Skip to content

An exercise in using AI for keyword research

Let’s bring together a few things we’ve talked about here often. Let’s see if we can use ChatGPT to help create better articles for our websites, and faster too. This will be a list of steps to try while reading them. Thanks to Matt Diggity for the ideas.

First of all, let’s warm up by asking ChatGPT to provide some keywords. Something like: “Give me 20 long tail keywords to target on a website about [blue widgets]”.

The results will show that the system understands the products or services under discussion. It might even come up with some good ideas straight away. If so, pick one out, or provide something better, and ask: “Create an outline for an article about [how blue widgets can be used in the aerospace industry]”.

We should now have something that could be developed into an article. An attractive title might give us a bit of direction. Ask: “Give me some interesting titles for an article about [how blue widgets can be used in the aerospace industry]”.

Next steps

Take this a stage further by asking: “Write a 100 word introduction to an article called [Widgets take off – the benefits of using blue widgets in aerospace]”.

What’s been missed out? Ask: “Provide a list of semantically relevant topics missing from this [Paste in your introduction]”. Did it come up with more ideas that ought to be included? Get the system to write a paragraph about one or more of these.

Now we should have an article coming together, and it will have taken just a few minutes. This can be the basis for something that can be edited for technical sense and accuracy by an expert, and for grammar and structure by someone else. The result will not be just any old article – it could well be one with real SEO potential.