Most people use Claude like a smarter Google search.
They open a new chat. Type a question. Get an answer. Close the tab.
If that sounds familiar, you're using about 10% of what you're paying for.
Here are 20 features ranked by how much they change the way you work, and most people paying $20 or $100 a month have never touched them.
The basics most people skip
1. Projects
The single most underused feature in Claude.
A Project is a persistent workspace. Every conversation inside it shares the same context, the same files, and the same instructions, permanently. You don't re-explain yourself every session.
If you've ever opened a new chat and typed "okay so I'm working on a SaaS app and the stack is…" that's a sign you don't have a Project set up.
Create one for each thing you work on. Your startup. Your content. Your code. Claude will know exactly who you are before you type a word.
2. Custom Instructions per Project
Inside every Project, you set a system prompt that runs automatically on every conversation.
"You are an expert copywriter. I run a crypto marketing agency. My audience is founders. Always write in a direct, punchy style. No em dashes. No fluff."
You write it once. You never explain yourself again.
This is the difference between a Claude that sounds like you and a Claude that sounds like a press release.
3. Project Files
Upload your documents to a Project and they stay there.
Your brand guidelines. Your style guide. Your product spec. Your competitor research. Claude reads them automatically at the start of every session.
Stop copy-pasting context. Start uploading it once.
The power features nobody talks about
4. Extended Thinking
Most people don't know there's a toggle that makes Claude slow down and actually reason through a problem before answering.
Turn it on for anything that requires real judgment: strategy decisions, complex debugging, legal or financial analysis, anything where a fast wrong answer is worse than a slow right one.
The output is noticeably better. The cost is a few extra seconds.
5. Code Execution
Claude doesn't just write code. It runs it.
Drop a CSV and ask it to analyze the data. It'll write Python, execute it, and show you the output, charts, tables, summaries, all inside the chat.
No Jupyter notebook. No local environment. No setup.
Upload your data. Ask your question. Get your answer.
6. File Uploads
PDFs, spreadsheets, images, code files, Word docs, PowerPoints. Claude reads all of them.
Drop a 200-page contract and ask "what are the termination clauses?" Drop your competitor's pitch deck and ask "what's their positioning?" Drop your Q1 financial report and ask "where are we losing margin?"
You've been copy-pasting things you could've just uploaded.
7. Vision
Paste any image directly into the chat.
Screenshots of broken UIs. Mockups for feedback. Diagrams to explain. Charts to analyze. Whiteboards to transcribe.
If you can see it, Claude can work with it.
8. Web Search
Claude can access the internet in real time.
Ask about something that happened today. Research a competitor that launched last week. Pull current pricing. Check the latest documentation.
It's not always on by default, look for the search icon in the toolbar and make sure it's enabled.
The ones that make you noticeably faster
9. Artifacts
When Claude builds something, a chart, a tool, a calculator, a mini-app, it appears in a live preview panel next to the conversation.
Not a code block you have to copy somewhere. A working thing, right there.
Every artifact gets a shareable link. Send clients a live calculator instead of a screenshot. Share a working tool instead of a description.
10. Data Analysis Mode
Upload a spreadsheet and ask Claude to find patterns, flag anomalies, or build a visualization.
It'll execute code, show you the result, let you iterate.
This replaced a full afternoon of Excel work for me.
11. File Creation
Claude can generate downloadable files, PDFs, spreadsheets, slides, and more, directly from chat.
You get an actual file, not text you have to manually format.
12. Voice Mode
Talk to Claude. Full conversational back-and-forth, not just dictation.
Use it on your phone while you're driving, walking, thinking. Use it to brainstorm out loud. Use it to prep for a presentation by talking through your ideas.
Most people have never used it. It's been sitting in the app this entire time.
The integrations nobody sets up
13. GitHub Integration
Connect a repo and Claude reads your actual codebase, not a snippet you paste.
Ask it to explain an unfamiliar module. Ask it to review a PR. Ask it to find where a bug might originate. It has real context, not a fragment.
14. Google Drive
Connect your Drive and Claude can pull documents directly into context.
No more copying from a Google Doc and pasting into chat. Point at the document. Ask your question.
15. Zapier and Tool Integrations
Claude connects to Zapier, JIRA, Intercom, and a growing list of tools directly inside the interface.
This isn't API work. It's a native connection you set up once in settings. Claude can then trigger workflows, check tickets, and respond to data inside your actual tools.
The Max-specific things
16. Claude Code (included in Max)
Claude Code is a terminal agent that reads your codebase, plans tasks, and ships changes, without you writing a line of code.
It's included in the Max subscription. Completely separate product. Most Max subscribers don't know it exists.
Install it with npm install -g @anthropic-ai/claude-code. Run claude in any project directory. Give it a task in plain English.
17. Multiple Models in One Subscription
You don't pay per model. You pay for Max, and you get access to Sonnet, Haiku, and Opus in the same interface.
Use Haiku for fast, cheap tasks. Use Sonnet for most work. Use Opus for the problems that actually need to be solved right.
Most people default to one model and never switch. The difference in quality, and cost, is significant.
18. Priority Access to New Features
New models and features ship to Max first, before Pro, before free.
When Claude 4 dropped, Max users got it days before anyone else. When new capabilities launch, you have them before your competitors.
You're not just paying for more usage. You're paying to be first.
The habits most people never build
19. Start a New Chat Between Topics
Context is finite. As a conversation fills up, Claude's quality measurably decreases.
Important instructions get buried. The model starts making mistakes it wouldn't make with a clean window. Most people blame the model. It's actually the workflow.
The fix: new chat for new topic. Every time. It takes two seconds and the quality difference is real.
20. Share Artifacts as Links
Every artifact Claude builds, a tool, a calculator, a chart, gets a public link.
Instead of screenshotting the output and sending a static image, send the live link. Clients can interact with it. Collaborators can explore it. The thing actually works.
Most people screenshot everything. There's a share button right there.
The cheat sheet
Bookmark the visual. Come back to it.
The people winning with AI aren't using a better model. They're using the same model with 20 features instead of 2.
If you found this useful, follow me @sharbel on X.
And if you run a business, we build content systems for founders in order to attract the right audience for you, and convert it into revenue. If you want to have these same systems installed for you, book a call with us at FounderFunnel.com.
