Low competition keywords: five places to find them and how to turn the list into a queue

A low competition keyword is a search phrase where the pages ranking today are ones your site can realistically outdo. It is not a fixed score, because it depends on your site. The fastest way to find them is to start from what you already have: queries where Search Console shows your pages on page two or three, questions people ask in your market in full sentences, and narrow phrases that stack a place, an audience or a use onto a broad term.

What "low competition" means for your site

A phrase is easy or hard only in relation to who is trying to rank for it. A new site and a ten year old site see the same results page and face different odds. So the useful definition is practical: low competition means at least one of the top results is a page you could beat with a better answer, and the page types that dominate are ones you can produce.

The step that decides whether a candidate is worth writing for is a manual read of the results page, which is covered step by step in our guide to SEO competition research. This page is about the step before it: how to generate a list of candidates worth reading.

Source one: queries you almost rank for

Open the Performance report in Search Console, described in Google's help for the report, and set the date range to the last three months. Sort by impressions and keep queries where your average position is between 8 and 30. These are phrases Google already connects to your site. A page that sits on page two for a query needs a stronger answer, a better title or a few internal links, which is far less work than starting from nothing.

Check which page is getting the impressions. If it is a page that only touches the topic, a dedicated page for that phrase is the move. If it is a strong page, improve it before writing a second one, since two of your pages chasing one phrase split the signal.

Source two: full-sentence questions

Short, broad phrases attract large sites. Questions written the way people speak do not. Collect them from these places.

• The suggestions that appear as you type a seed phrase into Google, with each letter of the alphabet and each question word added after it.
• The "People also ask" boxes on a results page.
• Threads on forums and question sites where your buyers talk, taking the title of the thread as the phrase.
• Your own inbox, sales calls and support tickets. Every question a customer asks twice is a page.

Keep the customer's wording, not a tidier version. The page that matches how people phrase it has the edge.

Source three: stack modifiers onto a broad term

Take a head term you cannot win and add one qualifier at a time. Each addition narrows the audience and thins the field.

Broad term	Plus a place	Plus an audience	Plus a use or platform
accounting software	accounting software for UK landlords	accounting software for freelance designers	accounting software that links to a bank feed
email templates	email templates for a school fundraiser	email templates for solo consultants	email templates for a shared inbox

These are illustrations of the pattern, not researched targets. Test any you like with the results page read before you commit.

Source four: results pages with weak pages in the top ten

Search your candidate phrases in a private window. Weak signals worth noting include a forum thread in the top five, a page that answers only half the question, a result that is years out of date, or a page on a site unrelated to the subject. Any one of these means Google is short of a good answer and will take yours. A page full of well-known publishers each covering the topic in depth is a different picture, and a candidate to leave for later.

Source five: new and changing topics

When a product launches, a rule changes or a tool is renamed, people search for it before good pages exist. Watch the news in your field, your suppliers' release notes and the changelogs of the software your buyers use. The first accurate, dated page on a changed subject often holds its position for a long time, provided you keep it current.

Turn the list into a short queue

1. Put all candidates in one sheet with the source, the exact phrase and the page you would write.
2. Merge phrases that are the same question worded differently. One page can serve all of them.
3. Run the results page read on each one and record the weakest result and why it is weak.
4. Write the pages in order of closeness to a buyer. A phrase that sits near a purchase decision beats one that draws curious readers.
5. Review rankings after six to eight weeks and promote the winners with internal links from your stronger pages.

A query with low monthly volume can still bring the best visitors, because a narrow question comes from someone who knows exactly what they want. Our note on how Semrush scores keyword difficulty shows how one tool turns links into a number, which helps when you sort a long list before the manual read.

Questions about low competition keywords

What is a good keyword difficulty score for a new site?

Calibrate against your own data: find the difficulty of queries where you already rank in the top ten and take the top of that range as your ceiling. Scores differ between tools, so use one tool throughout.

Does low competition mean low volume?

Often, yes, and that is useful. Narrow phrases have fewer searches and also fewer strong pages. Many small, precise pages add up to meaningful traffic.

Are free tools enough to find them?

Search Console, autocomplete, the "People also ask" boxes and your own customer questions cost nothing and surface the best candidates. A paid tool helps sort a large list.

How many low competition keywords should I target per page?

One main phrase and the reworded versions of the same question. If two phrases need different answers, give them separate pages.

How long until a page ranks?

Review at six to eight weeks and judge by impressions in Search Console first. Impressions rising from zero shows Google has placed the page.

Low competition in the answers assistants give

There is a second place to look for open space, and it has no volume column: the questions buyers type into assistants. The long, specific question about your field is where an assistant draws on a handful of pages, and where a clear page from a small site can be one of them. Ask ChatGPT, Gemini and Perplexity the narrow questions on your list and note which sources they cite; where the cited pages are thin or old, you have found an opening. AI Knows Us (aiknowsus.com) shows, for a set of questions about your business, which sources each assistant relied on and whether your pages were among them.