Keyword Research

I start with seed topics, not keywords. I ask what my audience actually cares about: problems they have, questions they ask. That comes from customer conversations, support tickets, forum browsing. Then I expand those seeds using Google's autosuggest, related searches, and People Also Ask boxes. I also feed seeds into Google Ads Keyword Planner to get volume estimates and alternative ideas. But volume alone is deceptive. For each candidate, I evaluate three things: search volume, [keyword difficulty](/keyword-difficulty/), and search intent. I use a difficulty metric from a tool like Semrush, but I also check the SERP manually. If top results are all big brand homepages or Wikipedia, I move on. If there are smaller sites ranking, there's an opportunity. I also look at the intent: are the results informational, commercial, or transactional? If I plan to write a guide but the SERP is full of product pages, the intent mismatch means I will not rank. I cluster related terms into topic groups. For example, "vegan dinner recipes", "easy vegan dinners", and "plant-based dinner ideas" all belong together. That group becomes one pillar page and supporting articles. I then map one [primary keyword](/primary-keywords/) and two or three secondary keywords to each URL. That mapping becomes my content plan. I also use [keyword analysis for SEO](/keyword-analysis-for-seo/) to refine the grouping and ensure coverage. [Tracking keywords](/tracking-keywords/) allows me to monitor how those terms perform after publishing. I always cross-reference with [SEO competition analysis](/seo-competition-analysis/) to find gaps competitors are neglecting.

Don't chase volume. I've seen people target keywords with 10,000 monthly searches, but the SERP is dominated by Wikipedia or major brands. You will not rank. Instead, prioritise terms with moderate volume but lower competition and clear intent. Don't target too many keywords on one page. I have done that and it dilutes the topic, confusing both users and Google. Pick one primary and a couple of related terms maximum. Another trap: ignoring intent. I once optimised a commercial keyword with an informational page. Zero conversions. Even after I fixed the content, the SERP still favoured product pages. That is why I now check [intent SEO](/intent-seo/) before writing a single word. Also, watch out for seasonal trends. A keyword might spike in December and disappear in January. Use Google Trends to check. Don't rely solely on difficulty scores; always inspect the SERP. A low difficulty score means little if the top results are from highly authoritative domains. And don't forget [long-tail keywords](/long-tail-keywords/). They often bring higher conversion rates because intent is more specific. For example, I ranked for "buy organic coffee beans online" with a small site, while "coffee beans" was impossible. Finally, stop when you have enough to create or optimise a page. I used to spend hours searching for the perfect low-competition term. Now I stop once I have a solid primary keyword and a few supporting terms. The goal is to write the page, not to perfect the keyword list.

I used to think keyword research was about finding high-volume terms and stuffing them into content. I was wrong. I stuffed keywords into headings and body text, but my pages ranked poorly and read terribly. Google's algorithms got better at recognising keyword stuffing, and my rankings dropped. That forced me to focus on relevance and intent instead. I also ignored search intent completely. I optimised for "best running shoes" with a blog post, but the SERP was all e-commerce and review sites. My page never gained traction. Once I switched to "how to choose running shoes" for an informational page, it started ranking. Another mistake: I overlooked competitor gaps. I now use SEO competition analysis to find terms competitors rank for but with weak content, and I build better pages. I also skipped keyword clustering. My site had isolated pages on related topics without internal linking. Clustering improved my topical authority and rankings. I also treated keyword research as a one-time task. I set it and forgot it. But search behaviour changes, and competitors enter niches. I now schedule quarterly reviews to spot shifts. I also misjudged long-tail queries. I thought they were too low volume to bother with, but I have since seen that long-tail keywords bring more targeted traffic with higher conversion rates. I misunderstood intent many times before I understood its central role. Also, I misinterpreted keyword difficulty scores. A low score does not guarantee ranking if the SERP is full of authoritative sites. Now I always verify with manual SERP inspection. I used to rely on outdated volume data, and I ignored branded versus non-branded terms. Finally, I wasted time searching for the perfect keyword list instead of writing strong content.