The most common new website SEO mistake is targeting keywords that established sites have spent years dominating. You publish an article on a competitive topic, it sits at position 87, gets zero clicks, and you conclude SEO does not work. It worked — you just aimed at the wrong target. Here is exactly how to find the keywords a new site can actually rank for, using entirely free tools.

Quick Answer: New sites should target long-tail keywords (three or more words) where the current top results include small specialist sites rather than major brands. Use Google autocomplete and People Also Ask to find specific questions, then validate them in Google Keyword Planner. A keyword is winnable for a new site when the top results are not dominated by Forbes, Wikipedia, or major national brands.

Step 1: Understand why keyword difficulty matters so much for new sites

Every keyword has an existing set of competitors occupying page one. Those competitors have accumulated backlinks, authority, and ranking history over months or years. As a new site, you are starting from zero authority and competing for the same positions.

You cannot outrank a site with 500 backlinks from high-authority publications when you have zero. But you absolutely can outrank a small specialist site with ten backlinks from mediocre sources — because at that level, content quality and on-page optimisation can overcome the authority gap.

The goal: find keywords where the current competition is at your level or below.

Step 2: Start with seed topics, not seed keywords

Write down five to ten topics your website is about — not specific keywords, just broad subject areas. For an SEO tools site: broken links, site speed, backlinks, keyword research, technical SEO. These are your starting points for discovery, not your final keyword targets.

Step 3: Use Google autocomplete to find long-tail variations

For each seed topic, open an incognito browser window and type your seed topic followed by a space. Watch every autocomplete suggestion that appears. Then try these patterns:

  • Type your topic + "for" → "broken links for ecommerce", "broken links for beginners"
  • Type "how to fix" + your topic → "how to fix broken links on wordpress"
  • Type "what is" + your topic → "what is a broken link in SEO"
  • Type your topic + "without" → "check broken links without a tool"

Each autocomplete suggestion is a confirmed real query that people are searching. Write down every specific, multi-word phrase you find.

Step 4: Mine Google's People Also Ask boxes for question keywords

Search each of your seed topics and look at the PAA box in the results. Every question there is a confirmed high-frequency search query. Click each question to expand it — three to four more questions appear. Click those to generate more.

In three rounds of expansion from one seed topic, you can collect 20–30 specific, long-tail question keywords. These are perfect for new sites because questions are often specific enough to have low competition while still having meaningful search volume.

Step 5: Validate volume with Google Keyword Planner

Take your list of long-tail keywords to Google Keyword Planner (free at ads.google.com — you need a Google Ads account but do not need to run any ads). Enter your keywords and check monthly search volumes.

Filter for keywords showing at least 100 monthly searches. Keywords below 100 monthly searches are still worth targeting if they are highly relevant to your business, but prioritise those with confirmed search demand.

Step 6: Manually evaluate the SERP competition for each keyword

This is the most important step and the one most people skip. For each promising keyword:

  1. Search it in Google in an incognito window
  2. Look at positions one to five — who is ranking?
  3. If you see Wikipedia, Forbes, BBC, or major brands in every position → too competitive, skip it
  4. If you see small specialist blogs, niche sites, or community pages → this is winnable
  5. Click through to the top-ranking content — is it well-written and comprehensive, or thin and outdated? Thin, outdated content is your opportunity.

Step 7: Sort your winnable keywords into priority order

By now you have a list of winnable keywords with confirmed search volume. Sort them by:

  • Commercial value to your business first — keywords whose searchers are more likely to become customers
  • Volume second — higher volume within your achievable range
  • Competition third — the easier the competition, the faster you will rank and see results

How many keywords should a new site target?

Focus on completing one topic cluster of five to eight related keywords before starting another. One finished cluster of well-optimised, interlinked articles builds topical authority that makes each subsequent article in that cluster rank faster. Spreading across twenty unrelated keywords produces no topical authority and slower rankings for everything.

Summary: How to find keywords a new site can rank for

  • Step 1: Understand that new sites need low-competition targets, not high-volume head terms
  • Step 2: Start with broad seed topics, not specific keywords
  • Step 3: Use Google autocomplete patterns to generate long-tail variations
  • Step 4: Mine PAA boxes for confirmed question keywords — click through three rounds
  • Step 5: Validate volume in Google Keyword Planner
  • Step 6: Manually check SERP competition — small specialist sites in top five = winnable
  • Step 7: Prioritise by commercial value, then volume, then competition ease

Continue reading: How do I write a blog post that Google will actually index and rank?