For most of my last two years, the main thing was my fractional CFO work. AI training and consulting sat on the side, more of a side project (very exciting though!). I still have two active fractional CFO clients, and I identify as a CFO!
This year I decided to change that and give the training real focus and structure: a three-session Claude in Action engagement for finance teams, in a corporate and mixed-cohort format. Over the past couple of months, I've trained several of them, across different industries and team sizes from 3 to 30.
I agree, a handful of teams isn't a perfect sample. But I found enough repeating patterns to see what works and what doesn't. I want to share them, whether you're rolling out Claude or any other AI tool on your own, or bringing in a trainer like me to help your team get there faster.
Claude in Action Updates
I'm wrapping a couple of corporate Claude in Action engagements this week, and I'm taking new ones for July. If your team could use it, reach out by email (just hit reply).
The mixed Claude in Action cohort wraps up today. What a group! I really enjoyed working with them.
The next cohort starts in July, and I'll announce the dates soon. There's a waitlist; sign up for a discounted rate and to hear everything first.
The newsletter keeps going, though I'll be traveling, so I hope you have a lovely few weeks ahead.
The Patterns
Most of these engagements were started by the CFO or a team leader. And most of those leaders chose not to sit at the center of the room. They sponsored the work and let the team own it.
The teams that moved fastest had a leader who wanted to make the team better, not squeeze more out of them. The framing was always "let's make your work easier," never "let's see who we can cut."
And that expectation paid out.
One team rebuilt their financial model and the full budgeting and FP&A process around it in four weeks. Past performance analysis, trend work, the levers and dependencies, all of it. That used to be a multi-month, full-time project.
Small tasks like reconciliations and simple reports went from a couple of hours a week to 20-30 minutes. These are usually the tasks everybody hates.
Month-end close came down by at least a day across every team. Accruals, lease accounting treatments, the parts that always run long.
Every team found tasks and workflows they could run faster with Claude. Some saved 3 hours a week, some closer to 10. And this isn't a static number; it keeps improving.
What sped up adoption
A champion inside the team, not the CFO. Usually a tech-savvy finance analyst or FP&A specialist with initiative and the team's trust. They collect use cases, track them, and keep the cadence going between sessions.
A warm start. A team already poking at Claude beats a cold start every time. If they are cold, plan for one to two weeks just to get them experimenting.
A simple use case tracker. The teams that kept one had a clear picture of what was working. The ones that didn't were running on memory and good feelings.
Letting the team surface the use cases. This was the surprise. People pre-judge their messy tasks, "this won't work, too many exceptions, my patterns are too weird," then test it anyway and watch Claude handle the mess better than they expected. The tasks they assume are off-limits are often the best candidates.
What didn't work
Sessions scheduled too close together. Two days between sessions and nobody has had time to apply anything. Space them seven to fourteen days.
Starting during close, audit, or budget season. People have no room to learn a new tool when they are buried. Pick a quieter window.
Handing the team a use case list. Work with them, not for them. The use cases they bring you are better than the ones you would assign, because they understand their own work.
The bigger picture
Here is how I think about Claude in a finance team. Three modes.
Mode 1, do it for me. Execute the work and hand back a clean output. Reconciliations, statement reviews, cash rec.
Mode 2, help me analyze. Break down results and surface what matters. Trends, fee calculations, variance drivers.
Mode 3, help me decide. Weigh the options and pressure-test the call. Stress-test an assumption, challenge a recommendation, question a model.
Almost every team I worked with showed real time savings and real enthusiasm for the tool. Almost none has moved to Mode 2 or Mode 3 yet.
That is the next frontier, and it's where the bigger payoff sits, when Claude stops being a faster pair of hands and starts changing how you think and decide.
First, a real thank you to my paid subscribers. Your support keeps this going, and several of you asked for more step-by-step instructions and frameworks you can actually use. So that is what this edition's subscriber section is.
I'm walking through one properly nasty task we automated with a client, start to finish, with the exact prompts and the checks around them, using a bank reconciliation as the example.
Closing Thoughts
The final note: the standard training structure is three sessions, but in almost every engagement we customized it. Some teams didn't have every tool approved yet. Some were further along and skipped the first session. The framework stays the same. The path through it depends on where your team starts.
I'm continuing with this, no question. There wasn't one unhappy team. Even the people who thought they were already advanced walked away having learned something.
If this is the kind of thing your team could use, reach out. I'm not taking calls for the next three weeks while I'm traveling, but email always works (just hit Reply!).
Until next Tuesday.
Anna
We Want Your Feedback!
This newsletter is for you, and we want to make it as valuable as possible. Please reply to this email with your questions, comments, or topics you'd like to see covered in future issues. Your input shapes our content!
Want to dive deeper into balanced AI adoption for your finance team? Or do you want to hire an AI-powered CFO? Book a consultation!
Did you find this newsletter helpful? Forward it to a colleague who might benefit!
Until next Tuesday, keep balancing!
Anna Tiomina
AI-Powered CFO