Keyword research for blog posts is the process of finding the questions your readers type into Google, grouping them so each post answers one main question, and checking the results page to see what kind of article Google already rewards. It is different from researching keywords for a shop or a service page. Blog searches are mostly informational, so the best targets are specific questions and long-tail phrases, not the broad head term with the biggest number in a tool.
I do this the same way every time, and I’ll show it with one worked example from start to finish. The example is illustrative. I invented the blog for this post, and I’ve left out search volumes on purpose, because the method matters more than numbers that would be stale in a month.
What Makes Blog Keyword Research Different?
A blog post doesn’t sell. It answers. That’s the whole difference. So the keyword you want is the one a curious person types before they know what to buy, and it usually reads like a sentence: “why are my basil leaves turning yellow” rather than “basil plant”.
That changes three habits. I care more about the shape of the question than the volume. I treat a whole group of related questions as one post. And I let the search results decide the format, because Google has already tested what readers click.
If you need the broader starting point, my guide to seed keywords covers where the first broad terms come from. This post picks up after that, once you have a seed and need to turn it into actual blog posts.
The Worked Example We’ll Follow
Picture a small blog about growing herbs indoors. Its seed is “indoor herb garden”. The owner wants a pillar post on that seed and a set of supporting posts around it. Every table below builds on this one blog, so you can scroll back and check how each decision follows from the one before it.
Where Do Blog Post Questions Come From?
Real questions come from four places, and each one sees something the others miss. Table 1 shows what I pulled from each source for the herb blog.
Table 1: Question sources for the “indoor herb garden” seed (illustrative)
| Source | How I use it | Example questions it surfaced |
|---|---|---|
| People Also Ask | Search the seed, open 3–4 questions so the box expands, copy every new one | “What herbs grow best indoors?”, “Do indoor herbs need direct sunlight?” |
| Google autocomplete | Type the seed plus “why”, “how”, “can”, “best”, and each letter a–z | “indoor herb garden without sunlight”, “indoor herb garden for beginners” |
| Forums and communities | Search the seed on Reddit and niche gardening forums, read thread titles and top replies | “basil keeps dying on my windowsill”, “grow light for herbs worth it?” |
| Search Console queries | Performance report, Queries tab, filter for question words | Queries the site already gets impressions for but has no post about |
In the table above, look at the third column for the forum row. Those phrases are messy and personal, and that’s why they’re valuable. Nobody’s keyword tool would have handed me “basil keeps dying on my windowsill”, yet it’s exactly how a frustrated reader thinks.
Autocomplete deserves one note of caution. Google’s help page says autocomplete predictions reflect real searches but also says the feature doesn’t simply show the most common queries. So I treat it as proof that a phrasing exists, never as a measure of demand.
For an existing site, the Search Console row is the one I’d start with. Those queries already earn impressions for you. Here’s the short version of how I pull them:
- Open the Performance report and set the date range to the last three months.
- Open the Queries tab and add a query filter with a regex for question words like how, why, what and can.
- Export the list and mark every question that has no post answering it yet.
The regex itself is in my Google Search Console regex guide. Free. Fast. And it’s your own data, not an estimate.
Should You Keep the Long-Tail and Zero-Volume Questions?
Yes, most of them. On a blog, the long tail is where a new site actually wins, because the broad term is usually held by big publishers and the specific question isn’t.
Here’s the trap. It catches nearly everyone. You run the forum phrases through a keyword tool and half of them show zero. That number rarely means zero searches. It usually means the tool can’t see that far into the tail, or it has folded the phrase into a shorter cousin. I cover how to check demand anyway in my guide to zero volume keywords.
My rule is simple. If a question shows up in at least two of the four sources in Table 1, it’s real enough to keep on the list, whatever the tool says.
How Do You Choose One Primary Keyword Per Post?
This is the step most bloggers skip. It’s also why their posts end up competing with each other for months, while neither one climbs past page two because Google can’t tell which of them should own the query. Every post gets exactly one primary keyword. The close variants that mean the same thing become its secondary keywords, and they live inside that post as sections or FAQ answers.
Table 2 shows how I grouped the herb questions.
Table 2: Grouping questions into posts (illustrative)
| Post | Primary keyword | Secondary keywords folded in | Why these belong together |
|---|---|---|---|
| Pillar | indoor herb garden | indoor herb garden for beginners, how to start an indoor herb garden | Same broad “getting started” need |
| Post A | best herbs to grow indoors | what herbs grow best indoors, easy indoor herbs | One list answers all three |
| Post B | indoor herb garden without sunlight | do indoor herbs need direct sunlight, grow light for herbs | Same problem: not enough light |
| Post C | why is my basil dying indoors | basil leaves turning yellow indoors, basil keeps dying on windowsill | Same plant, same diagnosis |
Read the last column of Table 2 before anything else. Every row is grouped by the reader’s need, not by shared words. “Grow light for herbs” and “without sunlight” share almost no words, yet they’re the same problem.
When I’m unsure whether two questions belong in one post, I search both and compare the first page. If several of the same URLs rank for both, Google treats them as one need, and I write one post. That idea is the basis of SERP based keyword clustering, if you want the full method.
Does the SERP Want a Listicle or a Guide?
Before writing, I open the results page for each primary keyword and record the format that dominates. Five minutes, tops. It saves me from writing a 3,000-word guide for a query where Google clearly prefers a quick list.
Table 3: Reading the SERP format, right vs wrong (illustrative)
| Primary keyword | What the top results look like | Wrong format | Right format |
|---|---|---|---|
| best herbs to grow indoors | Numbered lists of herbs with short care notes | Long step-by-step guide | Listicle, one herb per H2 or H3 |
| indoor herb garden without sunlight | Explainers comparing light options | Product roundup | Problem-solving guide with a comparison table |
| why is my basil dying indoors | Troubleshooting articles organised by symptom | Basil history and varieties | Symptom-by-symptom diagnosis |
| indoor herb garden | Long beginner guides with setup steps | Short list | Full pillar guide |
Compare the “wrong format” and “right format” columns row by row. The wrong choice in each row isn’t a bad article. It’s a good article for a different search, and that’s what makes it hard to spot.
I also note what the top results all miss. For Post C, every ranking page might list symptoms, but none might show how fast basil recovers after you move it. That gap becomes my information gain, the one thing my post adds.
Turning the Keyword Into a Brief
A keyword alone is not an assignment. Not even close. The writer, even when the writer is me, needs the decisions from Tables 2 and 3 in one place. So each primary keyword becomes a short brief before anyone drafts, and the draft then follows my SEO content writing rules.
Table 4: The brief for Post C (illustrative)
| Brief field | Filled in for Post C |
|---|---|
| Primary keyword | why is my basil dying indoors |
| Secondary keywords | basil leaves turning yellow indoors, basil keeps dying on windowsill |
| Search intent | Informational, troubleshooting |
| Format from the SERP | Symptom-by-symptom diagnosis |
| First paragraph must say | The three most common causes, in plain words |
| Information gain | Recovery timeline after each fix |
| Links up to | Pillar: indoor herb garden |
| Links across to | Post B (light problems) |
Look at the last two rows of Table 4. They connect this one post to the cluster, which is the next step. My full content brief template has more fields than this, and you can copy it straight into a doc.
Building the Cluster Around the Pillar
Once every post has its primary keyword, the cluster almost draws itself, because the groups in Table 2 already tell you which post is the broad parent and which ones answer the narrower questions underneath it. Here’s the herb blog as a simple map.
[PILLAR] indoor herb garden
/ | \
[A] best herbs to [B] indoor herb [C] why is my basil
grow indoors garden without dying indoors
sunlight
\____________/
(B and C link across:
low light kills basil)
In the map above, every supporting post links up to the pillar, and the pillar links down to all three. The line between B and C is the one most people forget. It exists because the reader’s real problem crosses both posts, and a link there helps the reader more than any keyword does.
I record each of these assignments in a keyword map so no two posts ever chase the same primary. My keyword mapping guide shows the sheet I use. If you’d rather track the raw research first, the columns in my keyword research template fit straight in front of it.
The Mistake I See Most Often
Honestly, it’s targeting the pillar keyword with every post. A blogger writes three posts that all orbit the same phrase:
- “indoor herb garden tips”
- “indoor herb garden ideas”
- “indoor herb garden guide”
Three posts, one need. Google picks whichever it likes on a given week, and none of them ever settles into a steady position because the site keeps sending mixed signals about which page is the real answer.
Go back to Table 2. Each row there has a different reader problem in its last column. If two of your planned posts would have the same text in that column, they’re one post. Merge them before you write a word, because merging later means redirects and lost time.
The second mistake? Trusting difficulty scores too early. I check the SERP first and the score second, and I explain why in my post on keyword difficulty.
What the Worked Example Shows About Keyword Research for Blog Posts
Table 1 proved that the best blog questions come from sources a keyword tool doesn’t read, especially forums and your own Search Console. Table 2 showed that grouping by reader need, not by shared words, gives each post one clean primary keyword. Table 3 showed that the SERP decides the format before you write.
Table 4 and the cluster map tied it together. Each keyword became a brief, and each brief knew where it linked. That’s the whole method, and it works the same for a herb blog or a B2B blog.
If you want me to run this process on your niche, my keyword research service starts with 5 keywords free to try. Running a store rather than a blog? My pair guide on keyword research for ecommerce handles category and product keywords, which work very differently.
Frequently Asked Questions
How Many Keywords Should One Blog Post Target?
One primary keyword, plus a handful of close variants that share the same intent. In my grouping example, each post carried one primary and two or three secondaries. If you find yourself adding a keyword that needs its own section and its own answer, it probably deserves its own post.
Can I Do Keyword Research for Blog Posts Without Paid Tools?
Yes. People Also Ask, Google autocomplete, forums and Search Console are all free, and they cover the question-finding work. Paid tools help most with volume estimates and competitor data, which matter less for informational posts than people think.
Should a New Blog Target Long-Tail Keywords First?
In most cases, yes. A new blog rarely outranks established publishers on a broad term in its first months. Specific questions with a clear answer are where a new site can win early, and those wins make the pillar stronger later.
Do Blog Posts Need Different Keywords Than Product Pages?
Yes. Blog posts target informational searches, like questions and how-tos. Product and category pages target commercial searches from people ready to buy. Mixing the two on one page usually means it ranks for neither.
Last updated: October 2026 by Mizanur Rahman
