Will AI Redefine SaaS? Linear CEO’s Take on the Future of Software

When the claim "SaaS is dead" circulates, Linear co‑founder and CEO Karri Saarinen offers a nuanced counterpoint: AI will not kill SaaS, but it will fundamentally reshape its value proposition. The competition shifts from a checklist of features to the ability to capture organizational context, support critical decisions, and intelligently orchestrate AI agents.

AI as an Amplifier, Not a New Story

Linear’s mission has always been more than a simple ticket‑tracking tool; it aims to reduce friction in teams and let product, design, and engineering focus on higher‑impact work. AI should therefore be inserted where it amplifies this existing mission, not merely added as a chat window or a superficial "we also have agents" claim.

From Tools to Judgment

The next generation of software will provide judgment rather than just tools. Teams will act as "judges"—defining problems, extracting context, and delivering high‑quality decisions. This requires a control layer built on "context" that translates human intent, historical decisions, and collaboration relationships into machine‑executable instructions.

Workflow Design Beats Model Power

Linear quickly dismissed the "add a chatbot then plug a model" approach because the real bottleneck is unclear workflow design. The key questions for integrating AI agents are:

When should AI help a person understand a problem?

Which context should be automatically injected?

Should the output be a suggestion, an issue, a code draft, or a reviewed result?

When does a human need to take over?

Answering these determines whether AI becomes a gimmick or genuine productivity.

Context, Priority, and Decision Records as the Real Moat

In real organizations, the hardest thing to copy is not a button or a form but the accumulated context, priority, decision history, and collaboration relationships behind those interfaces. Linear aims to become the control layer that provides agents with goals, constraints, and organizational memory.

Measuring AI Impact Beyond Tokens

Industry metrics that flaunt token counts, PR numbers, or code generation volume are misleading. Linear focuses on a quality‑closed loop: zero‑bug tendency, one‑week SLA for bug resolution, and ensuring AI tackles high‑frequency, clear, verifiable quality work. The three concrete success indicators are:

Reduced time from problem discovery to resolution.

Actual decline in bugs and regressions.

Improved user feedback after new feature releases, not just more feedback.

AI should accelerate execution, but judgment must remain deliberate. Linear practices "slow design, fast execution": spend time defining problems and evaluating solutions, then move quickly to implement.

Future Product Teams as Judges

In the coming years, product teams will resemble judges rather than assembly‑line workers. AI agents will handle repetitive execution, but humans must still articulate why a decision is made, set standards, and steer long‑term direction. The three most scarce capabilities will be:

Defining the problem.

Extracting and structuring context.

Making high‑quality judgments.

Organizations that embed these abilities into their workflows will master agents instead of being driven by them.

Key Takeaways

SaaS won’t disappear; it will shift from selling features to selling context and decision systems.

AI’s value lies in removing friction, not just adding another entry point.

The most valuable skill will be problem definition, context synthesis, and high‑quality judgment.