Evolving Keyword Targeting into Prompt Targeting
In the 25 years or so search engine optimization has been around there have been a lot of changes. However, none more drastic than the AEO/AIO transformation. This AI/LLM transition has reached all facets of every industry including digital marketing. Focusing just on SEO, it has prompted behavioral change that has influence SERP changes. Perhaps the biggest change in the way we do SEO is keyword research and content generation.
Old school Keyword Research and Content Generation
Keyword research used to be pretty straight forward, you would consider your products or services and look at your competitors and do a little ven diagram. Pretty simple and effective for a while. Take that to the next step, you would revolve your content around those keywords and also create some content to illustrate why you were better/different than the competitor. Keyword libraries would dictate everything on your website that you are selling or offering. Hub pages would often have a topical primary keyword and secondary keywords that page would try to target. This way of research is still relevant today but it has transformed into prompts.
Evolution of machine learning and NLP
Since the inception of SEO, queries have gotten more and more accurate and we usually don’t even think about it. This is due to machine learning, specifically NLP (natural language processing.) Without getting too into the weeds, which is worth reading, Google uses NLP to accurately show query results that match intent as closely as possible. Now, as we enter the AI era things are changing. Its not about targeting keywords but more so ideas or themes rather than on specific phrase.
SERP Changes
Pulled directly from a AI Overview:
“Google’s AI Overviews drastically reduce traditional organic website traffic. Recent studies show these summaries appear in up to 58% of searches, pushing organic click-through rates (CTR) down by 34% to 61%. Users now frequently consume the AI’s instant answer and leave without clicking any external links (zero-click search).”
This has prompted SEOs (pun intended) to change their strategy when targeting “keywords.” All of the sudden instead of targeting short tailed or long tailed keywords, you are focusing on targeting complete thoughts and ideas.
What is the difference between keywords and prompts? Keywords are phrases that are design to match intent in their result. They can be long tailed or short tailed keywords, the goal of this was the drive clicks. In the past Google guessed the intent of the query based on the keyword and any context surrounding it. Prompts are more wholistic. Essentially Google is throwing the kitchen sink at you with this new way of generating results from queries. An LLM result will give you:
- The main topic and the context around it
- Supporting themes and additional notes
- Specific terms or concepts readers expect to see
- The desired structure, tone, or level of expertise
How to begin thinking of targeting prompts instead of a keyword?
Here is the trick, although the search landscape has changed with AI overviews, and keywords have evolved into prompts, etc. I am going to say something unpopular but I think it is true. Writing has not changed!! A lot of SEOs will say you have to cluster your topics together at the top of the page. You have insert more FAQ style writing. You have to include bullet points (which was good to do anyways). This is a load of malarkey.
Machine learning, NLP, BERT, and all the other technology companies have come up with to decider meaning and intent of OUR writing is because we write like humans because we are humans. The one thing I have always agreed with Google on is this: We are supposed to write original, informative, pieces of content that fits the goal or narrative of why we are writing. If we are writing a blog post, I want that piece to be informative and interesting. Something I would want to share.
“Content that ranks highest is people-first content: highly readable, deeply authoritative, and heavily based on first-hand expertise rather than generic AI generation. Search algorithms consistently favor authentic, comprehensive material that directly satisfies the user’s search intent.” – Google Developers
Don’t try to manipulate your content just to get featured in an AI overview. If your content is ranking or capturing any AI overviews maybe it is a problem with your internal linking, or your content is thin and it sucks, or schema markup is absent.
How to generate a proper prompt
Think back to fourth or fifth grade. Think about what a good sentence entails. You have a verb, a noun, an adjective and maybe an adverb
A good prompt can change the results of what you are looking for significantly. So you have to ask yourself when you are writing your content. What do I want this piece of content to show up for when people are thinking of prompts to ask an LLM. That is how I think of it anyway.
“When shaping a prompt, the goal is to make sure the main phrase shows up naturally, and the surrounding context supports it.
A useful prompt usually includes:
- The primary keyword and the angle you want
- A few variations or close synonyms
- Related terms that readers expect to see
- Any modifiers that sharpen the topic (for example, “for beginners,” “for teams,” or “in 2026”)”
What do I do if my content isn’t showing up in LLMs?
Call West Loop SEO to come help add some depth and update/reshape old content of yours.
