How to Get Keywords for SEO: A Step-by-Step Process

You've probably heard it a thousand times: "SEO starts with keywords." And it does. But almost nobody talks about what actually comes before that. We've watched teams waste months optimizing for keywords their customers don't use, ranking for traffic that never converts, or targeting terms that Google itself answers before anyone clicks through to a website.

This is the step-by-step process we use when we need to actually figure out how to get keywords for SEO—not the version that assumes you already know what your audience is searching for.

Start With Audience Discovery, Not Tools

Here's what trips people up: they open Google Keyword Planner and start typing. Two weeks later, they have a spreadsheet of 500 keywords and no clear sense of which ones matter.

We learned this the hard way. Early on, we'd run keyword research for a client selling chicken coops and come back with a list dominated by "chicken" and "coop"—terms that generated thousands of monthly searches. The client was thrilled by the volume numbers. Then we'd spend months ranking those pages and see almost no business impact. Turns out, people searching "chicken" were usually doing homework for a school project. Not exactly our buyer.

The fix was embarrassingly simple: we stopped researching keywords and started researching people.

Before you touch a tool, answer these questions internally. Write them down. Actually answer them—not the answer you think sounds good, but the real answer based on how your customers talk.

Who is your actual buyer? Not the persona you wrote in a deck three years ago, but the person who has given you money or seriously considered it. What's their background? Are they a hobbyist, a small business owner, a professional? Do they have budget already allocated or are they exploring?

How do they describe their problem? This matters more than you'd think. Someone managing a small farm talks differently than someone starting a backyard project. "I need to scale my poultry operation" is not the same as "I want to raise chickens in my backyard."

What language do they actually use? This is where most guides fail. They tell you to "do audience research" and leave you hanging. Get specific. Spend time in the places where your audience hangs out naturally—Reddit communities, Facebook Groups, YouTube comments, Twitter threads. Copy the exact phrasing people use. When someone asks a question, note the words they choose. Don't paraphrase it through your marketing brain; write down what they actually wrote.

We keep a shared document of real customer questions and phrases, grabbed directly from support emails, phone calls, and community forums. That document is worth more than any keyword research tool because it's in your audience's actual voice, not SEO's standardized vocabulary.

Building Your Seed List From Real Data

Once you've done that listening, generate your initial seed keywords—the 5 to 20 phrases that form the foundation of your research. This isn't a guess; it's a synthesis of what you just learned.

If your audience is asking "Can I raise chickens in an apartment?" that's a seed keyword (with some cleanup). If people keep mentioning "predator-proof coops," that's another one. If customer conversations repeatedly mention "backyard chicken laws" as a pain point, that belongs on the list.

Your seed list should reflect your business capability too. There's no point targeting "commercial poultry farming" if you sell coops for five-bird operations. Be honest about what you can actually speak to and rank for. This filters out wasted effort early.

Free Expansion: Google's Native Features First

Now the expansion starts, and this is where most people get it backwards. They jump to paid tools and expensive subscriptions.

We always start with Google's free features because they're direct signals from your actual audience. Google Autocomplete, People Also Ask, and Related Searches aren't estimates or aggregated data—they're real queries people have typed into Google.

Take one seed keyword. Type it into Google and watch the autocomplete suggestions populate. These are high-frequency queries related to your topic. Capture all of them. People Also Ask, the box that shows questions related to your search, often reveals intent you didn't anticipate. Someone might be asking "how much do chicken coops cost" when you expected them to ask "how to build a chicken coop."

At the bottom of the search results, Google shows Related Searches. These tend to be a mix of long-tail variations and related topics. All of it is free and all of it is real behavior.

Google Trends is the underutilized tool here. It shows you search volume patterns over time. This matters because some keywords are seasonal ("raising chickens" spikes in spring). Others are one-off trends. If you're planning content, you want to know which keywords have staying power.

This phase usually generates another 30-50 keyword variations from a small seed list. And we haven't paid anyone yet.

Tool-Based Expansion: When and What to Use

Once you've exhausted the free options, tools help you add structure and scale. But not all tools are the same, and most people use them wrong.

Google Keyword Planner is technically free (if you have a Google Ads account). It gives you search volume ranges and competition levels. That's it. No intent filtering, no difficulty scoring, no traffic potential insights. It's a starting point, not a destination.

Ubersuggest and SE Ranking's free tiers give you more—keyword suggestions, some difficulty estimates, basic competitor data. They're fine for small research projects. Neither will break your budget.

Ahrefs, when you get to the paid tier, adds something genuinely useful: Traffic Potential. This metric shows you the actual monthly traffic the top-ranking page receives, not just the search volume. That distinction matters more than it seems.

We input our seed keywords into whichever tool fits our budget, apply filters for search volume and word count (three-word phrases and longer tend to be more specific), and let it generate a list of expansion keywords. The tool organizes what we already found through Google and adds depth.

Then we filter ruthlessly. High search volume alone doesn't mean anything. Low volume with low competition? That's interesting. High volume with high difficulty? Probably skip it unless you're already an authority site.

The Traffic Potential Trap Nobody Talks About

Here's where things get weird and most guides miss it entirely.

Search volume is a lie. Or not a lie exactly, but a incomplete picture that can mislead you into months of wasted work.

Say you find a keyword with 10,000 monthly searches. It seems juicy. You spend time optimizing content, building backlinks, tweaking technical SEO. Six months later, you rank number one. Then you check your analytics and see… nothing. Maybe a few hundred visitors, but nowhere close to 10,000.

There are three reasons this happens.

First: AI Overviews. Google now answers certain questions directly in the search results—definitions, quick facts, simple how-tos. If someone searches "what are backyard chickens," Google shows the answer right there. No click needed. You can rank #1 and get zero traffic because the query is already answered. This is a 2025 reality that most keyword research guides were written before encountering.

Second: search volume doesn't equal traffic potential. The top-ranking page often ranks for 20 or 30 related keywords, not just the one you targeted. Its actual traffic is 2 to 3 times higher than the primary keyword's search volume. But you don't see this until you analyze the page directly.

Third: CTR varies wildly. A keyword with high commercial intent (someone trying to buy something) might convert 10% of clicks. A keyword with informational intent (someone learning) might convert 0.5%. You're optimizing for volume, not value.

This is why we obsess over Traffic Potential instead of search volume. If Ahrefs shows that the #1 page gets 50,000 monthly visitors from a keyword with "only" 5,000 monthly searches, that tells you the keyword has hidden depth. The SERP page is answering related questions and pulling long-tail traffic you can't see in the raw volume number.

The practical fix: always check the actual traffic of top-ranking pages. Use Ahrefs, SE Ranking, or SimilarWeb. If the #1 page shows 1,000 monthly visitors but the search volume says 10,000, you've got an AI Overview or low-CTR problem. Skip it or go deeper.

If the #1 page shows 8,000 monthly visitors from a 2,000 search volume keyword, you've found a goldmine. That keyword has intent depth. Target it.

Matching Keywords to the Right Content and Funnel Position

Once you've narrowed your list, the next mistake is obvious in hindsight but brutal when you miss it: optimizing a page for the wrong type of keyword.

Keywords have intent. Someone searching "how to build a chicken coop" wants information. Someone searching "buy chicken coop online" wants to make a purchase. These are completely different intents, and ranking for the second keyword on a blog post about DIY construction means traffic that never converts.

This is where the work gets annoying because it requires actual thinking instead of just running a tool.

Take your narrowed keyword list and run the primary keyword through Google. Look at the top 10 results. Are they mostly blog posts? Product pages? How-to guides? Local business listings? The SERP itself tells you the intent that Google (and therefore your audience) associates with that keyword.

If eight of the top ten results are product pages from e-commerce sites, that keyword has commercial intent. Your homepage or product page should target it.

If seven of the top ten are long-form blog posts and guides, it's informational. A blog post is the right fit.

Mixed intent is the tricky case. "Chicken coops" might pull both buyers and researchers. Google often shows a mix of product pages and guides. In this case, you'll want to target it with a page that covers both angles—a comprehensive buying guide that also includes DIY options, for example.

We assign keywords to pages like this: brand and product keywords go to the homepage or product pages. "How-to" and educational queries go to blog content. Local queries go to location-specific pages. Question-based queries often live in FAQs or resource hubs.

The key is matching the intent of the keyword to the intent of the page. If the mismatch doesn't align with what Google is ranking, you'll rank but convert nothing.

Long-Tail Keywords Are Where Beginners Actually Win

Most keyword research guides tell you to target long-tail keywords—three or more words, lower volume, higher specificity. Then they treat it as optional advice for later.

It should be the main strategy.

We've built entire content strategies around three-to-five-word queries that have 50-200 monthly searches. These keywords are easier to rank for, have clearer intent, and often convert better than broad terms. A question like "how much does a chicken coop cost for five birds" has less volume than "chicken coop," but it's also less competitive by about 1,000 times. You can rank in weeks instead of months.

The psychology works against this. Broad keywords feel prestigious. "Ranking for chicken" sounds better in a status update than "ranking for backyard chicken coop setup cost." But the second one actually makes money.

Question-based long-tail keywords are especially valuable. "Where can I buy chicken coops near me" tells you someone is ready to buy. "What are the best chicken coop designs for predator protection" tells you what features matter to your audience. These aren't vanity metrics—they're signals.

Start with long-tail. Get wins. Build authority. Then move up to broader terms if you want. This is the path that actually works.

Learning From Your Competitors Without Copying

Competitive analysis is the part where people usually make a mess. They look at the top-ranking page for their target keyword, see what it's optimizing for, and try to replicate it.

That's not the insight you want.

What you actually want to know is: what is that page actually ranking for, not just what it looks like it's targeting?

If you're competing for "backyard chicken setup," run that URL through Ahrefs Site Explorer or SE Ranking's competitor module. The tool will show you all the keywords that page ranks for. You might see that 40% of its traffic comes from long-tail variations you hadn't considered. Or that it ranks for 15 variations of "chicken coop placement" that you didn't know existed.

This is where Traffic Potential analysis becomes actionable. You're reverse-engineering what's actually working.

Second, note the content structure. Not to copy it, but to understand what Google and users expect for this query type. If the top-ranking page is 2,500 words with embedded videos, you probably need similar depth. If it's a concise 800-word guide, don't waste time adding padding.

Third, look for gaps. The top-ranking page might be good at answering the main question but weak on pricing or specific use cases. That's your angle. You'll rank differently, often higher, by being more helpful on the dimensions competitors missed.

The Free vs. Paid Tool Reality

We use free tools for everything up to 20 keywords. Google's native features, Keyword Planner, Ubersuggest's free tier. They work fine.

Once you're managing 50+ keywords or need to track competitor ranking changes over time, the ROI of a paid tool flips. Automation saves hours. Traffic Potential metrics are worth the subscription alone if you're serious about competitive analysis.

But the expensive tools don't replace the work. They accelerate it. And they definitely don't replace thinking about what your audience actually needs.

Final Pass: Validation Before You Commit

Once you've narrowed your keyword list to something actionable—maybe 20-50 targets depending on your content capacity—do one final check.

Pick your top five keywords. Search each one. Read the top three results carefully. Could you write something better? Do you actually have expertise or data that would rank higher? Be honest here. If Google is already ranking content that's better than what you could produce, adjust your target list.

Check recent search trends. Is this keyword growing or shrinking? If it's volatile or seasonal, plan your timing.

Think about the effort-to-payoff ratio. A keyword with 500 monthly searches and clear commercial intent might be worth three months of effort. A keyword with 100 monthly searches and informational intent might not be.

This isn't a technical step. It's just reality-checking before you spend time you can't get back.

FAQ

Why shouldn't I just start with a keyword research tool?

Because you'll end up with 500 keywords and no sense of which ones matter. Tools show volume numbers, not whether your audience actually uses those terms or whether they convert. You need to understand how your real customers talk about their problems first, then use tools to expand and validate. Start with listening, not spreadsheets.

What's the difference between search volume and traffic potential?

Search volume is what Google says people are searching for each month. Traffic potential is what the actual top-ranking page receives in real traffic. They're often wildly different because AI Overviews answer some queries directly, pages rank for 20 related keywords you don't see, and click-through rates vary by intent. Always check the actual traffic of top pages instead of trusting volume numbers alone.

Should I target broad keywords or long-tail keywords first?

Start with long-tail keywords. Three-to-five-word queries with 50-200 monthly searches are easier to rank for, have clearer intent, and often convert better. You'll get wins faster and build authority. Once you're established, move to broader terms. Most beginners waste months chasing prestigious-sounding broad keywords that never rank.

How do I know if a keyword is worth optimizing for?

Run it through Google and check the top three results. Could you write something better? Does it match your actual expertise? Then use Ahrefs or SE Ranking to check if the #1 page gets meaningful traffic. If the top-ranking page shows 1,000 monthly visitors from a keyword with 10,000 search volume, there's a problem. If it shows 8,000 visitors from 2,000 searches, you've found a goldmine.