Welcome to another Retro Newsletter. If this is your first time here, I highly recommend reading this newsletter first. It will give you the basics.

Your team is probably using AI tools. Maybe a few. Maybe more than anyone's keeping track of. And there's a good chance nobody has stopped to ask: is this actually working for us?

Not "are we using AI?" That ship has sailed. But: which tools are genuinely helping, which ones are quietly creating more overhead than they save, and which ones are just there because someone said we should try them?

That's the retro this month.

At the end of your next sprint, ask your team one question before the retro starts: "Name one AI tool you used this week. Did it save you time or cost you time?" Just the data. No discussion. You might be surprised what surfaces.

This retro is not about whether AI is good or bad. It's about whether your team's AI adoption is intentional or accidental.

There's a difference between choosing to use a tool and just absorbing it into your workflow because someone told you to. This retro helps your team distinguish between them and decide what to do about it.

Time: 60–90 minutes

Works for: Any team using AI tools - engineering, product, design, leadership, or cross-functional

Setting the stage is about container building, so remember to create a container that’s safe for your team. AI adoption carries pressure from leadership, from hype, and from the fear of being the team that didn't keep up. That pressure can make honest reflection hard.

Sometimes sharing it directly is helpful:

"We're not here to justify our AI use or celebrate it. We're here to understand what's actually helping, what isn't, and what we want to do differently. So that we can share that with the people who need to know and also make better decisions for ourselves."

CheckIn Idea - A one-word question to get the voices in the room.

"Share one word that describes your relationship with AI tools right now."

Give people 30 seconds to think, then go around the room. No discussion, just listen. This will give everyone in the room insight into what’s happening for people and might even be surprising.

For gather data, get everyone to write down individually on stickies all the tools they are currently using and for what, be specific here, so you can create clusters that give you interesting and valuable data.

Use the colour coding to show what’s, making them:

Cluster the stickies by tool or by task type and take a moment to look at the full picture together before moving on.

Now look at what's on the wall and look for patterns and insights based on what you see. Look at how the stickies are clustered and also how the colours are clustered and forming.
Some ideas for questions to ask.

Question 1: What do we notice about this data? - Where might we be getting value?

Look for patterns. Is the value concentrated in specific tools? Specific tasks? Specific team members?

Question 2: Where might we be feeling loads of pain? What could be causing the pain?

Look for a pattern here again, but don’t be afraid to dig a little. What might be a pain for one person is easy and adding huge value to someone else, and so that might be an opportunity to learn from each other or pair up.

Then you can also ask questions about what the team might let go of, or what could improve if we knew more, had more support, or had more tokens?

🔀 5. Decide What to Do

Think about whether there are tools some are adopting that everyone might benefit from, and what's actually making you, your team, or your code worse. Or are there some prompts that people can share, or a library that you can create that will help others in the team to be more effective?

Move into actions and push for specificity here:

The best outcomes from this stage are small and concrete. A team that decides to stop using AI for a specific task for two weeks, then reassess, is more useful than a grand AI strategy.

Use dot voting to find where the team's energy is, and commit to no more than two action items.

One last thing to consider is how you can ensure you collaborate and share insights effectively on an ongoing basis. Things move fast and having a place to share ideas could be beneficial. .

🚪 6. Close

Close with a recap of the decisions and clarity on the actions, and then an appreciation and a reflection.

What are you taking from today?

🔥 Additional Ideas

You don't have to run this retro in full. If your team is just starting with AI tools, try a lighter version: gather the tools you're using, rate each one - helping / neutral / overhead. Decide on one to keep, one to drop, and one to try differently. Ten minutes. Surprisingly powerful.

Or run it in reverse: start with your team's biggest pain points, then ask if there is an AI tool that could genuinely help here, and have we tried it?

When the team lands on a tool they want to stop using, resist the urge to move straight to "what do we replace it with?" See if you can stay in the problem space for a bit longer.

"What did we hope this tool would do for us?
What would that actually look like if we got it?"

Sometimes the answer is a different tool, but sometimes it's a better conversation, a clearer process, or just more time to think. AI tools often land in teams as solutions before the problem is fully understood.

Many of you found me through Agile coaching or a CSM course we did together. What you may not know is that what made that work meaningful, for me and, hopefully, for you, was never really about the frameworks.

It was always about people. About change. About what it actually takes to show up differently, not just do things differently.

That's where my work is heading now. Leadership and personal development. Helping people be different, not just better skilled. If that sparks something in you, I'd love for you to explore:

And if you have a brilliant retro idea, a technique that worked, or an insight your team had, send it to me at [email protected]. I'm building a community resource, and your idea could be in it. 🙌

If this retro surfaced things your team wants to go deeper on, or if you want to build your facilitation skills to handle exactly these kinds of conversations, I'm running a Retro Masterclass in May.

Three sessions. One conversation. Real change in how you run retrospectives.

📅 May 6, 13 and 20 | 5:30–7:30pm SAST 💻 Online 💸 R1,800

We cover setting up retros that actually fit your team, facilitation in practice (the resistance, the silence, the team that's stopped believing), and what it takes to create real improvement over time. After all three sessions, you get a 1:1 coaching call with me, bring a real team issue, and we work on it together.

👉 Book your spot here

🗓️ Motion — An AI scheduler that automatically builds your day around tasks, deadlines, and meetings. When a new meeting drops in, it reshuffles your to-do list around it. Fascinating to try, and a useful prompt for an AI retro question: would you trust an algorithm to plan your day? 👉 usemotion.com

🎨 Neal.fun — A collection of beautifully made, slightly strange interactive experiences. Spend the world's money. Watch the size of things. See how long you'd survive on other planets. Equal parts educational and oddly moving. A good place to send a curious mind on a slow afternoon. 👉 neal.fun

🌍 Radio Garden — Spin a globe and tune into live radio stations from anywhere on earth. Jazz from Tokyo. News from Lagos. A tiny station from rural Iceland. It's a three-minute reminder that the world is enormous and mostly doing fine. 👉 radio.garden

🍯 Honey never expires. Archaeologists have found 3,000-year-old honey in Egyptian tombs — still perfectly edible. Some things just don't need improving. 🐝

Till next time

Jo