Hi Cursor Team,
I want to share this message after using Cursor for almost two years and submitting multiple rounds of feedback during that period. To be direct: the product increasingly feels like it is being developed without a mature product development process, and major UX decisions seem to be made without considering the impact on daily, heavy-use users.
This isn’t based on emotion, it comes from consistent, observable product behavior over a long time.
⸻
1. Frequent, disruptive UI/UX changes (evidence-based)
Over the past two years, core UI elements have repeatedly been:
• moved, renamed, or removed without prior communication
• redesigned in ways that break established workflows
• changed in ways that were later reverted, indicating insufficient UX validation
These patterns are visible simply by reviewing Cursor’s update history, Discord discussions, and community feedback.
This level of volatility does not align with how mature product organizations operate.
⸻
2. “Use an older version” is not acceptable for a paid product
On multiple occasions, the team or moderators have responded to breaking-change complaints with:
“You can download an older version if you preferred the previous behavior.”
However:
• older versions expire
• server-side updates break them
• users are forced to manually test which versions still work
This is not a sustainable or professional fallback in a paid software product.
Stability is a product responsibility, not the user’s.
⸻
3. Extremely fast shipping cadence without visible governance
Cursor’s rapid release velocity is impressive, but the lack of structure around it is equally evident:
• major UI decisions shipped without beta channels
• no feature flags for experimental changes
• no opt-in system for new UX
• no communicated deprecation timelines
• breaking changes pushed directly to all users
These are observable patterns in the changelog, Discord, and product updates.
This suggests high speed, but without the scaffolding expected from a mature change-management process.
⸻
4. Public announcement behavior reinforces this perception
The concern is not hypothetical, it’s supported by public behavior:
• several major UI updates were promoted on X before stability issues were resolved
• big feature announcements (Agent demos, Claude Code integrations, multi-file actions) were marketed before core workflows were stable
• after these announcements, rapid hotfixes and adjustments followed
This creates an impression that visibility sometimes takes priority over stability.
This is not a criticism of ambition, but of sequencing.
⸻
5. The competitive landscape has changed
When I first started using Cursor, it had no real competitors.
Today, alternatives like Claude Code and several AI-native dev environments offer:
• stable, predictable UX
• minimal disruption
• clear change-management policies
Cursor is no longer operating in a vacuum, and high volatility is now a real competitive risk.
⸻
6. Mature companies follow principles that Cursor is currently missing
Industry leaders rely on predictable UX and structured release processes:
• Figma: “Do not surprise the user. Predictability is sacred.”
• Apple : “Evolve slowly. Maintain familiar workflows.”
• Airbnb: “Nothing ships without internal dogfooding and staged rollout.”
• Notion: All experimental features go through opt-in Labs.
• Slack: Deprecations are announced months in advance.
• VSCode: Stable channel + insiders channel; never break muscle memory.
Cursor today does not follow these principles and the difference is visible in the day-to-day experience.
⸻
7. What Cursor needs immediately
To prevent further user frustration and long-term churn:
1. A proper change-management system
2. Feature flags for experimental UX
3. An opt-in beta/canary channel
4. Predictable deprecation policies
5. A stable channel that guarantees no breaking UX changes
6. Stronger UX governance around core workflows
These are standard practices in every successful, high-usage productivity tool.
⸻
8. This comes from loyalty, but loyalty has limits
I want Cursor to succeed.
I’ve used it for nearly two years and recommended it countless times.
But if the current pattern continues — where every few weeks major UX behaviors change without communication, validation, or migration paths — you will steadily lose your most loyal users, especially now that stable alternatives exist.
The people who rely on Cursor every day are telling you exactly what hurts.
Please listen.
Best,
A loyal user