AI Can Do More for You. But Can It Help You Think Better? analyst blog

If you follow AI influencers, you’ve probably seen the same promise many times.

AI can write your SQL.

It can clean your data.

It can build a dashboard.

It can create a presentation.

It can summarize a meeting.

All of that is useful. But I think we’re sometimes looking at the value of AI from the wrong direction.

We keep asking:

“What can AI do for me?”

As analysts, I think we should also be asking:

“How can AI help me deliver more value?”

For years, we’ve focused on becoming better at producing analysis. We learn new tools, improve our technical skills, and find ways to work faster.

AI can accelerate a lot of that work.

But being faster at producing an answer doesn’t necessarily mean we’re making better decisions.

Consider what happens after you’ve spent several days working on an important business problem.

You’ve gathered the data, validated the numbers, investigated the trend, and eventually arrived at a recommendation.

At that point, the traditional AI conversation might be:

“Give my analysis to AI and ask it to create the presentation.”

Useful.

But I’d rather ask:

“What if I’m wrong?”

Give AI your analysis and your recommendation.

Ask it to challenge your assumptions.

Ask what evidence could contradict your conclusion.

Ask what alternative explanations might exist.

Ask what information you’re missing.

Then change the perspective.

“What would a skeptical executive challenge?”

“What would someone responsible for the operational impact worry about?”

“What would someone who disagrees with me say?”

You’re not asking AI to make the decision for you.

You’re asking it to challenge your thinking before someone else does.

And I think that’s where AI becomes much more interesting for analysts.

An analysis can be technically correct and still lead to a bad recommendation.

The data might be accurate, but the question might be wrong.

The trend might be real, but the explanation might not be.

A KPI might have improved, but the improvement might not actually matter to the business.

These aren’t data-quality problems.

They’re judgment problems.

And judgment is where analysts create much of their value.

I’ve started experimenting with AI in this way in my own work.

I’ve used it to onboard itself into my business context. I’ve asked it to review presentations from an executive perspective. I’ve used research agents to compare our performance with external benchmarks.

The common thread is that AI isn’t replacing my analysis.

It’s giving me another perspective.

That’s important because when we’re deeply involved in an analysis, it’s easy to become attached to our conclusion. We know how much work went into it. We understand the data. We’ve already explained the story to ourselves.

AI can introduce some healthy friction.

It can ask questions we didn’t think to ask.

It can point out assumptions we didn’t realize we were making.

It can help us explore the argument against our own recommendation.

That doesn’t make the analysis better automatically.

It gives us an opportunity to make it better.

I don’t think the future of the analyst is about competing with AI over who can write SQL faster.

If AI can handle more of the execution, that’s an opportunity.

We can spend more time understanding the business problem, exploring alternatives, challenging our assumptions, and thinking about what should happen next.

Because the cost of a bad recommendation can be much greater than the cost of a slow analysis.

A faster analyst can produce more answers.

A better analyst helps the business act on the right one.

So yes, let AI write the SQL.

Let it clean the data.

Let it build the first chart.

But before you walk into the meeting, give it your recommendation and ask one more question:

“What am I missing?”

Maybe the biggest opportunity with AI isn’t getting it to do more of our work.

Maybe it’s using AI to help us avoid being wrong when the decision really matters.

Start with Data, End with Value analyst blog.