The Claude Quickstart for Content Creators

Go to claude.ai and click "Sign up." Once you're in, you'll land on the main screen. That's your workspace.

The layout is intentionally simple:

• chats on the left
• typing area in the middle
• Profile and settings in the bottom left

You'll figure it out in about 30 seconds.

It's $20/month and the free plan caps out the moment you start doing real work. I've been on Pro since week one, and the upgrade pays for itself in a day if you're spending 30-40 minutes editing every AI caption right now.

Pro gives you:

• longer conversations
• better performance
• more room to work with files and projects

Before you start working inside Claude, take 5 minutes to connect the tools you already use. You do this once at the account level, and Claude can pull from them anywhere, inside any project, without you reconnecting every time.

The ones worth connecting right now:

• Google Drive, so Claude can read your docs, sheets, and slides without you uploading them
• Gmail, so Claude can pull from emails when you need it
• Google Calendar, for scheduling context
• Canva, if that's where your design assets live

Connect whatever applies to your setup. You don't have to connect all of them, but Google Drive is the most useful one to do first. Once connected, any Google Doc you reference in a project stays live and auto-updates. No re-uploading when you make changes.

Quick thing before we go further, because this is what trips everyone up. Claude isn't one product. It's a whole family, all running on the same AI, each built for a different job:

• Claude Chat - the web version at claude.ai. Conversations, projects, files. This is the one we're in.
• Claude Cowork - the desktop app. Everything Chat does, plus it can actually touch the files on your computer.
• Claude Code - made for developers writing code. You can wave at it and keep walking.
• Claude in Chrome - a browser extension that lets Claude see your screen. Cool, but a later problem.

Only two of these matter to you today: Chat and Cowork. The rest are worth a look down the road, but trying to learn all four at once is how you end up with 12 tabs open and nothing done.

Chat is the easy way in. You type, it answers, pretty much like the ChatGPT you already know. Captions, brainstorming, emails, content planning, quick questions. This is where you'll live for most of this guide.

Cowork is the desktop app, and it does the thing the browser simply can't: it creates, edits, and organizes folders right there on your computer, and it remembers your context from one session to the next instead of forgetting you the second you close the tab. That memory part has a name. It's called persistent memory, and it's the whole reason Cowork feels less like a chatbot and more like a coworker who already knows the account.

One last thing so the settings don't throw you. Claude runs on different models, which are just different versions of the brain underneath. Anthropic swaps and renames them constantly, so don't get attached to the labels. There's a fast one for simple stuff, a reliable everyday one, and a heavyweight for the hard thinking. Easiest move while you're finding your feet: flip the adaptive thinking toggle on and let Claude pick for you.

You'll see everything it has stored about you. Revise all information in this section to ensure that all information is accurate before transferring to Claude.

Keep things like: your business, your audience, your offers, your communication style, how you like AI to respond, your current projects.

Delete things like: one-off random requests, outdated projects, old goals that don't apply anymore, conflicting information.

FYI: There is no save button. Once you close this section, the new info will be saved.

Now that you have cleaned up your memory, open a fresh ChatGPT conversation and paste this in:

I'm moving from ChatGPT to Claude and I need to bring you with me, basically.

Based on everything you've saved about me in memory, and what you've picked up across our past conversations, write me a complete brain dump covering:

1. Who I am: name, what I do for work, what business I'm building, who I serve
2. How I think and make decisions: what I prioritize, what I dismiss, how I weigh trade-offs
3. How I communicate: the words and phrases I use, the ones I avoid, the tone I default to, how formal or casual I am
4. My background: career, skills, anything that shapes how I approach problems
5. What I'm currently working on: active projects, products, deadlines, what I'm trying to figure out right now
6. How I prefer to work with AI: do I want pushback or agreement, short answers or long, do I like options laid out or one strong recommendation
7. Recurring patterns: anything you've noticed about how I work, what frustrates me, what makes me trust an output

Write it as a single document I can paste into a new AI tool. Be specific. Use direct quotes from our past conversations if you remember any. Don't be diplomatic. If I'm someone who gets impatient with fluff, say so. If I push back hard on certain types of advice, say so.

The goal is for a new AI to read this once and pick up where you left off, without me having to re-explain who I am.

Then let ChatGPT generate the document. Copy everything that GPT gives you from this prompt into a Google Doc. You'll use it in the next step.

Read through this document and again ensure all info is correct and update accordingly. Sometimes ChatGPT remembers weird stuff.

Paste the cleaned-up Google Doc into this section and click Add to Memory.

Find the section called "Instructions for Claude." That's the box. Whatever you place there becomes the starting context for every new conversation. This is the setup step.

Here's the full template. Copy the whole block, paste it into the Instructions for Claude field, then fill in the brackets right inside Claude.

1. IDENTITY
I'm a [your role: content creator, coach, course creator, copywriter, digital product seller] who [what you do and who you serve, e.g., teaches women 30-45 how to use AI without their content sounding generic]. I'm building this while [your current situation, e.g., working a full-time corporate job], so [the constraint that shapes your work, e.g., my time is limited and my output has to count]. My audience [where they're at right now, e.g., already uses ChatGPT but keeps rewriting everything because the outputs don't sound like them].

2. HOW I WANT YOU TO RESPOND
[How you want context handled, e.g., Skip unnecessary explanations and execute when you already have enough context.] [How you want feedback delivered, e.g., Don't give me soft feedback, lead with what's wrong.] [How you want copy delivered, e.g., When I ask for copy, give me clean paste-ready text with no intros or outros.] [How you want strategy delivered, e.g., When I ask for strategy, show tradeoffs first before executing.]

3. CURRENT GOALS
I'm currently [your stage and main initiative, e.g., launching a digital product business]. My focus right now is [your top 2-3 priorities, e.g., building lead magnets, creating launch content, and increasing sales from Instagram]. My main priority is [what you're optimizing for, e.g., consistency and speed without sacrificing voice].

4. BUSINESS CONTEXT
I primarily create [content types and platforms, e.g., Instagram content including Reels, carousels, stories, and captions]. I also write [other content you produce, e.g., emails and sales pages]. Most of what I need help with is [your highest-frequency Claude tasks, e.g., short-form social and marketing copy].

5. AVOIDANCES
Never use [specific words you hate, e.g., em dashes, "elevate," "transformation," "unlock"]. Never use [structural patterns you hate, e.g., motivational coaching language, "It's not X, it's Y" sentence structures]. Never write like [a tone you reject, e.g., LinkedIn]. Never open with [phrases you'd never use, e.g., "Here's the thing," "Let's be real"]. Avoid [overall vibe to avoid, e.g., sounding polished or corporate]. Write [your actual default, e.g., conversationally, like you would in a voice note].

That's it. Seriously. People think there must be some hidden advanced setup. There isn't. This alone usually improves outputs immediately. And later, when you start adding brand files and voice documents, it gets even stronger.

Good examples: Content Business, Launch Content, Email Writing, Client Work.

Bad example: Stuff.

You're going to end up with multiple projects eventually, so make them easy to identify. Once your project is created, the connectors you set up in Section 2 are already available here. You don't need to reconnect anything.

You can create a project from scratch in Cowork, or import an existing project from Chat if you've already built one there. The difference is what you can do inside them depending on which mode you're in.

Create your project in Chat if:

• You're writing content, captions, emails, or copy
• You're referencing Google Drive docs or uploaded files
• You're using Claude the way most people use it, conversations back and forth

Create your project in Cowork if:

• You want Claude to read or write files that live on your actual computer
• You're building a folder-based system (brand docs, content banks, research files)
• You're using skills or connectors that require the desktop app

If you're just getting started, Chat is fine. Most of the tasks in this guide live in Chat. Cowork becomes more useful once you've got files on your computer you want Claude to access directly.

Your conversations inside a project help shape context. But they do NOT permanently rewrite your instructions.

So if you tell Claude "always do this from now on" inside one conversation, it may remember temporarily in that thread. But if you want something permanent, update the actual instructions or files directly.

Think of projects like ongoing workspaces, not permanent brain surgery.

Take the next piece of writing you'd normally start in ChatGPT and run it through Claude using this format instead. That's when you'll see what the setup actually changes.