Keyword research decides what every page on your site is trying to achieve. If it is wrong, even excellent content attracts the wrong people, or nobody at all.
I learned this on our own site. At one point our pages were getting search clicks and impressions, but almost no enquiries. When we looked at the queries, a large part of the traffic came from people who wanted to learn something, not people looking for someone to hire. The traffic looked good in a report and did nothing for the business.
Three things to clarify first
Before writing anything: who you want to reach, what words they actually use, and what they want when they type them.
The most common mistakes
- Looking only at search volume. A phrase with 100 monthly searches from people ready to buy is worth more than one with 10,000 searches from students.
- Using your own vocabulary instead of your clients'. A real example from Hungarian: the formal word for "application development" gets about 20 searches a month, while the everyday phrase for "app development" gets an order of magnitude more. Same service, very different audience size. Every language has cases like this.
- Choosing phrases that are too broad. "Website design" is a fight with everyone. "Website design for machinery dealers" is a phrase you can win, and the people searching it are closer to buying.
- Ignoring the competition. If the first page is all big brands and marketplaces, pick a narrower angle.
- Not looking at your own data. Search Console shows what people already search for when they find you. Pages ranking just outside the top ten are often the cheapest wins.
The four kinds of intent
- Informational: "what is SEO". They want to learn.
- Navigational: a brand name. They want a specific site.
- Commercial investigation: "best SEO agency for B2B". They are comparing.
- Transactional: "SEO quote". They want to buy.
Most enquiries come from the last two. Informational content still matters, but it should lead somewhere, not stand alone.
Step by step
1. Write down how your clients talk. Twenty or thirty phrases, the way a client would say them on the phone, not the way your brochure says them.
2. Expand. Look at "People also ask" and related searches in Google. Look at the questions you get by email.
3. Check volumes. Google Keyword Planner is free with an Ads account. Ubersuggest, Ahrefs or Semrush give more context. We use Ubersuggest data a lot for comparing languages and markets.
4. Group by intent. "Website price", "how much does a website cost" and "website cost" belong on one page, not three.
5. Assign each group to one page. If you do not have the right page, that is your content plan.
6. Prioritise. Start where intent is strong, competition is realistic and you already have some visibility.
What you end up with
A simple table: phrase, volume, intent, competition, target page. It becomes the backbone of your SEO for the next year. You stop guessing, and every page has a clear job.
If you want a second opinion on your list, send it to me. Keyword research is the first step of every SEO project we run, including international ones, where each language needs its own research, not a translation of the first.

