How to Do Keyword Research for a New Blog: A Beginner's Guide

You've written your first three posts. You've hit publish. And then you sit there, refreshing your stats page like a gambler at 2 a.m., watching a grand total of four visitors bounce off your homepage.

I know that feeling. My first blog got exactly 31 visits in its first month. Twenty-two of them were me. The problem wasn't my writing. It was that I'd spent three weeks writing about topics nobody was searching for. I had picked keywords based on what I found interesting, which is the single most common mistake new bloggers make.

Keyword research for a new blog is different from keyword research for an established site. You don't have domain authority. You don't have backlinks. You have nothing but a blank canvas and the ability to pick your battles carefully. That constraint is actually a gift, because it forces you to be precise.

How to do keyword research for a new blog without wasting six months

The core principle is simple: find keywords that real people search for, that you can realistically rank for, and that align with what you actually want to write about. Miss any one of those three, and you'll either waste time or burn out.

I learned this the hard way. My first blog post targeted "best productivity apps." Volume: enormous. Difficulty: brutal. I ranked on page 7. It got zero clicks in four months. Meanwhile, a post targeting "best productivity apps for freelance writers with ADHD" brought me 340 visitors in its first 90 days. Same topic, different specificity, completely different outcome.

Step 1: define your topic cluster before touching any tool

Before you open a single keyword tool, write down five to seven broad topics your blog will cover. If you're writing about home coffee brewing, your clusters might be: espresso machines, pour-over technique, bean selection, grinding, water quality, and gear reviews.

Why does this matter? Because keyword tools will drown you in data. Without a filter, you'll end up with a 400-row spreadsheet and no idea where to start. Your clusters are the filter.

Here's what I do: I open a blank document and write each cluster as a heading. Under each, I list ten questions a beginner in that space might ask. Not keyword-researched questions. Just human questions. "How fine should I grind for a V60?" "Do I need a scale?" "Why does my espresso taste sour?"

That list of sixty or seventy questions becomes my research foundation. Everything that follows is about validating and prioritising those questions.

Step 2: use Google Keyword Planner as your starting point

Google Keyword Planner is free. It's not the most elegant tool, but it gives you something no other free tool does: actual data from Google's own search volume. You need a Google Ads account to access it, but you don't need to run any ads.

Enter one of your seed questions, and the tool returns related keywords with average monthly searches. Look for terms with 100 to 1,000 monthly searches. Below 100, the keyword might be too niche. Above 1,000, you're likely competing with established sites.

One caveat: Keyword Planner shows broad ranges rather than exact numbers. If it says "100-1K," you're working with rough data. That's fine for a new blog. Precision matters more once you have traffic to optimise.

A mistake I made early on was obsessing over exact volume. I'd reject a keyword because it showed 90 searches instead of 110. That's absurd. The difference between 90 and 110 monthly searches is noise.

Step 3: check difficulty with free and paid tools

Volume alone tells you nothing about whether you can rank. You need a difficulty score. Semrush and Ahrefs both offer keyword difficulty metrics (they call it KD, or Keyword Difficulty), and both have free tiers with limited daily searches.

Semrush's free plan gives you ten queries per day. Ahrefs offers a free keyword generator tool with similar constraints. For a new blogger, ten queries per day is genuinely enough if you plan your research sessions.

Tool	Free tier	Best for	Limitation
Google Keyword Planner	Unlimited (with Ads account)	Volume data, location targeting	No difficulty score
Semrush	10 queries/day	Difficulty, competitor analysis	Daily cap on free plan
Ahrefs	Limited free tools	Backlink data, difficulty	Best features are paid
Google Search Console	Free (once you have a site)	Actual queries driving traffic	Only shows data after you rank

What difficulty score should you target? For a brand-new blog, I'd aim for keywords with a KD under 15. Some experienced bloggers say under 30, but those folks usually have at least some domain history. If your domain is three weeks old, be conservative.

Step 4: analyse search intent before writing a single word

Search intent is the reason someone types a query. Google's entire ranking algorithm is built around matching intent. If your post doesn't match what the searcher wants, you won't rank no matter how good your writing is.

There are four broad intent types:

• Informational — "how to grind coffee beans without a grinder"
• Navigational — "Blue Bottle near me"
• Commercial — "best espresso machine under $500"
• Transactional — "buy Baratza Encore grinder"

For a new blog, informational and commercial queries are your sweet spot. Navigational queries are useless to you (people are looking for a specific brand). Transactional queries are dominated by e-commerce sites.

Here's a quick test: Google your target keyword and look at the top five results. Are they blog posts? Product pages? Forum threads? If the top results are all Reddit threads, Google has decided that searchers want community opinions, not a well-structured article. You can still try, but know what you're up against.

Step 5: prioritise and build your content calendar

You'll end up with more keyword ideas than you can possibly write about. Prioritisation is everything.

I use a simple scoring system. For each keyword, I rate three things on a scale of 1 to 5: volume, difficulty (inverted—low difficulty scores high), and my personal interest in writing about it. Multiply the three numbers. Anything scoring above 40 goes into the calendar. Below 20 gets discarded or saved for later.

That last criterion—personal interest—matters more than most guides admit. If you're writing about something you find tedious, your readers will feel it. A 15% drop in writing quality because you're bored will cost you more than a slightly better keyword ever gains you.

Fill your calendar with two to three posts per week for the first three months. After that, review what's getting traction and double down on those clusters.

What changes for a brand-new blog

Established sites can target competitive keywords and rank within weeks. You can't. A domain that's been around for five years with hundreds of backlinks has trust signals you simply don't have yet.

This means your strategy has to lean heavily on long-tail keywords. These are queries with four or more words, lower search volume, and less competition. They're also more specific, which means the traffic they bring is often more engaged.

One post of mine targeting a 70-searches-per-month keyword brought me 12 email subscribers in two months. A different post targeting a 5,000-searches-per-month keyword brought 900 visitors and zero subscribers. Specificity wins.

How long before you see results?

Realistically, three to six months before Google starts sending consistent traffic. Some posts will rank faster. Others will sit dormant for half a year and then suddenly climb. This is normal, and it's the hardest part of blogging to accept.

I've had posts that got 3 visitors in month one and 400 visitors in month eight with zero changes. Google needed time to trust the content and gather engagement signals. Patience is a strategy, not a virtue.

Should you use AI to generate keyword lists?

You can, but treat the output as raw material, not validated data. AI tools are good at brainstorming related questions. They're terrible at estimating search volume or difficulty. Use them in step one, then validate everything in steps two and three.

The bigger risk is publishing AI-written content at scale. Google's helpful content system has become increasingly good at identifying mass-produced text that doesn't demonstrate first-hand experience. Your advantage as a new blogger is your specific perspective. Don't throw that away for speed.

The AI search factor nobody warned you about

Here's something that's changed in the past two years: you can rank number one and still get almost no clicks. AI Overviews now answer many informational queries directly on the results page. If your target keyword is "what is keyword research," Google will show a generated summary and the user may never scroll down.

Does that mean keyword research is dead? No. It means you need to target queries that AI can't easily answer. Comparison queries, experience-based queries, queries that require a specific point of view. "What is keyword research" is now a losing battle. "How I chose my first 20 blog keywords and what happened" is not.

My rule: if a query can be answered in three sentences, AI will answer it. If it requires a story, a framework, or a personal recommendation, you still have a shot.

This isn't doom and gloom. It's a filter. The keywords worth writing about are now, by definition, the ones that require a human. That should make your job easier, not harder.

Tracking and iterating: the part most bloggers skip

Once you have ten or fifteen posts live, install Google Search Console. It shows you the actual queries people use to find your site. This is where the real learning happens.

You'll discover two things. First, some posts rank for keywords you never targeted. Second, some posts get impressions but no clicks. Both are gold.

When a post ranks for an unexpected query, write a dedicated post for that query. You've already proven Google associates your site with the topic. When a post gets impressions but no clicks, rewrite the title and meta description. A 2% click-through rate on 1,000 impressions is 20 visitors. Bump it to 5% and you've tripled your traffic with a ten-minute edit.

I check Search Console every Monday morning. It takes fifteen minutes. Over six months, that habit has been worth more to my traffic than any single tool subscription.

What about paid tools?

If you're serious about blogging as a long-term project, Semrush or Ahrefs will pay for themselves once you're earning anything. But don't subscribe on day one. Use the free tiers, publish twenty posts, and only upgrade when you've hit the limits of what free data can tell you.

Most new bloggers quit before they reach that point. The tool isn't the bottleneck. Consistency is.

Write the posts. Track the data. Adjust the titles. Repeat for a year, and you'll be ahead of 95% of the people who started when you did. Not because you found a secret keyword, but because you kept going long enough for the compounding to kick in.