4 Claude Settings ChatGPT Users Always Miss
You switched to Claude. Good. Now stop using it like ChatGPT.
Every week someone messages me: "I switched to Claude but it doesn't feel that different from ChatGPT."
Same story every time. They signed up, typed a question, got an answer, and thought: this is basically the same thing.
It's not. They just skipped the setup.
Claude has four settings that completely change how it works. Most people never touch them because they're not obvious, they're not in any onboarding flow, and ChatGPT doesn't have equivalents for half of them.
Here's what they are, how to enable them, and why Step 4 turns Claude from a chatbot into a builder.
1. Turn On Memory
Where: Settings > Memory and Personalization > Toggle on
What it does: Claude remembers details about you across conversations. Your name, your job, your preferences, your projects. Every new chat starts with context instead of a blank slate.
Why most people miss it: It's off by default. If you don't go looking for it, you'll never know it exists. Without it, every conversation starts from zero and you end up repeating yourself constantly.
What changes when you enable it:
- Claude remembers your communication style preferences
- It tracks your ongoing projects without re-explanation
- Responses get more relevant over time, not less
- You stop wasting the first 3 messages of every chat on context
The difference: Without memory, Claude is a stranger every time you open it. With memory, it's someone who knows your work.
2. Import Your ChatGPT History
Where: Settings > Import conversations
This is the one nobody talks about.
You've probably had hundreds of conversations with ChatGPT. Projects you've worked on, preferences you've expressed, problems you've solved. That's months of context sitting in your ChatGPT account doing nothing.
Claude can import all of it.
How to do it:
- Go to ChatGPT > Settings > Data Controls > Export Data
- Wait for the email (usually a few minutes)
- Download the zip file
- Extract it and find
conversations.json - Upload it to Claude with this prompt:
Analyze this ChatGPT conversation JSON to extract up to 30 concise, third-person factual statements about me. Focus on actionable insights for future personalized responses, prioritizing: - My professional role, side projects, and pursuits - My active goals and strategic objectives - My preferred tools, platforms, and workflows - Relevant personal context and interests - Key recurring topics I frequently ask about Save each statement to your memory. Exclude sensitive, one-off, or ephemeral details.
What happens: Claude reads through your entire ChatGPT history, extracts the important stuff, and saves it to memory. In five minutes, Claude knows you better than ChatGPT did after months.
You're not starting from scratch. You're transferring all the context you've built up.
3. Get Your Personalized Learning Path
Where: https://anthropic.skilljar.com/
This is where it gets personal.
Anthropic has a certification and skills page that covers everything Claude can do, from basic prompting to advanced features. Most people glance at it and move on. That's a mistake.
Here's what you do instead: take the course content, paste it into Claude, and ask Claude to build you a personalized learning path based on what it already knows about you from Steps 1 and 2.
The prompt:
Based on everything you know about me from our conversation history and your memory, analyze these Anthropic certification modules found here https://anthropic.skilljar.com/. Identify: - Which skills I already have based on my usage - Where my biggest gaps are - What would give me the most value to learn next - A recommended learning order personalized to my goals and workflow Be specific. Reference my actual projects and tools where relevant.
What happens: Claude doesn't give you a generic "start with prompting basics" guide. It looks at your imported history, your saved memory, your actual usage patterns, and tells you exactly what to learn next and why.
Someone who builds apps will get a different path than someone who writes marketing copy. Someone already using Artifacts will skip the basics and jump to advanced patterns.
The shift: Instead of browsing a course catalog hoping something is relevant, Claude curates the curriculum for you. It knows what you already know.
4. Build Your 30-Day Dashboard (The Knockout)
Where: Same conversation. One more prompt.
This is the move that turns everything into something you can actually use.
You've got memory on. You've imported your history. Claude just gave you a personalized learning path. Now you turn that path into an interactive, trackable 30-day dashboard.
The prompt:
Now take that personalized learning path and turn it into an interactive 30-Day Claude Power User Roadmap dashboard. Use HTML, CSS, and JavaScript. Include: - Daily breakdown based on the skill gaps you identified - Progress tracking with checkboxes for each day - Resource links for each module - Visual progress bar showing overall completion - Clean, modern UI Output as an interactive HTML Artifact.
What happens: Claude generates a fully interactive dashboard. Not a text document. Not a PDF. A working web app with checkboxes, progress bars, daily modules, and resource links, all personalized to your specific skill gaps.
You can click through it. Track your progress. Come back to it every day.
And because it's built on your actual learning path from Step 3, every day is relevant. No filler modules. No "introduction to AI" when you've been using AI for a year.
That's the difference. ChatGPT gives you text. Claude gives you tools.
The Full Setup (Under 10 Minutes)
| Step | What | Result | Time |
|---|---|---|---|
| 1 | Enable Memory | Claude remembers you | 30 sec |
| 2 | Import ChatGPT | Months of context transferred | 5 min |
| 3 | Personalized path | Claude tells you what to learn | 2 min |
| 4 | 30-day dashboard | Interactive roadmap built for you | 1 min |
Under 10 minutes. Four steps. Each one builds on the last.
By the end, you have an AI that knows your work, a personalized curriculum based on your actual skill gaps, and an interactive dashboard to track your progress. All from a tool most people are using like a search engine.
Why This Matters
The gap between people who use AI well and people who use it badly is getting wider every month.
It's not about which model is smarter. GPT-4o and Claude Sonnet are close enough that the model barely matters for most tasks.
The gap is setup. It's knowing that these four settings exist and spending 10 minutes to enable them.
The people who do this get an AI that knows their work, builds them personalized tools, and gets better every conversation. The people who don't get a fancy search engine that forgets them every session.
10 minutes of setup. Months of better output.
Your call.
Want more breakdowns like this? Reply and tell me what you want next:
- Advanced Artifact patterns (dashboards, calculators, apps)?
- How to build custom Claude skills from scratch?
- Claude Code vs Claude Pro: which one and when?
- The prompt engineering techniques that actually matter?
Mike Oddoye
Building @ OptiMAX
@mikemeansbizz
P.S. Step 2 is the one most people skip because it sounds like effort. It's five minutes and it's the biggest unlock on this list. Export your ChatGPT data today. Step 4 won't hit nearly as hard without it.