When LLMs use synthetic queries to get product-specific info, they are looking for specific traits.
These traits are based on the user's prompt/query.
For example:
If I'm an enterprise SEO manager looking for SEO software, my query might look something like this:
"I'm an SEO manager in a B2B enterprise company. What is the best SEO software for {use cases}. I have a ${value} buy-in budget}.
So, some examples traits derived from this query can be:
- The user is an enterprise-level buyer
- The budget
- The use cases
Etc...
In the table below, you can se some synthetic query examples an LLM (this was from ChatGPT Deep Research) can generate to find "evidence" about product/user traits.
Thus, SaaS companies can do the following:
- Make a list of their audience/user segments
- Map out product traits to these user segments
- Map out product traits to use cases
- Make sure these are mentioned in their documentation, feature pages, changelogs, blog posts etc..
- Make sure these are also mentioned (contextually) in their off-site content.
As AI search will be getting more granular and personalized, it's important to think beyond standalone queries and dive into traits and attributes.
Cheers! ✌
seo aisearch saasseo
This post was originally shared by on Linkedin.