Keyword Research from Scratch: A Practical Guide That Skips the Fluff

A friend launched a cooking blog and asked me to help with keywords. She'd picked "healthy recipes" as her main target. Monthly search volume: 1.2 million. Her domain was two weeks old with zero backlinks. That keyword choice was going to waste six months of her time.

Most keyword research guides start with "open your keyword tool and enter a seed keyword." That's step 5, not step 1. The real starting point is understanding what you can realistically rank for, and that depends on your site's current strength.

Before You Touch a Keyword Tool

Check your domain's current situation. How old is the site? How many pages are indexed? Do you have any backlinks at all? A brand new site with no authority needs to target keywords that an established site wouldn't bother with. That's not a disadvantage — it's a strategy.

We've seen new sites rank on page one within 4-6 weeks by targeting keywords with 100-500 monthly searches. These aren't exciting numbers, but 20 pages each bringing 15-30 visitors per day adds up to 300-600 daily visitors. For a new site, that's excellent.

The mistake is chasing volume. The skill is finding specificity.

Finding Keywords People Actually Search For

Start with Google itself. Type your topic into Google and look at three things: autocomplete suggestions (what appears as you type), "People also ask" boxes, and related searches at the bottom. These are real queries from real people. No tool needed.

For our cooking blog friend, "healthy recipes" autocomplete showed "healthy recipes for weight loss on a budget," "healthy recipes with chicken breast," "healthy recipes for picky eaters." Each of those is more specific, less competitive, and targets someone with a clear need.

Mine your existing data. If your site has any traffic at all, Search Console's Performance report is gold. Filter for queries where you appear on page 2 (positions 11-20). These are keywords Google already associates with your site but hasn't fully committed to ranking you for. A focused content improvement or a new dedicated page can push these to page one.

Check forums and communities. Reddit, Quora, niche Facebook groups, and industry forums show you the exact language your audience uses. SEO tools show you search volume. Communities show you pain points. The best keywords target pain points with decent search volume.

Evaluating Whether You Can Actually Rank

Search volume means nothing if you can't compete. Here's a quick manual check that works better than most difficulty scores:

Google your target keyword. Look at the top 5 results. If they're all from sites like Forbes, Wikipedia, HubSpot, or major brands, move on. If you see forums, small blogs, Reddit threads, or thin content in the top 5, that's your opening.

Also check: are the top results actually answering the query well? Sometimes the entire first page is mediocre content that happens to rank because nothing better exists. Write something genuinely better, and you have a real shot.

We ranked a client's page for "best invoicing software for freelancers" within 8 weeks. The top results were mostly listicles from affiliate sites that clearly hadn't tested anything. Our client had actually used five invoicing tools for their freelance work. That real experience showed in the writing, and Google rewarded it.

Long-Tail Keywords Are Not a Consolation Prize

There's a perception that long-tail keywords are what you target when you're not good enough for short-tail ones. That's wrong. Long-tail keywords convert better because they carry clearer intent.

Someone searching "shoes" could want anything. Someone searching "waterproof hiking boots for wide feet under $100" knows exactly what they want and is probably ready to buy. If your page is the one that answers their specific query, you'll get a visitor who is already 80% convinced.

We find that pages targeting long-tail keywords with 50-200 monthly searches often outperform pages targeting broader terms with 5,000+ searches — in actual revenue and conversions, not just traffic numbers.

Organizing Keywords Into Content Plans

Once you have a list of potential keywords, group them by topic clusters. One pillar page covers the broad topic. Supporting pages cover specific subtopics and link back to the pillar.

For a site about project management:

Pillar page: "Project Management for Small Teams" (targets medium-competition term)

Supporting pages: "How to Run a Sprint Planning Meeting," "Free Project Management Templates," "Asana vs Monday.com for 5-Person Teams," "How to Track Project Hours Without Micromanaging"

Each supporting page targets its own long-tail keyword while strengthening the pillar page through internal links. After 3-4 months with consistent publishing, we've seen pillar pages jump 20-30 positions as the cluster fills out.

The Tools We Actually Use for Keyword Research

Google Search Console for existing keyword opportunities. Ubersuggest's free tier for quick volume checks and related keyword suggestions. Google's autocomplete and "People also ask" for question-based keywords. AnswerThePublic when we need a broader view of questions around a topic.

If budget allows, Ahrefs or Semrush for competitor keyword analysis. Seeing what your competitors rank for — and what they don't — is where paid tools really shine.

Common Mistakes We Still See

Targeting one keyword per page and ignoring related terms. A single page can rank for dozens of related keywords if the content is thorough enough. Don't create separate pages for "how to do keyword research" and "keyword research process" — they're the same intent.

Ignoring search intent. If someone searches "keyword research," they want a guide. If they search "keyword research tool," they want a tool or a comparison. If they search "keyword research service," they might want to hire someone. Same base keyword, three completely different pages needed.

Refreshing keyword research too rarely. Search behavior shifts. We review our keyword targets quarterly. Some keywords that were easy 18 months ago now have serious competition. Others that seemed impossible have become winnable because the old content ranking for them hasn't been updated.

Keyword research isn't a one-time task. It's an ongoing practice that shapes every content decision you make. Start with what your audience actually needs, validate that enough people search for it, confirm you can compete, and then write something genuinely useful. That process hasn't changed in ten years, even as the tools around it keep evolving.