I like how Claude helps with brainstorming, resuming data, create documentation and automating tasks. I love that I can plug it with other platforms like Jira to create tickets, Notion for documentation creation, and Mixpanel for dashboard construction. Claude is really helpful, especially when compared to other AI tools. Therefore, it is really easy to use and setup. Review collected by and hosted on G2.com.

Sometimes, Claude can hallucinate. It's not always perfect, and I need to check after. Also pricing can be a bit expensive for a company. Review collected by and hosted on G2.com.

I use Claude for everything, as it's like my right-hand man. It's my sounding board, helping me brainstorm and work through ideas. It executes mundane tasks, allowing me to stay creative and focused on the bigger picture. I love Claude for giving me the freedom and flexibility to do the things I love, making me more productive and helping me get more time back. The Claude Co-work feature is fantastic because it lets me set up multiple tasks with ease, requiring less oversight. The browser MCP is incredible as it takes over my browser to handle tasks for me, almost like an employee. I also appreciate Claude for its user interface, features, and writing style compared to ChatGPT. The initial setup was super easy. Review collected by and hosted on G2.com.

Sometimes I find it jumps the gun and gives a confident answer without fully checking to make sure it is correct and makes assumptions. Also, sometimes the writing style is repetitive and sounds like AI writing patterns. Review collected by and hosted on G2.com.

The Opus models and their usage limits were recently adjusted, and they now feel sufficient. The recent changes that add more usage to the plan also make it more cost-effective.

That said, Opus is still quite slow compared to models like GEmini 3.5 Flash, or the Cursor models.

For coding work, Claude is best for architecture and bug fixes, especially because it hallucinates less in its max thinking mode.

MCPs support is quite good as well, with support for GitHub, Google, and Notion, etc.

I don’t like the Claude UI much, mainly because it isn’t as easy to set the thinking mode as it is in Cursor or Copilot, where it’s simpler. Review collected by and hosted on G2.com.

The SKILL setup on claude app is not as easy unlike the copilot which offers custom agent with custom tool calls etc making it easy to select which agent to use. Review collected by and hosted on G2.com.

Claude’s writing quality is what stands out most to me. The output is clear, well-structured, and easy to adapt to a specific tone—friendly, direct, executive-ready, and so on. It’s also excellent at turning rough notes into something genuinely actionable, whether that’s summaries, emails, enablement docs, call follow-ups, or FAQs. I’ve found it especially strong at reasoning through nuanced scenarios and delivering responses that feel thoughtful rather than templated. Review collected by and hosted on G2.com.

Like any AI, it can be confidently wrong if you don’t give it enough context, so I still feel the need to sanity-check the facts. I’d also appreciate more control and visibility into how it prioritizes instructions when prompts start getting complex. Review collected by and hosted on G2.com.

I love using Claude to automate my recurring tasks each day. It helps me keep track of new product developments and changes, pushing updates to Slack and Asana. It's essentially my personal assistant and saves me tons of time with its integrations and connections. I love how Claude gives me the answers I need fast and does the work for me. Its connections are great, especially with Asana, and it handles many manual, time-consuming tasks like creating templates and assigning tasks. Compared to Chat GPT, Claude works better and has more connectors. Review collected by and hosted on G2.com.

Claude needs explicit direction or else it tends to lead me down the wrong path. In order to get the best results, Claude needs proper prompts and all of the resources required. Review collected by and hosted on G2.com.

Claude is very efficient, if not one of the best or probably the best AI that could code very well, it is available query basis/token and provided we give a good prompt, it connects frontend and backend so smoothly, that one can build a software themselves.

the pricing is legit great compared to other AI like gemini or chatGPT as its paid version will give access to their opus model which is vastly better. their ui is very simple for new users as well, but some integrations like skills can be difficult to use in first try. Review collected by and hosted on G2.com.

as of support, as an user who have even hosted a website using claude, i have so far never needed to contact customer support, so i would not have a say on that. the performance of the models is superb but we have limited queries. most of the time tokens are completed before our code is completed by sonnet 4.6. Review collected by and hosted on G2.com.

Claude has become an indispensable tool for my engineering projects. It excels at breaking down complex technical problems, writing and reviewing code, explaining architecture decisions, and helping debug tricky issues. The depth of its technical knowledge is impressive — whether I'm working on system design, algorithms, or infrastructure, it consistently gives thoughtful, accurate responses. It feels less like using a tool and more like having a knowledgeable engineering colleague available 24/7. I use it daily and it has genuinely made me more productive. Review collected by and hosted on G2.com.

The token usage is a little hard to manage Review collected by and hosted on G2.com.

The cross chat memory is the best feature of Claude. I work on a number of apps that are interrelated and Claude has a solid grasp on all of them. I can reference one and it has instant access to the relevant parts. Review collected by and hosted on G2.com.

There is no method/methodology to keep the context of a conversation that has length, current. The workaround is breaking things like builds into parts but, unfortunately, that's a bit like 10 steps forward, 1 step back. Review collected by and hosted on G2.com.

I use Claude daily for a variety of tasks in marketing like writing copy, drafting campaign briefs, rephrasing content, and more. It really cuts down the time I spend staring at a blank page and helps reduce back-and-forth in campaign creation. The quality of content and design is noticeably better, picking up context well and not requiring over-explaining, which means less heavy editing for me. Claude's consistent tone and style across long pieces are impressive, and the ability to translate vague creative directions into concrete language is beneficial. Its long context window is a big plus, reflecting the full picture without much back-and-forth. I find Claude's writing quality and reasoning superior, with less editing needed and a more natural tone. The initial setup was very easy as well. Review collected by and hosted on G2.com.

The message or token limits are the biggest frustration. It cuts off mid flow. Breaks momentum. Telling how much a task is going to consume before going ahead and doing it and leaving me without any credits to do anything. There should be range of tokens consumed and I want to approve on that before going ahead and using it all up. Although the initial setup was very easy, setup with Figma was a struggle. Review collected by and hosted on G2.com.

What I like most is that Claude feels like a cross-functional teammate, not just a writing tool. A lot of my day as a PM is spent translating between engineering, design, ops, and the business side, and Claude is genuinely good at keeping all of those lenses in view at once. It can walk me through a technical concept from scratch, then help me reshape that same idea into a spec for eng, a leadership-ready summary, and clear talking points for design.

The most helpful part is how much of the “connective tissue” work it takes off my plate: drafting PRDs, synthesizing messy research and interview notes into clear themes, prepping stakeholder updates, and turning a rough thought into a structured document I can actually share. It’s fast, it asks sharp clarifying questions instead of guessing, and it plugs into the tools and files I already use.

The real upside is the time and headspace it gives back. It compresses hours of drafting and formatting into minutes, which frees me up for the judgment calls, prioritization, trade-offs, and talking to users; that actually need me. Overall, it feels less like using software and more like having a tireless thinking partner on call. Review collected by and hosted on G2.com.

What trips me up most is that Claude is often too verbose and overly hedged. I’ll ask a direct question and get a long preamble, a stack of caveats, and three different options when I really wanted one clear recommendation. I’ve set preferences to keep it concise, but I still find myself trimming the response or re-asking the question just to get to the point.

The bigger issue for me as a PM is trust. It can sound just as confident when it’s wrong as when it’s right. It’ll state a number, a “fact,” or a summary in the same assured tone either way, so I can’t take its output at face value for anything that feeds into a decision. I still have to verify, and that eats into the time it’s supposed to save.

It also loses the thread in longer sessions. I’ll provide context or a format up front and then have to repeat it later, and the same prompt can come back structured differently from run to run. The tooling around it; connectors, re-authentication, setup, can be fiddly enough to interrupt the flow as well.

None of this makes it less useful. It just means it’s a strong first drafter and thinking partner, not something I can hand off unsupervised yet. Review collected by and hosted on G2.com.