In marketing, there’s a term for someone who shares information and influences other people’s decisions. They’re called a maven.
Think of that friend who seems to know everything about everything. They know which company you should hire, which product you should buy, and why you’re probably doing something wrong.
Sometimes they’re genuinely knowledgeable and helpful. Other times, they’ve read a few articles and suddenly consider themselves an expert. And occasionally, they have their own agenda.
Well, lately we’ve been dealing with a new kind of maven in the marketing world.
The Rise of Skynet…err, I mean ChatGPT
I can’t begin to tell you how many client meetings I’ve been in recently where someone pulls out a printout, screenshot, or email explaining everything ChatGPT thinks is wrong with their website.
There’s a few emotions I have to feel about this.
On one hand, I love seeing clients take an interest in their marketing. You’d be surprised how often we’re chasing down business owners for information we need to do the work they’re already paying us to do. So, when someone comes to a meeting prepared with questions, that’s actually refreshing.
And let’s give AI some credit. Sometimes it finds real problems. Things worth investigating or improvements we genuinely hadn’t considered.
But then there’s the other side.
When AI Becomes the Expert in the Room
We’re seeing more business owners use tools like ChatGPT and Claude to audit their websites, troubleshoot SEO problems, and tell them everything their marketing company should be doing differently. The problem isn’t necessarily that they’re using AI. The problem is that they’re treating its answers as facts.
A lot of people don’t understand how important context is when asking AI a question. They’ll type something like, “Analyze my website and tell me everything wrong with my SEO.”
Sounds reasonable, right?
But what information is the AI actually working with? Has it reviewed the live website? Does it know what keywords the business is targeting? Has it seen the site’s actual performance data? Does it understand the industry, competition, or what the marketing team has already accomplished?
Sometimes the answer to most of those questions is no. And even when AI has access to that information, it can still make incorrect assumptions. That’s how we end up having conversations like these:
“ChatGPT says we need another 100 to 150 citations to improve our local rankings.”
“Claude says our homepage needs another 1,000 words of content.”
Really? Says who?
Here’s the thing. Neither recommendation is automatically wrong. But neither one is automatically right. Could accurate, relevant business citations help your local visibility? Absolutely. Does that mean you need to submit your business to 150 directories, many of which nobody has ever heard of? Probably not.
Should your homepage explain what your business does, where you operate, and why someone should choose you? Of course.
Does that mean every business needs a 1,500-word homepage? No. Google doesn’t have some magical word-count requirement that suddenly makes a homepage rank better. And that’s the distinction that gets lost when people start treating AI recommendations like instructions. Just because something can be improved doesn’t mean improving it will make a meaningful difference.
The Internet Is Full of Outdated Advice
Here’s another problem.
Go search the internet right now, and you’ll find thousands of articles explaining how to improve your SEO. Some were written last month. Others were written five, ten, or even twenty years ago.
Some of that older advice is still perfectly valid.
A lot of it isn’t.
Search engines have changed. How people search has changed. Google’s understanding of websites has changed. And now we’re dealing with AI-powered search experiences on top of traditional search results.
Yet, AI can still surface recommendations based on information that’s outdated, incomplete, or simply irrelevant to your particular business. And here’s what makes this especially frustrating: AI can deliver questionable advice with exactly the same confidence as good advice. It doesn’t always tell you, “Hey, I found this recommendation in an article from 2012, and I’m not entirely sure it still applies.” It just gives you another item for your marketing agency’s to-do list.
Suddenly, we’re spending time explaining why we don’t need to implement a recommendation instead of discussing the things that could actually move the business forward.
Research Is Great. Context Is Better.
We’ve talked with doctors who genuinely appreciate when patients take an interest in their health and come prepared with questions based on their own research. But there’s a reason you still consult a medical professional. Reading about symptoms doesn’t give you the experience or context needed to make a proper diagnosis. Marketing obviously isn’t medicine, but the same principle applies.
You can research SEO, AEO, GEO, local search, or whatever acronym someone comes up with next. You can ask AI hundreds of questions and get hundreds of answers. That doesn’t mean every answer applies to your business. A marketing professional should understand not just what could be done, but what should be done, what’s already been tried, and what is most likely to make a difference.
Should You Stop Using AI for Marketing Advice?
Absolutely not. I use AI myself. It’s an incredibly useful tool for research, brainstorming, troubleshooting, and finding things that deserve a closer look. But that’s exactly what it is. A tool. Not a replacement for experience, current data, or an understanding of your business. Because not every technical issue is an SEO emergency. Not every missing keyword needs another paragraph. And not every suggestion deserves a place on your marketing budget.
In conclusion…
Just kidding. That’s probably how this article would end if I’d let AI write the whole thing. My point is simple: Use AI, learn from it, but do not rely solely on it. Don’t confuse a confident answer with an informed one. The goal isn’t to do everything ChatGPT recommends. It’s to do the things that actually help your business grow.
– Steven Williams, VP of Technology