My 3 interesting things for you this month…
1. Resolving challenges with AI
Last month we talked about solving ‘How problems’ with AI. Hopefully some of you have made some progress. You can check out the full piece on my site, but the summary is rather than delegating your work to AI, use AI to help you be better at what you do.
When people try this, they tend to hit three common barriers. Here’s how to get past each one with practical steps you can try right away.
1. The answer didn’t feel useful enough
Sometimes you get a response that isn’t quite right. Maybe you already know what it said, it’s not really relevant or the answer won’t apply with your constraints.
If that’s the case, you can fix it with context and constraints. Think of AI like a new member of your team – it can do brilliant work, but only if you brief it well.
AI performs to the quality of the context you provide. Before you hit “go,” use a simple structure to improve your responses: goal, context, constraints and quality bar. Even a few lines can make a big difference.
And if the problem is complex, you’ll often get better reasoning from a more advanced model (probably paid ChatGPT “Thinking” – and don’t forget to turn on “Extended thinking” as well!).
A helpful follow-up prompt when the reply feels generic: “Given the goal, context, constraints and my quality bar, suggest three options. For each, show the exact steps taken, effort required and any risks.”
2. I don’t even understand what it means!
I was running a training program for a team last week, and ChatGPT gave an amazing answer to drastically improve the process. But… the team leader said “I’m not sure this is useful, no-one understands what these things means”!
When the language doesn’t land, the idea won’t either. The good news is AI is very good at rephrasing, summarising and explaining answers. A response that doesn’t land isn’t a dead end – it’s a learning opportunity.
Take the moment to build your own understanding:
- Use “Ask ChatGPT” on a highlighted phrase. Select any terms or phrases you don’t understand and click the icon to get a more detailed explanation.
- Ask follow up questions. Questions like “what’s the core principle behind this suggestion?” can help you get a deeper understanding.
- Ask for a plain-English rewrite. A simple prompt like “Explain this as if I’m a finance leader, avoiding jargon and giving me a before vs. after example.” can help AI rephrase in a way that lands.
The goal here isn’t to become an expert overnight, it’s to get just enough clarity to decide whether the idea is worth taking to pilot.
3. I understand the concept, I don’t know how to do this
If you get the idea but aren’t sure how to start, use ChatGPT to turn it into a small, safe experiment.
Why this helps: it lowers the friction to begin, turns a concept into a few doable actions and gives you enough structure to involve others without creating extra work.
Some practical steps to take:
- Ask for step-by-step instructions and a rough checklist based on your context, plus what “good” looks like for a first pass.
- Add a short note you can share with the people who need to be involved (what we’re trying, why now and what we need).
- Ask for any risks to watch out for so you don’t trip over basics.
If you want one prompt that does all of that: “Based on this idea and our context, outline the first three steps I should take this week, a simple checklist for each, what ‘good’ looks like for a first attempt, a short message to brief my team and the top two watch-outs.”
Run it once, then ask: What worked? What didn’t? What would we change next time? That’s enough to move from understanding to doing, without overcomplicating it.
2. Focus as a service: the most valuable product finance can ship
Here’s an interesting thought. What if the most valuable thing finance delivers this quarter is not the monthly report or the 2026 budget, but focus?
Clear, confident focus helps people put energy into the few things that will actually move the needle and stop spending time on the noise.
Here’s a 3 step approach to delivering this:
1. Figure out what matters most
You might know the target, but how well do you understand the drivers that get you there – and the few challenges that hold you back?
These drivers are what really shape performance. Ideally they’re linked to strategy and real business behaviour, not just variance to forecast. Once you pin them down, your work becomes more impactful and your partnering gets sharper because you can talk in terms of actions and impact.
Compare these two statements:
- “We need to improve revenue by addressing pipeline quality, pricing and retention.”
- “If we lift conversion from trial to paid by 2 points in mid-market this quarter, we’ll be on track.”
The second one tells people what to do next. That’s the level to aim for.
2. Coach for clarity
Once you have the driver, create an ongoing rhythm to keep that focus moving forward.
Meet partners regularly and talk about the most important numbers and the stories behind them. Keep the “what” and “why” of key metrics short and clear, then move onto the action-focused question of “how could we…”
Following the “what, why, so what?” framework, which I’ve covered previously, is useful here. Here’s an example of what that looks like:
What: “This week’s focus is trial-to-paid conversion. We’re at 21% against a 23% target.”
Why: “Drop last week driven by slower follow-up on day two.”
So what: “How could we get 30 more day-two contacts next week without adding cost?”
That “How could we…” question is doing a lot of work. It keeps the conversation forward-looking, invites options and respects the team’s expertise.
3. Bring accountability
Your job isn’t to come up with the answers of what to do next, it’s to ask the questions that facilitate partners figuring that out.
Once they have, your role is to make progress visible and keep accountability. Create a handful of commitments with an owner and data, then close the loop next time you meet. Keep the list small enough to fit on a slide or the top of an email.
If a commitment slips, stay curious: “What got in the way? What’s the smallest tweak that gets us moving again?” The accountability is in the visibility, not in micromanagement.
We spend a lot of time producing data, numbers and reports – in a world where AI will take over more of this work, providing focus for the business keeps us relevant and impactful.
3. Moving to action
Most finance teams are drowning in information but struggling for insights that drive action.
We write extended packs, host long reviews, yet the actions that change results are often late or just lost in the noise.
Over the summer, I’ve been working on a simple sequence with a few clients business partnering senior leaders in fast paced businesses, to shift from commentary to action. It turned a noisy pipeline of data into a rhythm of clear choices, faster delivery and – we hope – ultimately better performance.
I think of it as an Action Framework for Finance. Seven practical steps that any finance leader can run with a whiteboard, a calm voice and a good question.
1. Get rid of everything you can: delegate, delete, delay
Start by freeing capacity. Ask “What would break if we stopped this for two weeks?” If the answer is “not much,” park it. Delegate what grows others, delete what no one reads, delay what doesn’t move a driver right now. The point is to make space for the work that matters.
2. Use AI to add structure, process and scale to what remains
Use AI to sketch a checklist, a standard outline or a short note you can reuse. It’s not about perfection. It’s about a consistent first draft that saves an hour and gets everyone looking at the same shape.
3. Build a simple driver model for your organisation and align the team around it
Get to the small set of inputs that actually move results. One page is enough: the business goal, three to five drivers, the best proxy for each and who owns it. Share it widely.
When people talk about numbers through drivers, the conversation moves from “what happened” to “what we’ll change.”
(P.S. I’ve covered building a driver model in depth on my site – so if you need a place to start, make sure to give it a read.)
4. Prioritise the team’s time against the key drivers
Map your week to the model. If a meeting or analysis doesn’t support a driver, that’s where delegate, delete, delay comes in. If a driver is off track, assign real time to it.
5. Ask open questions that lead partners to commit to what they will do to move a driver
You don’t need to have the answers. You do need to ask the questions that surface them. Open, “how could we…” questions go a long way here.
6. Hold them to account for the commitments they made
Capture who will do what by when with a line on expected impact. Bring it back next week with a simple loop: what we tried, what we learned, what we’ll do next.
7. Prepare crisp choices for the decision maker: invest, pivot or stop
When a call is needed, arrive with three options that respect the driver model: invest, pivot or stop. One slide each, trade-offs clear, recommendation on the page. This is about turning analysis into a decision, not another discussion.
This is not theory. My aim is for it to become a weekly habit that moves energy from analysis to action.
In the next few newsletters, I will unpack some of the steps with examples you can apply, the pitfalls to avoid and approaches to use with business partners. Let me know if it works for you!
Want more insights?
Sign up to my interesting things newsletter to get tips like this and more.
