Cursor problems, start solving the easy stuff

Feedback
1

Cursor is an excellent product, but it seems to be at a plodding development progress at the moment. While no offense is intended, the product is beginning to exhibit signs of being managed by a young team with limited experience in growth and operational management. While the platform itself shows potential, your communication with your users and the documentation requires significant improvement.

Example “Composer” is a significant components of Cursor, but it is not mentioned or explained in the documentation. For example I just tried using the “search or ask” for Composer on the Cursor - Build Software Faster - the result was:
“You have reached your chat limit. Please try again later.” :rofl: it’s hard to take “Cursor” seriously if you don’t even update basic documentation and your search reaches the chat limit :face_with_raised_eyebrow:

A few examples of the documentation:

  1. Lack of Clarity on Reindexing and Dataset Management:

The documentation does not explain the technical foundation of reindexing. Is it powered by Retrieval-Augmented Generation (RAG) or another methodology? Understanding this would help users optimize their workflows, especially when working with large codebases. Additionally, providing concrete examples of when to use dataset indexing versus .cursorignore files would be invaluable.

  1. Best Practices for Context Optimization:

One of Cursor’s advertised strengths is providing “contextual awareness” for better coding assistance. However, there is little to no guidance on how users can structure their projects or configure Cursor to maximize this capability. Including real-world examples, data points, or even anecdotal use cases could make this section significantly more actionable.

  1. Comprehensive Feature Explanations:

While brevity is important, the current documentation leans too heavily on surface-level descriptions. For a platform that recently raised $60M in funding, and having users like my self that are paying 40USD/m it should be a reasonable to request that you put in the hours to provide detailed, and insightful technical resources that not only describe what a feature does but also why and how to use it effectively.

Some sections of the documentation read as though they were generated by a language model with limited knowledge of Cursor’s architecture and goals. This diminishes the credibility of the content when you encourage us to use Google or Stack overflow for the .gitignore typed .cursorignore files. :upside_down_face:
I think you could reduce the amount of clutter on the forum if you started to provide use case scenarios for optimal onboarding experience and clear technical insights to help your users.

Providing simple things like a roadmap would ease many users. BTW, framing your collaboration with Supermaven is not a roadmap.

Beyond technical issues, there is a more significant concern: communication. Cursor currently appears to be out of touch with its user base, which could have long-term consequences. For example, acknowledging user frustrations and outlining your roadmap for improvements would go a long way. Statements like:

“We understand Composer is a frustrating experience 50% of the time. We are sorry for the many instances where your carefully crafted code was deleted or altered without clear reason :joy:. We’re actively working to change this behavior to better serve you.”

… would demonstrate that you value your users’ feedback and are committed to improvement. This kind of transparency builds loyalty. Users will tolerate growing pains if they feel heard and understand the steps being taken to address their concerns.

In the rapidly evolving landscape of AI coding assistants, it’s impossible to stay permanently ahead of the curve. There are too many players vying for attention. However, solid communication can cultivate loyal users who will stay with Cursor even when the next shiny tool emerges. Without it, the platform risks becoming another temporary stop for lemmings and “content creators” that jump to the next trend once their favorite YouTuber makes a shocked-face thumbnail about a competing product.

10 Likes
2

Cursor is great, but it feels too much like a garage startup. With the kind of funding they have I expect a more professional experience.

4 Likes
3

well… while I 100% share @flight505’s sentiment and analysis here, they did only raised in August, if I’m not mistaken. Hiring and building out support structures also takes time and needs differnet skillsets then writing code.

I guess we can debate what is reasonable to expect here time-wise but personally I’d value good choices and signs of directionality here over fast ones. I do agree that the Supermaven merger might be cool down stream somewhen but is a lighthouse that no-one really needed right now.

Also posting this to encourage the team to do some soulsearching on how hands-on they wanna be with either the product itself or the community and be pragmatic about the other side of the coins on those decisions to not spread them selves too thin.

1 Like
4

Internet startup cash is already gone dude. Executive recruiter can have somebody the best of the best of the best the top 0.1% at the desk in less than 2 weeks.

Building a company takes time if you don’t have any money, you you still have 60 million bucks deploy it.
You just picked up Maven, what are these guys doing just sitting around doing nothing? They raised money as well.

Being guilty myself of taking all the gas off after a raise, but I was never an environment this competitive, this space is brutal. Every plugin in VS code that did AI coding it’s going to have their own IDE and make a play. Code Maven was actually one of the worst plugins, and they now have their own IDE.

I don’t think you can take gas off until you get big enough to duplicate yourself so many times that if you got to take a couple hours or half day that’s okay.

I think the team has gotten too much money too quickly. They’ve never gotten this before and they got lazy. It’s called burnout. Had it. They would have been better to have a smaller core team and build and release quickly rather than a lot of money they have to spend, you can’t really save it.

With 60 million comes a whole new problem you got to account for it and you got to report back constantly The worst thing that you can do is take outside investor money if you want to move fast like before.

You have to change your whole management mindset this is where a lot of companies fail because they need to hire managers that don’t do anything they just manage. You have to learn parliamentary procedure you have to learn how the board works you have to present you have to do everything else 90% of your job as the CEO with investors is catering to the investors 10% for the actual company.

I will be actively looking around for another IDE for sure.

1 Like
5

Hi, your critique is rather a log of your understanding journey about the product, not a review. Cheers!

P.S. Please come back with those more excellent comments.

6

:astonished: :point_right: New Tool, Cursor killer?

1 Like