One of the first things we do when we start a new SEO or AI optimization campaign is we put together a list of topics you want to create content on. And often when I present this list to clients, the first thing they say is, can you show us how much search volume there is for each of these topics so we know which ones to prioritize? And then I explain to them why I don't think we should prioritize by search volume.
So I thought it'd be fun to do a video explaining why that is.
Now, it's not because the topic doesn't matter. Picking the right keywords to create content for is, in my opinion, most of the strategy in any SEO or AI optimization campaign, whether you're creating new content or optimizing existing content, the purpose is always to rank for specific keywords. And if you're picking the wrong keywords, you really are wasting your time.
There are three things that need to be true for a keyword to not be a complete waste of time. First, ask yourself, would someone who's searching for this in Google actually be interested in becoming a customer of your business? If the intent behind the keyword is not to do business with you, it's a complete waste of time.
Now there might be some fringe exceptions. Maybe you can get a ton of traffic if you rank for some related term and that might bring you some backlinks which helps you rank for some other terms, but for the most part, I wouldn't worry about those keywords.
The second is, can you actually rank for this keyword or is it too competitive? If Google is only showing brands that are 10 times your size or 100 times your size on page 1, this might not be the best keyword to start with. Writing content for a keyword you can't rank for is a complete waste of time.
And the third factor is, are people actually searching for this term? Now with large language models, the question is now, are people or large language models actually searching for this term? Because keep in mind, when someone goes to chat, t or Gemini types in a prompt asking for a recommendation, that large language model is then going to do a series of searches in a search engine.
The question is whether those searches are performed by a human or a large language model. But if no humans and no large language models are searching for a given keyword, writing content for that keyword is a complete waste of time.
So this brings us back to my client's question. Why wouldn't we want to see the search volume for each of the terms we're considering targeting and prioritize them based on that? Well, there are a few reasons.
First of all, there really is no reliable data in terms of search volume. We get some data from keyword tools. Like hrefs and Semrush, but that data comes from Google Ads and it's very limited.
It's very common to receive a large amount of traffic from a keyword that hrefs says gets zero search volume. And in terms of large language models, there's really no reliable data for search traffic.
And the other reason is even if we had reliable search traffic data, that's only one of three factors. We also don't have reliable data in terms of how competitive a given keyword is and in terms of whether or not someone searching for this keyword has the intent of doing business with you. All we have is our intuition.
If you're going to write an equation for how valuable a potential keyword is, it would be search volume multiplied by the chance we have to rank for this keyword, multiplied by the chance that someone searching for this keyword is interested in being a customer of your business.
But since we can't reliably get any of those variables, a lot of SEO's do default to search volume because it's easy. You can put at least the keywords in a tool and it'll spit out a bunch of numbers and it seems really scientific. The problem is once you have some numbers and you can sort this list of keywords from highest search volume to lowest search volume, you stop using your intuition, which in this case is much more valuable than the search traffic numbers.
But here's the secret. There actually is reliable data in terms of what keywords have the highest potential. The catch is that that data is only available to you after you've created the content.
And that's in Google Search Console. Once you publish the content, you can see what terms it's showing up for. You can see where you're getting impressions, even if you're not ranking high enough to get clicks.
I have other videos talking about how to utilize that data. The important point I want to make here is that your goal should be to start creating content as quickly as possible. Quickly create 10 or 20 pieces of content and then you'll have real data you can work from.
