I’m not the type to update my operating system the second something new drops. Updates can mean things slow down, or the battery drains faster, so I wait and watch how it goes for everyone else. This time I read reviews for a couple of weeks, and by this past weekend the bugs seemed fixed, and the reviews were getting better. I was excited to update, partly because I was tired of feeling stuck in the past (what does that say about me?!), and partly because the platforms I run my business on keep updating too and eventually stop playing nice with an old operating system. So I updated my MacBook to Tahoe.
There’s always a price to every update. So was this time…I got locked out of Claude, which kept telling me something was wrong and try again. Well, I did, and nothing helped. It turned out, the authentication got messed up after the updates.
The fix I found after googling told me to delete some files, and even after double-checking that it wouldn’t affect any of my work, it still deleted 70% of my scheduled tasks that were running every day supporting my business. That annoyed me AF.
Working with tech and systems, an issue like this happened to me more times than I can count, and years ago this would have flipped me for days. Luckily, I’m more of a grown-up and experienced, so kudos to me for bouncing back quickly and looking at solutions instead of rage. I looked at it as: what can I learn here? Maybe this is my chance to do what I keep telling other people to do: document your process.
You gotta practice what you preach 🙂
Why founders don’t trust AI yet
Funny, just this weekend I was doing research on how small businesses adopt AI, and I found some interesting numbers. A lot of founders are struggling to bring AI into their business, and it comes down to three things: trust, ROI, and knowing where to start:
- Trust: 76% of small businesses already use AI, but only 14% say it’s fully built into how they run the business (Goldman Sachs’ 10,000 Small Businesses Voices survey).
- Return on investment: 74% of business owners who haven’t fully adopted AI say they’d move forward with clearer proof it pays off (PayPal’s Reimagine Main Street survey).
- Where to start: 51% of small businesses are what that same survey calls “Explorers,” trying tools but not committed yet, and 37% say they just don’t have the time to properly explore what’s out there and where to even start.
If any of this sounds like you, this is your proof: you’re not alone and it’s not just in your head.
Here’s how I worked through all three, while getting my scheduled tasks back, using this week as my own case study.
1. Don’t believe AI’s first no
The first thing I did after getting locked out was ask Claude to help me get my tasks back. It told me, flat out, that it couldn’t find them, that they were gone. I clapped at the keyboard, tried from different angles as I knew that wasn’t true. I created each task right there in Cowork, so it was more than frustrating to read he cannot find it.
Guess what, I started a new chat today and asked the same thing. It suddenly found everything. WTF, right?
Lesson one: when AI tells you something isn’t possible, I don’t take that at face value anymore, especially when I’m sure I’m right. I ask again, and a lot of the time, it finds what it “couldn’t” find the first time.
Trust comes with time and practice, same as with a person. You don’t trust someone the day you hire them, it takes watching them handle real work and correcting them a few times. The difference with AI is, when you’re frustrated with it, there’s nobody else to blame. It’s a tool, and it’s only as good as how you trained it, tested it, and corrected it.
I’ve written before about the moments a client’s trust in you is won or lost. This is the same idea, pointed at yourself instead of your clients: do you trust the AI you already pay for?
2. Find what to hand off
The second thing I want to share is the process I use for automating tasks in my business. It starts with pen and paper, or a screen and keyboard: I write out a list of everything I do every day. If you don’t have time to sit and write, use the microphone on whatever AI platform you’re on, or a dictation app, talk it out, and have AI clean up your rambling into something readable. That’s how I write a lot of what I publish, including this post, my jabber babber turned into an actual sentence.
I have calls: some are networking calls, some are client calls, and both need prep and a follow-up.
For networking calls:
- Who referred this person, what they do, their LinkedIn, Instagram, and website, all before we talk. I care less about small talk and more about asking real questions about their work.
- After the call, anything I promised goes into my project management system or my calendar so I follow through. If they mentioned wanting to be on podcasts, I pull up my networking tracker and connect them to people I know.
- Every person gets logged there too: name, location, profession, what they’re looking for, notes from our conversation, especially anything personal. A birthday coming up, a surgery, goes on my calendar as a reminder to check in. That’s the part that matters to me.
For client calls, I have AI pull up that client’s profile in Notion, where I keep everything: past calls, memos, open tasks, anything I need before we talk. After the call, I send a follow-up note so we’re both on the same page, and every task goes into their profile and onto my calendar.
That’s really it: prep for the call, and what happens after. I think of AI as my assistant on steroids. When I was a kid, I used to watch successful business owners on TV who had executive assistants running their whole operation, and I thought, how cool would that be. Now AI does that job for me.
3. Track your hours to find the return
Before I automated any of this, looking up information before a call, or checking in on client projects, took fifteen to twenty minutes. With five or six calls in a day, that adds up fast. Now that same review takes five to ten minutes. I know this because I took the time to measure. Oh boy, that was tough to watch…when you look at where your time goes, suddenly you realize you’re not as productive as you’d like to be.
You see the ROI once you know what you’re measuring and where your time is going in your business. Start with the pen-and-paper list, track what you’re doing and how long it takes, and that’s when you know what to automate.
If you have a team, ask them to do the same. Just for a week. If you use timesheets as invoicing for work done for clients, you’ll see lot of the billable hours as well as the overhead – that is a good indicator of where their time is spent on a day-to-day.
Look for the pattern. If everyone’s spending hours prepping for calls, that’s your first task to hand off, you buy back time for you and for them. Once you’re picking a tool based on what you measured instead of what’s trending, you’ll know it’s the right one.
If you want a hand figuring out your own first task, that’s exactly what a free Strategy Call is for. No prep needed, just tell us about one busy day. And if you want the longer list of tasks worth handing off, I put one together here: five tasks worth handing your AI assistant this week.
If you want more of this in your inbox, the Clarity Letter goes out most weeks with one useful thing to try.
