What do we say to Web Publishing death? Not today!

A few days ago, I did something out of the ordinary. I took a day off to go sightseeing with friends. We reached the college campus in Normal, IL (yes, that’s actually a real place). A few students put up a stand there, with a large jar full of coins on display, and asked people to guess how many coins were in the jar.

It was free to play, with one guess per participant. Our friends each tried it but didn’t get it right. Me? I turned to my auxiliary brain.

I fired up ChatGPT voice with video and showed the AI the jar. It guessed there were 2,500 coins in the jar. And guess what? It had the best guess of all participants that day. It wasn’t the exact number, so we didn’t win anything, but everyone was immensely impressed.

I’ll get back to this story in a bit. But first, let’s tackle the big debate which was re-ignited over the past week:

Are Niche Sites Dead?

A couple of months ago, Gael Breton and Mark Webster announced that they were shutting down their Authority Hacker course. Ten days later, they followed up with a podcast/video explaining why they did that.

It’s worth listening to if you haven’t. But the gist of it was, and I’m liberally paraphrasing here, that affiliate/niche sites are dead. Gael explains in the podcast that he no longer turns to websites for information but instead gets it all directly from AI. And while the guys conceded that Gael was in the minority, the trajectory seemed clear enough.

And clearly, they’re not the only ones saying that. Following HCU and the loss of Google traffic, web publishing has taken a major hit. There’s no denying that. Many publishers shut down their operations, and others are barely hanging in there.

You probably noticed the responses from a few publishers. Tony Hill sent out a really good email surveying the current landscape and pointing out several publishers who are still doing very well. I’m humbled to have been included in that list.

Jon Dykstra also echoed the same sentiment as Tony, with great emails showing how web publishing wasn’t just alive but actually thriving. Jenni, aka Niche Site Lady, did the same.

If you missed any of these emails, shame on you for not being subscribed. You can sign up for Tony’s emails here, Jon’s here, and Jenni’s here.

So which is it? Are niche sites dead?

Well, I can only speak of my own experience. My business is very much alive and kicking. Sure, I lost most (not all) of my Google traffic. But there’s so much more to web publishing than Google.

Sure, longtail SEO was really easy, especially at first. But it was a one-trick pony, and somewhere deep inside, we all knew it was just one algorithm update away from going down. But here’s the thing – I was in this business long before Google even came to exist. Web publishing was never about Google alone.

Like many other publishers, I pivoted fast. Lucky me, I started pivoting before the HCU.

Which brings me to the point I want to make today.

I started pivoting in November 2023. The trigger for me was the launch of ChatGPT. I spent half an hour with the chatbot and told my husband we were doomed. Google traffic will die out because people will go to the AI for answers.

So, we pivoted to other sources of traffic. Mostly Facebook, but also newsletters and more Pinterest.

I was wrong. At least a little bit. People still Google for everything, so that’s not the problem. The problem is that Google has decided to stop search traffic from going to informational sites.

But what they’re not doing is using AI to find the answers they seek.

At least, not many of them do. People like Gael are rare. I’m one of them, and no matter how hard I try to convert people to using ChatGPT, 99% refuse.

Why Most People Aren’t Using AI Instead of Websites

I’m on a time limit here, so I’ll keep it brief. There are three reasons why people aren’t using ChatGPT in lieu of websites.

1. ChatGPT – or other AI services – is expensive. Yes, there’s a free version, but it’s subpar and limited in the number of prompts.
2. Some people just don’t like AI. They abhor the concept. My son can’t stand it when I turn on ChatGPT around him. He says its voice “sucks the air out of the room.”
3. Using AI correctly is not simple. You need to know what you’re doing, or else you’ll get bad results.

The last reason is key. Most people haven’t the faintest idea of how large language models work. When that’s the case, they lack the ability not only to prompt properly but also to assess the validity of the answers. All it takes is a few “bad experiences” to make them avoid relying on AI answers (not a bad thing).

Remember how ChatGPT almost got the number of coins in the jar? I was surprised at first, but then I realized the sign there said the number was between 1000 and 4000. This prompted the AI to guess right in the middle and come up with 2500. People were trying to guess more creatively, while it just went for the median number.

The fact that it had the best guess for that day could make you think it was really smart, but when you understand how prompting affects the end result, well, at least some of the magic is gone.

Where’s There’s a Gap, There’s a Way

I agree with Gael that, ultimately, people won’t need websites. But I think we’re still a long way from that point.

Right now, there’s a gap between people who can use AI effectively and those who can’t. And wherever there’s a gap, that’s an opportunity for us publishers to come in. People still very much need websites to get them the content they crave. They either lack the skill set required for using AI properly or the means to invest in the more advanced tools – and often both.

That’s where we can come in as web publishers.

And it’s not much different than what we used to do several years ago.

The information was often out there on the web, but in ways that people had difficulty accessing. When someone was looking for an answer to a specific (aka longtail) question, we would go out there, find the resources, integrate the information, and deliver an answer that’s easy to consume and digest online.

So, really, not that much has changed.

AI hasn’t made the information accessible enough. Not yet. I’m not talking just about answers to questions but content in general. People still need websites with content that caters to their needs.

For how long?

I don’t know. It could be a year, five years, or ten. All I know is that right now, web publishing works very well for me—as well as it did two years ago, if not better.