Last updated: November 1, 2026.
The first thing people new to keyword research usually do is open an SEO tool. Semrush. Keyword Planner. Ubersuggest.
But I think in most cases that’s not the first place we should start.
Google itself is.
Because every day Google shows us the searches that millions of users actually make. And it’s completely free.
1. Google Autocomplete
As soon as you start typing “coffee machine” into Google, suggestions appear. For example:
- coffee machine recommendations
- coffee machine prices
- coffee machine cleaning
- coffee machine best brand
So what does Google base these on? Put simply, they’re searches people make most often and that Google thinks will be useful.
In other words, they don’t appear at random. They come from real user behavior. Google also says the predictions come from searches that have actually been made on Google. The language of the query, where it’s searched from, current interest in that query and the person’s past searches also affect the predictions (Google Search Help). But the predictions aren’t a ranked list of the most searched queries. That’s why I don’t read them like search volume data.
Google also doesn’t show some predictions at all because of its policies. The same help page explains that predictions that are violent, sexually explicit, hateful or potentially dangerous are blocked. So if a search isn’t in the suggestions, that doesn’t mean nobody searches for it.
When I’m researching a new topic, this is one of the first places I look. Because sometimes dozens of ideas I hadn’t thought of come up here.
2. People also ask
In many searches, you’ll see the People also ask box among the results. I think this is one of the most valuable areas on Google. Because it shows the other questions users ask about the same topic.
For example, a search for “coffee machine” might bring up questions like:
- What is the best coffee machine?
- What should I look for when buying a coffee machine?
- Drip coffee maker or espresso machine?
These questions actually give us new content ideas. Sometimes one of them can even be the topic of a new article on its own. I covered this box and the other areas of the search results page in Chapter 5.
3. Related searches
When you scroll to the bottom of the page, you’ll see the Related searches section. I think this area is very valuable too. Because Google is answering the question “What else do people researching this topic search for?”
Sometimes completely new keywords we never thought of come up here.
4. This is where we find long-tail opportunities
In Chapter 8 we gave the example of Trendyol, Turkey’s largest online marketplace: Trendyol → Trendyol integration → WooCommerce Trendyol integration.
We can already see most of these longer, more specific queries in Google’s suggestions.
Long-tail keywords are usually not words we sit down and try to come up with. They’re searches users are already making. We just notice them.
The name “long-tail” is a bit misleading. The “long” here has nothing to do with the number of letters. It’s about how specific the user’s search is. Sometimes even two words can be long-tail. espresso machine is two words, but it’s much more specific than machine. That’s why it’s better to judge long-tail not by word count but by how clear the user is.
When a user searches for coffee, we don’t know what they want. But when they say home espresso machine, they’ve given us much more information. The more detail the user gives, the more accurate the answer we can prepare for them. And Google’s suggestions show us these details ready-made.
How I look at long-tail
When I start a new project, I usually don’t start with the most general keyword. First I try to find keywords we can compete for, where it’s clear what the user wants, and that are directly related to our product or service. Because seeing the first wins in SEO both boosts motivation and, over time, makes it easier to compete for bigger keywords.
But not every piece of content has to be long-tail. Balance matters here too. If you target only very specific keywords, you may struggle to grow after a certain point. If you target only general keywords, the competition can be very high. I usually try to become visible on more specific topics first, then gradually move toward more general keywords. The growth story of many successful sites looks like this too.
Fewer people come from long-tail searches. But the people who come usually know what they want. The B2B client I described in Chapter 10, whose site gets 30-40 organic visitors a day and sells machines that cost around €20,000, is a good example of this.
What matters in SEO is not finding the most searched keyword, but finding the keyword that brings the right user.
Pushing the suggestions a bit further
Typing just the main keyword into Google and looking at the first four or five suggestions may not be enough. A few small techniques bring up many more ideas:
- Adding a letter to the end of the keyword. When you continue with “coffee machine a”, “coffee machine b” and so on, new suggestions appear for each letter.
- Adding question words. Searches like “how coffee machine”, “why coffee machine” and “which coffee machine” bring out the questions people are curious about.
- Adding something before the keyword. Like “home coffee machine” or “office coffee machine”. This is where you see how users describe the product.
- Expanding the questions. When you click a question in the People also ask box, new questions are usually added below it. Open a few questions and the list grows quickly.
Keep in mind that the suggestions change based on language, location and past searches. If you’re working for a different city, or don’t want your past searches to affect the results, you need to take this into account.
But is it right to turn every suggestion into content?
I think this is the most critical point. We don’t write content for every keyword just because Google suggested it.
First I ask myself these questions:
- Is this topic related to our business?
- Does the user’s goal fit us?
- Can we really create useful content?
- Should it be a separate page, or should we cover it within an existing page?
Because sometimes the suggestions are small enough to be answered within the same page. Not every suggestion means a new article.
Let’s look at an example
Say we sell coffee machines. Google gave us these suggestions:
- coffee machine cleaning
- coffee machine recommendations
- coffee machine prices
- coffee machine problems
It wouldn’t be right to turn all of these into separate product pages. If it were me, I’d think about it like this:
Category page
- Coffee Machines
Blog
- How do you clean a coffee machine?
- What should you look for when buying a coffee machine?
- How do you choose the best coffee machine?
So we can use Google’s suggestions not just as keywords but when building the site architecture too. I cover site architecture in detail in Chapter 18.
So will we never use SEO tools?
Of course we will. Tools like Semrush, Keyword Planner, Search Console and Google Trends are still very valuable.
But I usually like to start my research on Google. Because here I see the searches real people make. Then I validate these ideas with SEO tools. I explained how I read the volumes the tools show in Chapter 10.
Finding ideas in Search Console queries
If your site has been live for a while, you have another very valuable source of ideas: Search Console. The Performance report shows which queries your site appears for in search results. Google’s suggestions show the searches everyone makes. Search Console shows the searches Google associates with your site.
What matters most here are queries that get impressions but no clicks. To find them:
- Open the Queries tab in the Performance report.
- Turn on Total impressions, Average CTR (click-through rate) and Average position.
- Sort the list by impressions.
- Look at queries with high impressions, low clicks and an average position beyond the first page.
These queries can show that Google associates you with the topic but doesn’t see your page as a good enough answer. For example, if your coffee machine category page gets impressions for the query “how to clean a coffee machine” but no clicks, the user won’t find the answer they’re looking for there. That’s a good candidate for a separate blog post.
You can also use a regular expression (regex) in the query filter to separate out question searches:
how|why|what|which|whenThis filter shows the question searches your site appears for all together. I cover Search Console and regex filters in detail in Chapter 29.
There’s one more thing to know: Search Console doesn’t show every query. Very rare searches that aren’t made by more than a few dozen people over a two-to-three month period are left out of the table to protect user privacy. Google calls these anonymized queries. They’re included in the chart totals but don’t appear as rows in the table, and they’re dropped from the totals when you use a query filter (Google Search Central). Some very specific long-tail searches can fall below this threshold. So saying “nobody searches for this” about a search you can’t see in Search Console may not be right. Google’s suggestions and the People also ask box help fill this gap.
There’s another source of ideas outside Google: the questions customers ask before they buy. When the series was being published on X, a comment under this chapter said exactly this: seeing the Google query next to the customer’s real objection shows better what we should cover on the page. I explain how I bring the questions from the sales team into content in Chapter 15.
My usual working order
- I look at Google’s suggestions.
- I go through the People also ask section.
- I glance at Related searches.
- Then I back these up with Search Console and SEO tools.
This method shows me not just more keywords but also what people are really curious about.
I think this is the best part of keyword research. We’re not actually looking for keywords.
We’re finding the questions people ask.
And that’s exactly the starting point for creating good content in SEO.
Practice
- Type your main product or service into Google. Add letters and question words to the end and copy the suggestions into a list.
- In the same search, open a few questions in the People also ask box and add the new questions that appear to the list.
- In Search Console, sort the last 3 months of queries by impressions. Which 10 queries have high impressions and low clicks?
- Decide for each idea on the list: Will it be answered on an existing page, does it need a new page, or is it unrelated to your business?
- Next to each idea, write where you found it: autocomplete, People also ask, Related searches or Search Console.
- Sort your list from general to specific. Which are the first five searches you could compete for, where it’s clear what the user wants?
Previous: Chapter 11: What Is Keyword Difficulty? · Contents · Next: Chapter 13: How Do We Place Keywords on the Site?
This chapter was first published in Turkish on X (September 10, 2026).