Stop asking "what's our AI strategy?"

You'll just get bulls**t, not actual work.

If you’re a brand leader and that question is still on your meeting agenda, watch what happens.

You get twitchy. Someone suggests a chatbot. Someone else suggests hiring a head of AI. IT talks about vendors. Finance asks about ROI. Then you all go back to moving data between spreadsheets like it’s the default setting.

The problem with that question is it invites performance. You get decks. You get demos. You get everyone doing their bit to sound strategic.

What you don’t get is actual work being automated.

AI isn’t the destination. It’s a tool, and it’s getting baked into a lot of the stuff you already pay for. The better question is boring but that’s why it works.

How are we going to automate 30% of tasks by 2030?

The only target that matters

Pick a target that forces choices, and is close enough to feel uncomfortable. Put a date on it, even if it’s a rough date, not a perfect number.

30% of tasks by 2030 is a clean target. If you get to half that, you’d still have taken a big chunk out of the grind.

And it’s not a random number. It’s a simple target that’s big enough to change how work gets done.

It’s tasks, not jobs.

If you’re running a retail brand, you’re already running a service business internally. Merch serves trading, and finance serves everyone. Tech and systems teams serve everyone too. And no, this isn’t about cutting heads.

The point is to redesign how the work gets done. A lot of effort goes into moving data around, updating numbers, and reconciling systems.

That cost adds up. And in the AI-era, that work can be done differently.

A vague AI strategy doesn’t force any of that work to happen but a hard automation target does.

It makes you pick. It makes you prioritise. It makes you admit where ops are creaking.

It also makes board conversations easier. Despite the question “what’s your AI strategy?” coming up all the time. The board doesn’t actually need an AI strategy (and most wouldn’t recognise one that’s legit from one that’s bullshit anyway). It needs a decision framework for when automation is worth doing.

Where the real work is hiding

Go look at the places where teams hand over to each other. That’s where data gets copied, double-checked, and re-entered.

Look between merch and logistics. Between systems that were built for different decades.

That’s where the opportunity is. Not in a shiny demo. Not in a brainstorm about the future of AI. It’s in the grim, daily stuff that good people hate doing.

If you want a simple start, hunt for tasks with these tells. They’re boring, but they’re reliable.

• The same data is typed twice, usually in two different systems.

• The output is a spreadsheet someone emails round to get approved.

• Nobody can agree who owns the step, so it sits in limbo.

Finding the tasks and fixing them across teams is the job. It’s more people than tech.

You can often buy tools that sync product data across systems and cut manual re-entry. Buying the tool is the easy bit.

The harder part is agreeing what good data looks like, who signs it off, and what changes when you stop relying on heroic manual checks. This is why brands chase the wrong stuff.

The flashy customer-facing thing is easier to show progress on. It looks good in a deck. Meanwhile, the wins are in clear workflows and taking friction out. That’s the stuff that cuts cost, moves faster, and stops trading teams firefighting.

We’ve been saying this in retail for ages. Doing the boring implementation work, getting workflows clear, and protecting the brand is what separates the ones that get value from the ones that don’t.

Your hot new AI hire won’t save you

Here’s the thing. AI skills will get cheaper and easier to pick up over time.

It’s easier learning how to vibe code than to grasp the full reality of running an inventory based, fast growing business.

What’s scarce is context. The people who understand how an inventory-based retail brand really works, and can sanity-check what the machine got right or wrong.

You can’t learn that in a course. It’s lived experience. It’s knowing where the bodies are buried in your allocation logic, your size curves, and your month-end close.

So don’t build your plan around hiring a handful of specialists and hoping they’ll fix it for you.Get the leadership team clear on where work actually happens, how it flows, and why it breaks down. Then use AI and automation to remove the drag.

Think about it like spreadsheets. Most of your business doesn’t need to become a spreadsheet wizard.

But everyone should know what spreadsheets can do, and when to pull in the person who’s good at automating the button-clicking.

AI is the same…

What to say to your teams

The biggest thing I tell leadership teams is this. Stop pretending you’ve got it all mapped. Tell the truth, and set a direction.

AI’s going to change how we run this business. It’s going to change how we work.

We don’t know exactly how yet, but doing nothing isn’t an option. So we’re going to talk about it head-on, and we want everyone calling out where AI and automation can take busywork off people’s desks.

We’ll get better at it as we go.

It calms things down. It tells people you’re not outsourcing the future to a vendor. And it makes it okay for people to bring messy, boring problems to the surface.

Then you make it real.

You don’t need a big bang programme. You need a rhythm.

Pick a painful process. Map it end to end. Put the person who does the work in the room.

Bring someone who can configure the system and someone who can wire it up without breaking anything. Ship something small.

Share what worked and what didn’t.

If you want real automation by a deadline that matters, you won’t get it from one tool.

You’ll get it from a steady run of unsexy fixes that remove friction and stop good people spending their week doing admin.

And when someone asks what’s our AI strategy, you’ll have a better answer.

We’re aiming to automate 30% of tasks by 2030. Here’s where we’re starting.