What are people actually searching for?
Type a topic and get the real phrases search engines suggest as people type: the questions they ask and the ways they phrase a search. Free, from the source, no keyword database.
Add it and every idea is checked against your pages, so you see what you still need to write.
Queries search autocomplete for your seed plus every letter and question word, then merges and de-duplicates the results. Nothing is stored.
How this works
A paid keyword tool sells you a database of volumes and difficulty. This uses the free signal underneath it: the autocomplete endpoint, which returns the phrases people actually type. It runs the seed, then the seed plus each letter a-z, then each question word (how, what, why…), and merges everything. The questions are usually the most useful. They map to the exact queries to answer.
What the free keyword ideas tool searches
Type a seed topic and the tool asks search autocomplete for suggestions. It queries the seed, the seed plus each letter of the alphabet, and the seed after each question word such as how, what and why. Google is asked first; when that endpoint limits automated requests, another autocomplete source answers, and the result says which one supplied the phrases.
The answers are merged, lowercased and de-duplicated, with the seed itself removed. Suggestions that begin with a question word are grouped as questions; the rest are phrases. Up to 80 questions and 160 phrases are returned. A browser cannot call the suggestion endpoint directly, so the request is made for you, and the wide expansion is why a short seed still returns a long list.
How to read the ideas
These are phrases real people type, not a database of figures. There are no search-volume numbers and no difficulty scores, because those need a licensed data set this tool does not have. Treat the list as demand evidence and topic discovery: each phrase is a real query that somebody entered, which is the part that matters when you choose what to write next.
If you enter your site, each idea is also marked against your pages as covered, a near miss, or no page. That marking compares the significant words in a phrase with the words in your page URLs, so it is a fast signal rather than a full content audit. A near miss usually means a section to add, not a new page.
Common uses and why they matter
The question phrases are the most useful part. They map to the exact queries people ask, so they make natural headings and question-and-answer entries, and they often reveal the language a buyer uses rather than the language a seller uses. A question phrase that keeps appearing is a strong signal that this is how your audience frames the topic, which matters because a page written in the seller's vocabulary may miss the buyer's.
Because the list reflects live suggestions, it moves with the season and with current interest. Running it again later can surface new phrasing. The absence of a phrase is not proof that nobody searches it: autocomplete shows what is commonly typed, and results differ by country and language, so a list from one market may not match another. Run it in the market you serve to avoid a false picture.
Turn the ideas into content
Start with the questions you can answer well, then check whether any existing page already targets them. For a phrase marked as no page, decide whether it needs a new page or a section on an existing one. Write for the reader, then use the phrase where it is natural: in a title, a heading, and early in the page.
Group related phrases and pick one page per topic. Use keyword gap when you want a list built from your own sitemap, and on-page check to confirm a finished page uses the phrase it targets. For the wording of the title and description, use SERP preview to see how it renders, then improve the page that already has the strongest footing.
When to use related checks
Use keyword cannibalization if two pages start to target the same phrase, and content quality to review the page you write. The cannibalization guide covers the decision to merge pages, differentiate them or re-aim them, because combining two pages is not always right when they serve different stages of a task. Choose the fix that matches the intent of each page.
If you want to know which queries already bring impressions, that requires your own search performance data, which this tool does not read. Use it for discovery, then measure against the queries your site actually appears for. The two signals answer different questions, and this tool only does the discovery well. Use the first to find topics and your own query data to judge results.