Using a machine learning model trained on millions of keywords, we scrape live SERPs to work out what type of content you need to produce to satisfy the user's search intent.

Understanding the intent behind a keyword has become a hugely important topic to SEOs and digital marketers - how can you be visible for a keyword if the type of content you’re producing is not aligned with what a user wants to see?

Some keywords are explicit in their intent; usually, these are keywords with modifiers appended or prepended to them.

Examples include:

  • BUY used cars

  • Mailchimp VS Hubspot


The modifiers “buy”, “vs” and “specifications” give a very clear indication to the search engine what sort of content around “used cars”, “Mailchimp” and “iPhone 12s” you want to see.

But what does someone want to view when they search for something without any modifiers? What’s the intent behind a keyword like “CBD Oil”? In this case, it’s most likely “transactional”;

Surely if someone searches for this term it’s because they’re looking to buy CBD oil right?

This is what we mean by “keyword intent”.

At Keyword Insights we developed a similar, but slightly different metric. We call this “Keyword Context

When we talk about context, we mean “what is the contextual setting around this keyword?”. If you’re confused I wouldn’t blame you so let’s revisit our example of “CBD Oil”.

Our hypothesis is that the intent behind such a keyword is transactional. Indeed, using a few popular SEO tools intent classification tools our hypothesis is proved correct. All of them reported the intent of this keyword as "Informational".

Which makes sense. People searching for “CBD oil” will want to purchase CBD oil… eventually. Enter “Keyword Context”.

If you actually witnessed the SERPs for that keyword, you’ll notice a lot of the results tend to favour more “long-form” type and not “product pages”.

This is where our metric, keyword context comes in. For the keyword “CBD Oil”, Keyword Insights would show you this:

Basically, the intent behind the keyword is “transactional”, but the context of it, currently, is “Informational”.

We use the “Keyword Context” function, in parallel to the “Keyword Cluster” feature, to quickly find groups of keywords that should be bundled together and written about in a long-form way. In short, we’ll often be able to show you keywords that are inherently transactional in nature, but for which the context of which is informational.

Did this answer your question?