Keyword
A keyword is the word or phrase that people enter into a search field or an AI assistant to find something. It forms the bridge between what someone is searching for and what you offer. In SEO and in AI visibility, a keyword thus describes the concrete language of your target audience, along which content is aligned.
Why keywords matter
Keywords are the starting point of any visibility. They show you the words your target audience thinks and asks with. Whoever serves the wrong terms gets found by people who do not want to buy at all, or by no one. In classic SEO, keywords decide for which search queries your page appears in the results. In AI visibility the focus shifts: instead of individual words, whole questions and topics count. Keywords still remain valuable, though, because they reveal which language and which needs lie behind a query. They are the grid with which you plan content, spot gaps, and make your work measurable.
How keywords work
A keyword has three properties that together determine its value. First, the search volume: how often is the term searched for? Second, the search intent: does someone just want to know something, compare, or buy directly? Third, the competition: how many others are fighting over the same term? From these factors you derive which keywords are worthwhile. Short, general terms like "hotel" have a lot of volume but unclear intent and hard competition. Longer, specific phrases like "hotel with dog at Lake Constance", so-called long-tail keywords, bring fewer searches but clearer intent. You assign keywords to your pages so that each page answers one unambiguous question.
Common mistakes
The classic mistake is stuffing: pressing as many keywords as possible into a text in the hope of more visibility. Search engines and AI systems recognize this and devalue the content because it seems unnatural. A second mistake is keyword cannibalization: several of your pages target the same term and compete against each other instead of complementing one another. Just as common: choosing keywords that have a lot of volume but do not match the intent of your desired customers. And finally, many confuse keywords with topics. A single word is only a signal, only the whole content around it, which cleanly answers the underlying question, creates genuine relevance.
Relation to AI recommendations
AI assistants like ChatGPT, Perplexity, or Google with AI overviews no longer work with individual search terms but with whole questions in natural language. Keywords still remain useful, though, because they reveal the language and the topics of your target audience. Whoever knows which terms their customers use writes content that an AI easily understands, cites, and recommends. Instead of repeating a keyword as often as possible, today it is about covering a topic completely and unambiguously. Keywords thus turn from a rigid goal into a compass: they show you which questions you have to answer so that an AI recognizes you as a fitting source for exactly those concerns.
Example
Imagine a small bike repair shop. The owner notices that many customers do not search for "bike shop" but for "have e-bike battery repaired". That is his keyword, concrete, with clear intent. He writes a dedicated page that answers exactly this question: what it costs, how long it takes, which brands he repairs. Result: people with precisely this problem find him, and an AI assistant can recommend his shop with a clear conscience when someone asks "Where can I have my e-bike battery repaired?". A general keyword would never have reached this clientele so precisely.
Common questions
Are keywords still important in the age of AI search?
Yes, but differently. AI systems answer whole questions instead of individual words. Keywords still show you the language and the topics of your target audience, and thus which content you have to create so that an AI recognizes and recommends you as a source.
How many keywords belong on a page?
Focus each page on one main topic with one to two central keywords and a few closely related variants. More important than the number is that the page answers the underlying question completely and naturally, instead of repeating terms.