How predictions are chosen
- Common and trending searches that start with what you typed
- Your location and language
- Your earlier searches, if search history is on
Google removes predictions that break its policies, for example violent, hateful, sexually explicit or dangerous ones, and predictions about people that could be harmful.
How to use autocomplete for SEO
- Keyword ideas: type your main topic and note the predictions; they are real phrases people search.
- Questions: add “how”, “why”, “can” or “best” to find questions to answer.
- Long-tail keywords: try each letter of the alphabet after your keyword for more variations.
- Brand searches: check what appears after your brand name, such as “reviews”, “prices” or “login”, and make sure you have pages that answer it.
Autocomplete doesn’t show search volume, so check promising phrases in a keyword tool before you prioritize them.
Autocomplete example
A garden centre owner types “raised bed” into Google. Suggestions include “raised bed soil mix”, “raised bed ideas” and “raised bed size”. The owner writes a guide for each. One suggestion, “raised bed liner”, shows a question the owner had not thought of. The owner adds a section on it and links to the shop category.
How to collect autocomplete ideas
- Open a private browser window so personal history does not skew the predictions.
- Type your main keyword and write down the suggestions.
- Add each letter a to z after the keyword to find more.
- Add question words: how, what, why, can.
- Group the ideas by search intent and decide which page answers each.
Autocomplete mistakes
- Treating suggestions as proof of high search volume.
- Targeting a prediction that shows only in your location.
- Ignoring negative predictions about your brand, such as “scam”.
- Trying to manipulate predictions with fake searches, which is against Google’s policies.
Autocomplete FAQs
Why do predictions differ between people?
Language, location, trending topics and search history all affect them.
How can I use it for keyword research?
Use the suggestions as a starting list, then check demand in a keyword tool. See keyword research and long-tail keywords.
Turning suggestions into pages
Group similar suggestions and give each group one page rather than one thin page per phrase. For example, “raised bed soil”, “raised bed soil mix” and “best soil for raised beds” belong on a single guide with sections for each.