Using AI for SEO: Why You Still Need a Strategist

What Happens When AI Does Your SEO

Recently, I was teaching a “digital team member” how to do optimize a single page of a website with the eventual goal of offering this to clients. (Spoiler alert: I think SEO still requires a strategist at the helm). The idea was simple. Ask the “Digital Team Member” to connect to Google Search Console, pull the site data, and identify a page worth optimizing.

So I ran the “skill” on my own site first.

The “Digital Team Member” pulled sitewide data for drivetraffic.ca and came back with a great first suggestion. The biggest opportunity was a blog post on Google Ad Grants. Over three months, it had earned 623 impressions in Google search results, but only 3 clicks, a 0.48% click-through rate. For a topic that isn’t particularly competitive, that number is low — a listing with a well-matched title and meta description on a low-competition query should earn somewhere closer to 3-5% CTR. The “Digital Team Member” recommended updating the title and description to improve that rate.

Theoretically updating the title and description made sense, but I’m rarely a straight yes or a straight no gal — there’s almost always a better answer somewhere in between, and finding it means asking more questions.

The First Redirect: What the Numbers Don’t Know

My response was, “people searching for ‘Google Ad Grants’ aren’t necessarily looking to hire someone. They could be looking for information on the topic. The “Digital Team Member” was solely focused on the volume of impressions rather than intent of visitor to the page.

There’s a difference between someone who wants to know what a Google Ad Grant is and someone who wants to pay a consultant to manage one. Both might find that page; one of them is more likely to become a lead than the other, and that is the one we should focus on.

I asked the “Digital Team Member” to go deeper: pull the queries already driving traffic to that page, so we could see what visitors were actually searching for before deciding which direction to take.

They came back with a longer list. Most of the traffic was informational, as I suspected, but buried in the data were a handful of relevant phrases: “Google Ad Grants consultant Canada,” “Google Ad Grants expert Canada,” “Google Ad Grants Management Services, nonprofits, Canada.”

Those are the ones we should be chasing, even if they have a lower search volume. The “Digital Team Member” had the data all along but it took someone who understood the business to know which part of the data mattered.

The Second Redirect: Site Knowledge the AI Didn’t Have

Once we had the right keywords, the next step seemed obvious: optimize the blog post to target them. I said no again.

I knew that DriveTraffic already had a service page on the same topic: Google Ad Grants management for nonprofits. The blog post and the service page were competing with each other for the same searches (“Keyword cannibalization”). This is a common problem that’s easy to overlook when you’re looking at one page at a time, so I knew that I had to train the “Digital Team Member” to look into this when suggesting which page to optimize and how to perform the optimization. Optimizing the blog post would have made the cannibalization worse, not better. The right move was to rewrite the service page, make it the authoritative destination for anyone looking to hire a consultant, and redirect the blog post to it.

The “Digital Team Member” would not have caught this on their own. They were analyzing a page, not a site. The decision to restructure required knowing what else existed, what the business was actually trying to sell, and which page was the right home for a commercial audience.

What This Actually Shows About AI

The “Digital Team Member” was useful throughout the session; they connected to Google Search Console, via a connector in Claude.ai, pulled the data, surfaced the opportunity, dug into the query-level detail when I asked, and helped draft the updated service page copy.

But the two decisions that shaped the outcome, which keywords to prioritize and which page to optimize, came from an SEO Strategist who knew the business, the site, and what kind of client we’re trying to reach.

The on-page optimization cheat sheet I share with clients covers the tactical side: how to structure a page, what to include in your title and meta, how to signal relevance to Google. That part is learnable. Knowing when to apply those tactics, and where, is the part that takes a strategist.

If You Want to Follow Along on My AI Journey

You may have already guessed it: my “Digital Team Member” is an AI agent. I’m building this “team” out over time: new skills, new capabilities, and hopefully, eventually tools that business owners can use in conjunction with an SEO strategist.

If you want to follow along as I build this out, sign up for the DriveTraffic newsletter and I’ll share updates as they happen. No pitching. Just valuable information. While working on this blog post, I received an email from a previous client telling me were approved for a grant  they applied for thanks to my mentioning it in the most recent newsletter!

Want to see the before and after?

I saved snapshots of both pages before we made any changes: the original blog post and the original service page. If you’re curious what we were working with, you can download them below.