I was at dinner last week with a guy my husband works with. He uses Claude every day. His whole team does, and somebody on his team is building out workflows right now that they’re getting ready to roll out. This is not a guy who’s new to AI.
He knew I’d done some training for my husband’s team, so he asked what my tips were. What did I think people actually needed to know.
I started talking about connectors, how that’s been the biggest shift I’ve seen with the teams I’m working with lately, because connectors are what let all the different tools a team already uses come together in one place. I was a few sentences in when he stopped me.
“What’s a connector?”
That stopped me too. I’ve been in this world long enough that I assume it’s common knowledge by now. It is not. This guy uses Claude daily and had never heard the term. My husband jumped in here too, because he learned this from me a few months ago and set it up for himself. He was the proof sitting right there at the table.
So here’s the plain version, for anyone else who’s been nodding along without asking.
So what is a connector?
A connector is a bridge. It lets the AI you’re using reach into the tools you already work in, instead of only knowing what you type into the chat box.
You might hear it called a connector in Claude, a plugin in ChatGPT, or an integration if you’ve been using software for a while. If it’s connecting to one specific piece of software, you’ll sometimes see the term MCP, short for model context protocol. I wish I didn’t have to hand you an acronym, but it’s the term you’ll run into, so it’s worth knowing.
Here’s where I’d start
Your calendar and email. This is usually the easiest one to get approved, and it’s the one that shows you what’s possible fastest. Ask it to scan your inbox for anything you’re waiting on a client to answer, or to draft a reply to something you never circled back on.
Your file storage. Drive, SharePoint, Notion, whatever you’re on. Now it can find the actual document instead of you hunting for it. A new hire asking “where does this live” stops being a twenty minute Slack search.
Your CRM. Connect it and it can pull a contact’s history before you jump on a call, or check who you’re meeting today against what you already know about them.
Your notetaker. Fathom, Otter, Fireflies, whatever records your calls. It can tell you what actually got said and decided, without you re-listening to a forty five minute call to find the one line that mattered.
Your project management tool. This is the one I use the most. Connect your team’s system and your own task list, and it’ll give you one combined rundown of what actually needs to happen today instead of you checking two places.
If your company hasn’t approved connecting something like your CRM or your Drive yet, don’t wait to find out how this works. Try it on your own calendar and email first, on your own account. It’s twenty dollars a month. Once you’ve felt it work on your own stuff, you’ll know exactly what to ask for at work.
We all lose real time to context shifting. Trying to remember where you saved that one document. A new hire asking where something lives, and you realize you’ve never actually written it down, you just know. A connected AI can find that in seconds instead of you hunting for it.
What if there’s no connector for your software yet?
Not every piece of software has one built. I’m working with a team right now dealing with exactly this, they have a maintenance tracking system with no native connector and no MCP available. Here’s the part that surprised even me: you can tell the AI what the software is, what you need it to do, and ask it to build the connection itself. It can go find the public information it needs and build it. Or if that’s not possible, it can use your browser to get you the same outcome by navigating the tool directly. You don’t have to be technical for any of this.
Here’s what’s actually happening behind the scenes
Every app and program you use is built on code, and developers spend their careers building interfaces so you never have to see that code. AI can see it. It can write it. That’s what’s happening on the back end when you set up a connection, it’s coding, quietly, on your behalf. That’s also why you’re not stuck waiting for someone else to build the thing you need. If you can picture the workflow in your head, you can ask AI to help build it.
When ChatGPT and Claude first launched, people used to talk about the day you’d open your AI and it would just tell you what’s in your inbox that needs a response, and what’s sitting on your task list. I remember not being sure how soon that day would actually come.
It’s here. And as business owners, we get to decide how much of that context shifting we’re still willing to carry ourselves.
If you want help figuring out where a connector could take real weight off your plate instead of adding one more thing to configure, that’s exactly what we work through on a free AI Clarity Call.