 19, 2026Gemini API Can Now Load Tool Schemas on Demand Instead of in Every Prompt
Key Questions Answered
What are the key details on Half of issues are created by agents?
Linear released its MCP server in May 2025. MCP is a standard way to connect an AI agent to a service so that it can read and change data. According to Cordova, Linear was practically the first in its category to do this. The company then added a platform for embedding agents, integrations with Codex, Cursor and GitHub Copilot, and its own Linear Agent. Cordova considers it a serious mistake for a SaaS company to settle for being a database with an MCP server on top.
What are the key details on Sales without SDRs or forward deployed engineers?
When Cordova joined in 2023, Linear had about 40 people and one salesperson. Today the company has about 170 employees, and the only executives are the founders and her. Besides $100 million in ARR, Linear has more than 40 000 paying customers and net revenue retention (NRR) of 177%: that is, customers who were already paying now spend 1.77 times more in total than before. In August, employees sold $99 million worth of shares at a company valuation of $2.5 billion. Linear has never had SDRs.
What are the key details on AI budgets and pricing?
What sells fastest right now is whatever falls under a separate budget for AI tools: that money doesn't have to be fought for against other line items. But Cordova warns that buyers increasingly sign one-year contracts and paid pilots instead of three-year deals, because they don't know what they will need in two years. Some of the "$100 million ARR in a few months" stories are really a handful of $30–40 million contracts with defense companies or large law firms.
Ship AI Faster with Founders, Engineers & Weekly Architecture Reviews
Inside Vibe Club (hamidun.com/club) — weekly live masterminds with Zhemal Khamidun, Tech Radar, full LMS access, and 370+ production Claude Code skills.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.