Hey Reader,

This week I was connecting Claude to my analytics data (again) and let it do its thing.

And honestly? It’s impressive. You ask a question, it pulls the numbers, writes a summary and even makes a presentation or a website out of it.

No logging into GA4. No clicking through five menus to find the report that should’ve been on the home screen in the first place.

Which got me thinking...

If AI can just grab your data and tell you what’s going on, why do we need dashboards at all?

At what point do we ditch GA4 (and every other platform) completely?

In my opinion...

Not anytime soon.

I think we’ll still be using dashboards for at least the next 5 to 10 years. And I’m saying that as someone who thinks GA4 kinda sucks.

Here’s why.

1. We’re visual creatures.​
We want to see the thing. A line going up. A bar that’s way shorter than last month.

Yes, AI can make you a chart. But it makes it slightly differently every time. A different layout, a different date range, sometimes even a different definition of the same metric.

You can teach it. You can write the perfect prompt. But that consistency just isn’t there yet. But over time, this will get better.

And consistency is the whole point of a dashboard. You see the same view every Monday, so you notice right away when something’s off.

2. Your brain needs the reps.​
AI is great at throwing out ideas. “Maybe it’s the checkout page. Maybe it’s mobile users. Maybe it’s that new shipping banner.” Useful!

But if you never look at the data yourself, you lose the muscle. And at some point you won’t be able to tell whether the AI found a real insight...

or just made something up.

To be honest, that’s my biggest fear. It’s not that AI is bad. It’s that businesses get so comfortable letting it make decisions that nobody checks anymore.

Blind trust is way easier than thinking. It’s also way more expensive when it’s wrong.

3. More insights aren’t better insights.​
Here’s what I keep seeing with AI: it gives you too much. Twelve observations, eight recommendations, a few “opportunities” thrown in for good measure.

But in e-commerce you usually need one or two insights. Strong ones. Strong enough to test. If you can test it, you learn something.

That’s what actually moves your conversion rate. A 40-bullet report doesn’t.

4. AI still needs somewhere to get the data from.​
This one gets forgotten a lot. AI doesn’t magically know your numbers. Your data has to live somewhere it can actually reach.

If you’ve got a data team, that’s probably a data warehouse or a data lake (basically, one big central storage place for all your business data).

Your AI plugs into that, and GA4 is just one of many sources feeding it.

But most smaller e-commerce shops don’t have that luxury.

No data team. No warehouse. No budget for one either.

So what do they use?

Something easy. Something that’s already set up and collecting data for free...

Yep. GA4.

As much as I like to complain about it, for a lot of businesses it’ll stay the main place their data lives.

So here’s how I’m using AI with my data right now:

You first. AI second. Not the other way around.

And here’s the part I’m actually excited about.

I think AI will get more teams using data, not fewer. I talk to so many design, content and dev teams who want to use analytics. They just don’t know where to start, and dashboards feel intimidating.

AI could be their tutor. “What does this metric mean? Why did this drop? What should I check next?” Suddenly a designer can see where users get stuck. A content team can see what actually works.

So no, dashboards aren’t going anywhere. And neither are tools like GA4, at least for smaller teams. If anything, more people will be looking at them.

That’s my 2 cents this week.

And hit reply, I’m curious. Are you using AI with your analytics yet? Has it replaced any of your dashboards?

Happy analyzing,
Robert

P.S. If you wonder why you haven't heard from me for a loooong time, well... I had a kid. And I underestimated how much time I would have for anything else 🫣