AI Is Optimizing Your Email Marketing.
That Doesn’t Mean It’s Helping.

Marketing tip #1: If you’re regularly using a marketing tactic, such as email marketing, be sure to analyze your results ON A REGULAR BASIS.

What’s regular? That depends upon the frequency with which you use the tool. If you only send out email marketing once a month, you likely need 3-6 months of results to get the slightest inkling of meaningful results. But if it’s daily or weekly use, analyze your results more frequently.

Why do I mention this?

Our Marketing Tools Provide Lots of Reporting. But Is It Useful?

You likely whiz by your email marketing metrics every time you open the program to launch a new campaign. Your last email’s delivery rate, open rate, clickthrough rate, bounce rate, etc., etc. But what does it all mean?

How do you know if it’s successful?

Crickets.

All those statistics are nice. But do those numbers mean my efforts are working or failing?

Well, guess what?

I’m not going to bore you either, reciting XYZ statistics in LMN range mean “fantastic results” “middling results” or “not so great.”

That’s what is known as A-N-A-L-Y-S-I-S. 🙄

And, sure, I could sift through all those stats to analyze what is and isn’t working. (And clients do pay me for those services.)

But, why not improve upon hours of human analysis by testing out our AI tools to find out how well they analyze and crunch the numbers in nanoseconds? And will they have valuable recommendations?

Can AI Tools Provide Actionable Analysis?

I took my newsletter data and ran it through multiple tools—ChatGPT, Gemini, Claude, Perplexity AI—and asked them what I should fix. And honestly? They did exactly what they were supposed to do.

They gave me smart, reasonable, well-structured advice. How to:

• Improve click-through rates
• Strengthen calls-to-action
• Encourage sharing
• Segment the list

And they provided an array of analysis to go with its suggestions. But where it shined was in the follow up. Don’t take the first response as the gospel from AI results. Challenge it:

• “Is XYZ relevant if ABC isn’t bringing the expected results?” – Query AI further on its analysis.
• “You said these open rates were excellent, but then why have I not seen ________ result from my readers?” – Make it further defend its conclusions. Argue.
• “If I add in ___________________, will my results improve?” – Let it speculate. And then ask why it gave the response it did.
• “If I implement _________________ changes to my email marketing, when should I expect ____________ result?” – Make. AI. Work. FOR. You.

The Ai Advice Wasn’t Wrong. It Just Wasn’t Useful.

Here’s something nobody tells you about AI-generated marketing advice: It’s very good at being correct. It’s not very good at being relevant.

It’s not always relevant because AI doesn’t actually understand your business. It does understand patterns…extremely well. And most of those patterns are built on averages.

You don’t have an “average” problem.

On paper, my newsletter looked fine:

Healthy open rates.
Engaged readers.
Even direct replies.

If I followed the AI’s advice blindly, I would have spent even more time:

• Tweaking subject lines
• Experimenting with send times
• Adding more calls-to-action

Those are all reasonable recommendations. And they’re also a distraction because the real issue wasn’t any of those things.

The real problem was harder to see.

People are opening my emails… reading them… and then doing absolutely nothing.

Not because they weren’t interested. But because they didn’t need to.

I had already given them everything.

The insight.
The explanation.
The conclusion.

Clicking the link wasn’t necessary. It was optional.

And optional doesn’t convert.

AI May Unwittingly Lead You Astray

AI will happily analyze whatever you give it. But it won’t ask: “Are you even measuring the right thing?”

It sees your metrics are the problem. (Like a word problem. You know? Math!) Aside: I was terrible with word problems in school.

It doesn’t question whether your structure, intent, or strategy might be the issue instead.

So now you end up wasting gobs of time doing something dangerous: Improving things that don’t matter.

The Efficiency Trap

This is the real risk of using AI in marketing: It doesn’t make you worse. It makes you more efficient at whatever you’re already doing. Even if what you’re doing isn’t working.

So now instead of guessing slowly, you’re optimizing quickly. Just… in the wrong direction. Now what?

So What Is AI Actually Good For?

A lot, if you use it correctly. (I can’t emphasize this enough. Ask the wrong question, or in the wrong order, and you get useless answers.)

Here’s how I use it now for an email marketing analysis:

Dump the data in
Metrics, trends, campaigns.

Ask for patterns
“What stands out?”
“What’s underperforming?”

Ask for recommendations
Let it give you the obvious stuff.

Then do the part AI can’t do:
Challenge it.

Ask:

What’s missing here?

What is this assuming?

What if the problem isn’t in the metrics at all?

That last step is where the real insight lives.

Now that’s just on one platform. Here’s some broader advice for getting to the root of your analysis:

1. Start on your favorite AI LLM platform (Large Language Model – pop quiz!). Copy ‘n’ paste its analysis into a document. Now take your data AND the analysis to another platform. Tell it: “Here’s my data. And here’s what [platform name] recommended. Tell me what you agree and disagree with in its advice and why.” Rinse. Repeat. On other AI LLMs.
2. Challenge each platform’s assumptions and analyses. Make it explain its conclusions.
3. Context – You may be asking it to analyze your email marketing efforts. But also ask it to consider the context. How is it performing alongside your social media efforts? Traditional advertising? Events and webinars?
4. Curve Ball – “If there’s one major change you could make to my email marketing to make it wildly successful, without budgetary constraint, what would that be? And ask me any questions you need answered first to be the most accurate email marketer.” Don’t settle for easy answers or easy analysis. Press AI to work hard for your business.

The Uncomfortable Takeaway

AI is a powerful tool. But it has no sense of priority. And no instinct for what matters.

It has no understanding of your audience beyond patterns.

It will give you answers.

It just won’t tell you which ones are worth your time.

Final Thought

If you’re using AI to improve your marketing, you’re not wrong. Bravo for testing and challenging the new tools we have.

But if you’re not questioning its conclusions, you’re not really thinking—you’re outsourcing judgment. And that’s where things can (and do) go sideways.

Because the difference between “good advice” and “useful strategy” is context.

And that’s still your job.