top of page

AI Search Glossary

Whether you call it GEO, AIO or LLM-O, this AI Search glossary for marketers will help you learn the generative search terminology that matters most. And we’ll let you know when the industry agrees on a name.

A

.

AI (Artificial Intelligence)

Technology that enables machines to mimic human intelligence. It transforms how businesses operate and customers interact with brands through intelligent automation and personalized experiences.

.

AI Benchmarks

Performance standards that measure how effectively AI models accomplish specific tasks compared to other models, helping marketers evaluate and compare different AI tools.

.

AI Overviews

Google's AI-powered search feature that generates instant answers by summarizing information from multiple websites, fundamentally changing how users discover and consume content.

.

Astro

Astro is Wix's AI-powered business assistant that transforms website management through intelligent automation, helping users optimize their digital presence for human visitors and AI discovery.

B

.

Bot-training

Automated programs that crawl your site to build the knowledge base that allows LLMs to learn language patterns, improve their training data, and generate human-like content.

.

Bot-user

Automated programs that fetch information from your site on behalf of an LLM user.

C

.

ChatGPT

OpenAI's breakthrough conversational large language model (LLM) chatbot that sparked the current AI revolution, fundamentally changing how people search for information and interact with digital content.

.

Citation

AI’s acknowledgment of the original source in the name of transparency and credibility.

.

Claude

An advanced large language model (LLM) developed by Anthropic that's reshaping how search marketers approach content creation and optimization strategies.

.

Common Crawl

A massive, publicly accessible archive of web data that provides researchers and developers with insights into internet trends, website structures, and digital content patterns.

.

Copilot

Microsoft's conversational AI assistant, deeply integrated with Bing search. It leverages LLMs and sources information directly from the web. Copilot in Bing was the first major search engine to use "grounding" for AI search.

.

Custom GPT

Specialized versions of ChatGPT tailored for specific tasks or industries, offering marketers powerful automation tools that maintain a consistent brand voice and expertise.

D

.

DeepSeek

A Chinese AI company that develops large language models and AI systems. They're known for creating competitive AI models that often punch above their weight in terms of performance relative to their size and training costs.

E

.

Entities

The real-world people, places, things, and concepts that search engines and AI systems identify and understand. Search marketers optimize for entities as well as keywords.

F

.

F-score

A measure of a model's accuracy that considers both precision and recall. Search marketers may use F-score to evaluate the performance of AI-powered marketing tools or algorithms.

G

.

GPT (Generative Pre-Trained Transformer)

The foundational architecture that powers many modern large language models (LLMs), revolutionizing how machines understand and generate human-like text across countless applications.

.

Gating

A smart routing system in AI that automatically directs different types of inputs to specialized sub-models or "experts" that are best equipped to handle each specific task. This makes AI systems more efficient and accurate.

.

Gemini

Google's multimodal AI system that can understand and generate text, images, audio, and video. Gemini's capabilities significantly affect how Google understands and ranks content, requiring search marketers to adapt their strategies.

.

Grounding

The process of connecting AI responses to verified, real-world information sources, ensuring accuracy and reliability in AI-generated content and search results.

H

.

Hallucination

When AI systems confidently generate false or misleading information and present fabricated content as factual.

I

.

Information Retrieval

The process of finding and extracting relevant information from large datasets. It’s the foundation of modern search engines and AI systems that connect users with the information they request.

J

.

JavaScript

The programming language that brings websites to life with interactive features. LLMs process Javascript in a manner that is distinct to classic search engine crawlers.

K

.

Keyword Research

The process of identifying the terms people use to search for information. While still important, keyword research is evolving as search becomes more conversational and entity-based.

.

Knowledge Cut-Off Date

The point in time after which an AI system lacks information about events or developments.

L

.

LLM (Large Language Model)

Sophisticated AI systems trained on vast text datasets to understand and generate human-like language. LLMs have revolutionized search and content creation in digital marketing.

.

LLM.txt

A proposed web standard that lets website owners tell AI models what content they should use.

.

Llama

Meta's family of efficient large language models designed for research applications. It demonstrates that high-performance AI can be achieved with smaller, more accessible architectures.

M

.

Machine Learning

AI that learns from data rather than following pre-written rules. It's the engine behind everything from search rankings to automated ad bidding.

.

Mention

When an AI chatbot or research tool references a brand or website without linking to it. Mentions are great for visibility, but they may not drive traffic.

.

Mixtral

An open-source language model from Mistral AI that uses Mixture of Experts architecture to deliver strong performance while remaining freely accessible.

.

Mixture of Experts

An AI architecture that uses specialized sub-networks to handle different types of tasks, making models more efficient and potentially more accurate.

N

.

NLP (Natural Language Processing)

The technology that helps computers understand, interpret, and generate human language. It's the foundation of modern search and AI communication.

O

.

Open Source LLMs

AI language models that are freely available to use, modify, and build upon. They offer flexibility and control that proprietary models can't match.

.

OpenAI

The AI research company behind ChatGPT. Their work has shaped how we interact with AI today.

P

.

Parameters

The internal settings learned during AI training that determine how a model processes information and generates responses.

.

Perplexity

An AI-powered answer engine that provides direct responses to questions with sources and citations.

.

Prompt Engineering

The skill of crafting effective prompts to get AI models to produce exactly what you need. It's becoming essential for marketers using AI tools.

Q

.

Query

The search term, question, or request you submit to a search engine or AI model.

R

.

Reasoning Model

AI designed to think through problems step-by-step, potentially revolutionizing how search engines evaluate content quality and relevance.

S

.

Sentiment Analysis

A metric that quantifies the emotional tone within text data. It moves beyond simply counting brand mentions in LLMs to provide a measurable value of how those mentions are framed.

T

.

Tokenization

How AI models break down text into smaller pieces (tokens) they can process. Understanding this helps optimize AI tool usage.

.

Training Data

The information used to teach AI models how to perform tasks. Quality training data directly impacts AI reliability and accuracy.

.

Transformer

A neural network architecture that is particularly well-suited for processing sequential data like text. Transformers are the foundation of many modern large language models (LLMs).

U

.

User Experience (UX)

How users feel when interacting with websites, apps, or AI tools. Good UX directly impacts search rankings and AI adoption.

V

.

Vector Embedding

A way to represent text as numbers that capture meaning, enabling AI to understand semantic relationships between different pieces of content.

W

.

Wikipedia

A free online encyclopedia that serves as a crucial training source for AI models, making it influential in how AI understands and presents information.

X

.

XAI (Explainable AI)

Efforts to make AI decision-making transparent and understandable, helping users trust and effectively use AI systems.

Y

.

Yield

A performance metric measuring the return or success rate of marketing efforts, campaigns, or investments.

Z

.

Zero-Shot Learning

AI's ability to perform tasks it wasn't specifically trained on, which means it adapts to new situations using existing knowledge.

Get SEO & LLM insights sent straight to your inbox

Stop searching for quick AI-search marketing hacks. Our monthly email has high-impact insights and tips proven to drive results. Your spam folder would never.

*By registering, you agree to the Wix Terms and acknowledge you've read Wix's Privacy Policy.

Thanks for submitting!

bottom of page