Vectors in SEO for search engines and LLMs or AI are a way that machines connect specific topics together (entity SEO) to determine what the output for a user should be. When any one of the entities and vectors change the slightest, the output changes because the intent of the user is now different. This includes a single word, a number, or when a new vector or entity is introduced.
What made sense for a paragraph based on the person searching a first time is now a bullet list when they add a word or short phrase. A sound clip and an image could have been the best result for one query, but with a change like asking how to make an animal noise vs. what noise an animal makes, an instructional video clip with a written how-to guide may be best because sound with visual and written instructions may help the user better.
Once you master this concept, optimizing for AI, LLMs, and SEO becomes easy. Here’s how I explain it to clients, at conferences, and when hired to do private workshops for other agencies and companies.
If we’ve met you know I am a pet person, so I like to explain vector SEO with pet examples.
If you have a dog, you know that dogs modify their behavior when its time to go for a ride in the car. But have you ever wondered how the dog knows what your actual intent is with the ride? It’s the same as vectors and SEO.
Vectors for a dog can be:
- Slight changes in your tone of voice
- Items in your hand before, while, or after you tell them they’re going for a ride
- Sounds that change like pieces of paper crinkling vs. only a leash and keys
- Your scent if the dog can detect hormones, sweat, and breath smells from breathing
- The body language you present or your movements
Dogs can be trained to detect changes in your behavior and smell which is similar to how psychological service dogs can rely on hearing, smell, and visual queues. These are equal to vectors that search engines and AI or LLMs use. The different vectors create a different output from your dog.
If you promise your dog a ride in the car and a pup cup, you have no reason to be anxious or nervous, but you may show a bit of annoyance because you’ll have to hold them back from jumping through the drive through window and you’ll need to clean up the splatter. Going for a ride to the park on the other hand keeps you happy because it is fun for them and you get your exercise in.
When its time for the vet, your dog has likely heard you rustling papers if you have to bring them in. Your dog may also sense you’re a bit nervous or stressed, especially if it is for a health reason. Even if you say “go for a ride in the car” and “pup cup,” you may be emitting scents and sweating.
If these are the same signals you emit each time you take your dog to the vet, you’ve created a series of vectors that signals to your dog they’re going to the vet, and also for a ride in the car with the potential for a pup cup. That is how your dog knows it isn’t only a pup cup trip.
These are the vectors:
- A bit of annoyance is the entity and vector combination that becomes associated with definitely going for a pup cup. The dog’s vector result or output is excitement and happiness.
- Pure joy and no hesitation are vectors that signal to your dog to know it is time to play and be outside producing the vector result excitement, energy, and joy.
- Stressed movements, the sound of paperwork, and sensing anxiety through smell and hearing can take the same exact words and producing a new vector result which is hesitation. Even if they’re going in the car and getting a pup cup, they are also going to the vet.
This applies to cats too. A surefire way to get most cats to come out of hiding are the toys, treats, and the foods they love. There are also vectors that change the speed at which a cat will come, or simply not come.
- Carriers that are only used to go to the groomer or vet can be associated with avoiding you and hiding. Your cat will hear the carrier or see it and may be less trusting when they hear the treats.
- If they became tired with a toy, the noise it makes may no longer help them come out from hiding, or at least not as fast. Now you need to introduce a new vector to get a more optimal result. The user adding SEO modifiers to their query because the first result in the search engine is related here.
- Cans of food can work, but if they’re not hungry you may need the sound of a treat like puree tubes, whipped cream, or crunchy treats. In this output the vectors are not associated with the need of the user, your cat, so no result happens (i.e. you do not get exposure.) You need to find a better way to explain there is a benefit in coming to eat the food you’re setting out, or for a search engine or LLM to display your content to a user by having the right mix of information (entities and vectors) on your page and website.
Vectors are how algorithms associate topics together to create a result that meet’s a person’s needs. When one thing changes, the format and content of the output can change. If you all of the sudden create content that is not in the topic entity, you’re changing your vectors and website relevance. It doesn’t matter if you’re making money now because when the algorithms understand this is part of your new entity, you may lose a ton of traffic for the future, if not all traffic because you’re now in a different topical realm.
This is why it is important to stick to the topics that matter to your customers and not chase “one off successes” just because you made money or saw traffic and exposure. As machine learning (AI) takes off, vectors and entities will become more relevant than ever. It is all about associating topics with the goal of proper information retrieval. If you’ve been practicing these techniques combined with creating useful content and properly sourcing your content vs. gimmicks, you’re likely already seeing the wins.
