You’ve seen the graphic. You, in the center of a wheel, ringed by thirty AI tools, each one promising to transform your business. It looks like progress. Mostly it just gives you a low-grade stress rash.
Here’s what I want you to do instead. Throw out the tool stack and build a problem stack. Put one AI coworker in the middle, put your actual problems around the outside, and start handing off the work. In the video below I walk through exactly how to do it using the apps you already pay for.
The tool stack is the wrong way to think about AI
Most people are still treating AI like a pile of subscriptions. Another login. Another tutorial. Another thing to feel behind on. The mental model is that AI is something you bolt onto the business in addition to everything you already do.
That framing is backwards. You don’t have a tool shortage. You have a problem stack. It’s the handful of recurring jobs that quietly eat your week: tracking action items, managing email, finding files you know you saved somewhere, making social graphics, capturing ideas before they evaporate. The point of AI isn’t to collect more tools. It’s to hand off as many of those problems as you can.
So flip the picture. Instead of you in the center surrounded by tools, put Claude in the center and surround it with your problems. Same wheel. Completely different job.
Build your problem stack first
Start by writing down the things that take real time in your business. Not the strategic work that only you can do. The repetitive stuff. Here’s mine, and yours will look similar: action items from calls, email triage, finding and organizing files, graphic design for social, idea capture. Put whatever belongs in your circle. You do you.
Then ask one question about each item: where does this problem live right now? You’re already using something to wrangle it. A notetaker plus Asana or a spreadsheet for action items. Gmail or Outlook for email. Google Drive or your desktop for files. Canva for graphics. Notes or Notion for ideas.
Here’s the part that should make you exhale. You’re not switching any of it. We’re going to make the tools you already know work harder.
Why this beats learning ten new apps
When you start with the problem instead of the tool, two things happen. You build real AI skills, and you build them inside software that’s already familiar. No migration. No new system to learn at 9pm on a Sunday. You’re just teaching an assistant to reach into the places your work already lives.
Let me tell you on myself. I went looking for the exact LinkedIn graphics I used to open that video. I couldn’t find them. I dug around for a solid thirty minutes before they turned up in one of my Google Drives. The same Google Drive I had already connected to Claude. I could have asked Claude to find them in about ten seconds, and I didn’t, because I forgot. AI is as much a habit as it is a tool. The habit is the hard part.
How to actually connect things
In Claude, in your browser, go to the customize area and look for connectors. Scroll through. These are the apps you can hook up. Pick one, click connect, and authorize it (usually by signing in to that app once). Do that for each app where one of your problems lives.
Then practice. This matters more than people expect. You don’t get good results from AI by reading about it. It’s like piano. You don’t sit down and play like a pro on day one. You run scales. You play easy pieces. You make mistakes and try again.
So play a few easy pieces. Connect Canva and try this: “Go to Canva and find the social post I made for my lead magnet about the low-tech way to save five hours a week.” When I did that, it found the design even though I didn’t remember the exact title. Or connect Gmail and your notetaker and try: “Go to Gmail and [your notetaker] and summarize any interactions I’ve had with [client name] over the last 30 days.” Out comes meeting notes, action items, and current status. Meeting prep, done.
Your four steps for this week
1. List your problem stack. Write down the recurring jobs that drain your time.
2. Identify the app where each problem already lives.
3. Connect that app to Claude using the connectors function.
4. Test it on one real task and see if your week gets a little lighter.
Start with a problem, not a tool. If you connect a pile of apps without a job to do, you’ll never open them again.
One thing a connector can’t fix
If your biggest frustration is that AI content never sounds like you, a connector won’t solve that. That’s a different problem with a different fix, and it’s the next thing worth your attention. Get your problem stack hooked up first, then go teach the AI to sound like you. That’s the combination that actually makes AI feel less like a chore and more like a coworker.
