When I first joined Scale in 2019, Tuomas, the co - founder of Linear—my former colleague at Uber and also the CTO of Linear—sent me an email. He first congratulated me on joining Scale, and then mentioned that no wonder recently at Uber...

In 2019, when I first joined Scale, Tuomas, the co-founder of Linear—my former colleague at Uber and also the CTO of Linear—sent me an email. First, he congratulated me on joining Scale, then mentioned that no wonder he hadn't seen me at Uber recently. Next, he introduced the new app Linear he had developed and asked if I wanted to give it a try. We soon arranged a demo, although it was some time later that Scale became a customer of Linear.

Later, during my time starting a business back in China, for a while on LinkedIn, everyone was discussing whether a company had to expand wildly to succeed. The Linear team shared their experience—that a company could focus on making good products without wild expansion and still become an excellent company. At that time, Ganesh, our common leader at Uber (who later switched to VC), commented that Linear had found a great development path, but it was difficult to achieve such an efficient growth method in a highly competitive and winner-takes-all market.

As it turns out, the Linear team's expansion has indeed been very restrained, but its business growth performance is quite remarkable. I've been using Linear and also recommend it to the companies I advise. Compared with traditional tools like Jira, Asana, and Trello, Linear is much more user-friendly. It's evident that this is a company that takes product development, engineering, and user feedback seriously, creating better products through efficiency rather than resource input.

Last week, Linear announced a new round of financing. Before that, they had only raised funds twice. This time, Accel, which has achieved a lot in the AI field, led the investment. After the financing, Linear became a unicorn. Looking at the latest various tools—AI Coding tools such as Devin, Claude Code, etc.—you'll find that usually, only one or two task management tools are officially supported in their documentation: if there's only one, it's Linear; if there are two, they are Linear and Jira.

When Linear first started in 2019, there were no large language models yet, and it was hard to imagine that AI Agents would create a disruptive opportunity for task management systems. Analyzed according to many current methodologies, for a startup with just a few people to challenge a multi-billion-dollar giant like Jira in the task management field, the conclusion would surely be that it's impossible to succeed. But today, we've seen the possibility of Linear surpassing Jira, and this possibility stems from the pursuit of product efficiency, a good taste for engineering design, restraint in expansion, and the ultimate improvement of engineering efficiency.

I think this is precisely the belief that many people in software development lack. When developing software, people either think they have to go all out quickly or think they can only make a little money. I hope to see more and more successful cases like Linear.

via AI Exploration Station - Jike Circle (author: Panda AI Sugarcane)