Using AI to Grow a Motorcycle Blog: Three Hours Well Lost

I’m going to be honest with you. I almost didn’t ride last Sunday. The group ride fell apart through the week, the weather had been good all week instead of just Sunday like the forecast called for, and by the time I looked up from my laptop it was already the middle of the morning. But that’s actually what this post is about. Because those three hours I lost weren’t wasted. I was deep in AI work, and it changed how I’m thinking about using AI to grow a motorcycle blog and YouTube channel in a pretty significant way.

If you’ve watched my recent videos you already know I’ve been leaning on AI pretty hard. Titles, descriptions, channel direction, finding a niche, cleaning up noindex issues on the site. I’ve covered most of that already. But this particular Sunday session was different. It started as housekeeping and turned into something closer to a strategy overhaul.

The Sunday Morning Rabbit Hole

It started simple enough. I had finished up the noindex cleanup on waltinpa.com and I was looking for a small project. Something I could chip away at in spare time without having to sit down and tackle it all in one chunk.

What I landed on was the category structure on the blog. The site has been around for a while and the categories were a mess. Motorcycles was the parent category and I had something like 107 posts filed under it. The child categories existed but barely anything was actually in them. Motorcycle education, destination rides, motorcycle news. All the right buckets, just nothing sorted into them.

So I used AI to help me figure out which categories were worth keeping, which ones to cut, and how to restructure the whole thing. Once that was settled I went through and manually recategorized all 107 posts. That alone took a while.

The Detail I Didn’t Know I Was Missing

While we were working through the category cleanup, AI flagged something I had completely overlooked. Each category on a WordPress blog can have its own description, and that description is actually indexed by Google. You can drop a focus keyword in there, add some context, and give Google more to work with when it’s trying to understand what your site is about.

I had nothing in any of those fields. Blank across the board. So we went through and wrote descriptions for every category. Low-hanging fruit, and genuinely useful. Worth doing if you haven’t.

When AI Contradicts Itself and You Call It Out

Here’s where the session got interesting. After the category work I went back to tweak my YouTube title and description prompt. I’ve been refining that prompt over several sessions and it’s gotten pretty solid.

But while we were working, AI flagged one of my existing video titles as a weak example. The problem was that title was one it had generated using the exact same prompt we’d been building together.

I pushed back on that pretty directly. If this prompt produced that title, why is that title now being held up as a bad example?

What came back actually made a lot of sense. It said that when it first started working with me, WaltInPA read like a motovlog channel first with some gear and educational content mixed in. But after seeing the full content library laid out by category, it was identifying something different. It thinks there’s real potential for WaltInPA to become a go-to source for Pennsylvania motorcycle destinations. Rides, covered bridges, day trips. Maybe even useful to people in cars, not just riders.

So the advice changed. Before, it was telling me not to lean into location-specific titles because the search volume wasn’t there. Now it’s saying lean in hard. Write the post about riding to the Kennedy Covered Bridge or the Greisemer’s Mill Covered Bridge. Own that space. Build toward it intentionally.

What This Actually Means for the Channel and Blog

That shift matters because it’s not just a tweak to a title format. It’s a positioning change. It means the content I’m producing now, the destination rides, the Pennsylvania backroads stuff, all of that isn’t just filler between bigger videos. It’s the foundation of something.

I don’t know that I would have arrived at that on my own, at least not as quickly. Having three months of content laid out in front of a tool that can look at all of it at once and identify patterns is genuinely useful in a way that’s hard to replicate just by thinking it through yourself.

And look, I’ve said it before and I’ll keep saying it: if you’re running a channel, a blog, or any kind of content platform and you’re not using AI to some degree, you’re leaving a lot on the table. Not because it does the work for you. It doesn’t, and it shouldn’t. But it sees things you’re too close to see, and it asks questions you forgot to ask yourself.

Goofy Saved the Ride

I should mention that the Sunday didn’t end with me staring at a screen. Goofy texted me just as I was about to go cut the grass and asked if I wanted to ride. I did. We went out, it was a great time, and there’s a separate video covering that ride if you want to see it.

But the three hours before that? Worth every minute.

---

Ride With Me

If you’re getting something out of this kind of content, the best thing you can do is subscribe to the WaltInPA YouTube channel for more riding content. More Pennsylvania rides are coming, more behind-the-scenes channel stuff, and yes, more AI talk whether you’re ready for it or not.

If you know someone running a channel or blog who would find this useful, share this post with a friend who rides. It doesn’t have to be a rider. Anyone in the content space could use this.

And if you want to talk through any of this stuff, come hang out in the WaltInPA Discord where good people talk about riding. The conversation there is always worth dropping into.