How to do keyword research like an agency (not a 50-word list)

Most people learn how to do keyword research the wrong way. They open a free tool, type in their main service, copy the 40 or 50 phrases it spits out, sort by volume, and start writing for the biggest numbers. Six months later they are stuck, wondering why a brand-new site with ten pages cannot outrank companies that have been publishing for a decade.

The gap is not effort. It is method. In this guide we will show you how to do keyword research the way we actually do it for clients at Zarle Infotech, which is closer to how a real agency thinks than what most blog tutorials teach. You will get the full process, a prioritization framework you can use today, and the honest numbers from our own projects so you know what to expect.

Here is the short version before we go deep. Stop hunting for 50 keywords. Mine thousands of your competitors' keywords instead, find the small and winnable ones first, publish for those, and let the early wins compound into the authority that lets you take the bigger terms later.

Why the "50 keyword" approach fails

When you pull only 50 keywords from a single tool, you are almost always looking at the popular, high-volume, high-competition terms. Those are the ones everyone chases. A new or small site has no realistic chance of ranking for them yet, so you write, you wait, and nothing moves.

There is also a content problem hiding underneath. Small sites with thin pages are genuinely hard to rank. If the content on every page is not solid, weak pages drag the whole domain down, and no amount of keyword picking fixes that. Learning how to do keyword research properly only pays off when it sits on top of real, well-written pages.

So the first mental shift when you learn how to do keyword research is this: it is not about finding words to write about. It is about deciding which searches you can actually win right now, and which ones to save for later.

The agency method: mine thousands, not fifty

Here is where our process splits from the typical tutorial. Instead of pulling around 50 keywords, we pull 10,000 to 20,000 competitor keywords using paid tools, then filter down to the small, low-competition wins first.

Think about what that changes. When you look at 15,000 keywords your competitors already rank for, you are not guessing what your market searches. You are reading a map of proven demand. Buried inside that list are hundreds of phrases with 50 to 100 monthly searches, low difficulty, and clear intent that nobody is fighting hard over. Those are the terms a small site can rank for in weeks, not years.

This is exactly what we did for Dr. Prakash Hospital. The site started at roughly 800 to 1,000 visits a month and had been stuck there. We deliberately targeted the smallest keywords first, the low-volume, low-competition ones most agencies ignore because the numbers look unexciting. Those small wins compounded, lifted the site's authority, and pulled the bigger terms up behind them. The exact filtering and prioritization we run on that raw list is our own secret sauce, but the principle is not a secret at all: win small first, then scale.

If you want the deeper argument for why you should chase a wide net of terms rather than a handful, we wrote a companion piece on how many keywords to target that pairs well with this one.

Step by step: how to do keyword research the Zarle way

You do not need our exact stack to use the method. Here is the full process broken into steps you can run yourself.

Step 1: build your seed list from what you actually know

Before any tool, write down the words your customers really use. A doctor knows the procedures patients ask for. A product seller knows what buyers call the product. Pull from your sales calls, support questions, and Google Search Console if the site already exists. This raw language often beats what a volume-first tool hands you, because it reflects real intent, not just search counts.

Step 2: expand with a multi-tool stack, not one source

There is no single perfect keyword tool. We use a mix of Ahrefs, Ubersuggest, and SEMrush, plus local keyword analyzers to see what is working in a specific city or region. Each tool sees a slightly different slice of the search data, so relying on one gives you a narrow picture. Feed your seeds into a couple of them and export everything.

Step 3: pull your competitors' entire keyword footprint

This is the step most DIY guides skip. Take your top three or four competitors, drop their domains into your tool, and export every keyword they rank for. This is how you get from a 50-word list to a 10,000 to 20,000 keyword pool. You are no longer brainstorming in the dark. You are inheriting years of your competitors' proven research for the price of a subscription.

Step 4: sort for the smallest winnable terms first

Now filter aggressively. Look for keywords with low difficulty, honest intent, and modest volume, often in the 50 to 100 monthly search range. These are your first targets. They will not look impressive on a spreadsheet, and that is precisely why they are open. Everyone else is fighting over the head terms.

Step 5: match each keyword to intent and a page

For every keyword you keep, ask what the searcher actually wants. Are they looking for information, comparing options, or ready to buy? Then decide which page answers it: a service page, an around-service page, or a blog. No keyword should float without a home. Keyword research that does not end in a page-by-page plan is just a spreadsheet nobody opens again.

Step 6: publish, then read the data and double down

In the first two months, do not bet everything on 10 or 15 keywords. Go after 150 to 200 keywords to find where you actually perform. Some will surprise you by ranking fast; others will not move. Once you can see the winners in Search Console, double down on those themes and expand around them. SEO is not a set-and-forget task, it is constant checking of what is working, what is not, and why.

A prioritization framework you can use today

The hardest part of learning how to do keyword research is not finding keywords. It is deciding which ones to write first when everything looks important. Three factors matter: search volume, keyword difficulty, and intent. The mistake most people make is fixating on one, usually volume, and ignoring the other two.

Here is the simple scoring approach we lean on. Rate each keyword on all three, then let the combination tell you where it belongs.

The logic is straightforward. A low-volume phrase you can rank for this month is worth more than a high-volume phrase your competitors own and you cannot touch for a year. Business value and winnability beat raw volume every time. As your site gains authority from the quick wins, terms slowly move up the table, and yesterday's authority play becomes today's steady builder.

Intent deserves its own note. A phrase like "best orthopedic surgeon in Noida" carries strong local, buy-ready intent, so it is worth far more per visitor than a broad informational term that pulls readers from other countries who will never book. When we prove ROI for clients, we read this intent split directly, separating buy-intent traffic from branding and info traffic rather than celebrating raw visit counts. We go deeper on that in our guide to measuring SEO ROI.

What good keyword research actually gets you

Done right, this method compounds. Here is the real trajectory from one healthcare client we have run SEO for over about 18 months. The site sat at roughly 1,000 monthly visits and had been stuck there for a long time.

• First 6 months: 1,000 to 4,000 monthly visits
• Next 6 months: 4,000 to 7,000
• Last 6 months, after a strategy change: 7,000 to 21,000 to 22,000 monthly active users

That site now ranks number one on multiple keywords and top three on many more, and its authority is high enough that new posts take off quickly. The business impact matters more than the traffic: that client went from basically zero appointments from the website to a steady 5 to 6 appointments a day and 200-plus appointments a month from the site alone. None of that started with head terms. It started with small, winnable keywords that most people would have deleted from the spreadsheet.

Typical results are more modest and still worth it. We usually see around 30 to 40 percent growth month over month, and a site starting from zero can reach 500 to 800 monthly visits in six months. If you want a realistic timeline, read how long SEO takes before you set expectations.

Keyword research is only half the job

Knowing how to do keyword research is only half the job. Finding the right keywords does nothing if the pages behind them are weak. A few honest rules we hold to on every project:

• Depth wins. Aim for 1,500 to 2,000 word pages with a short summary up top. In our experience, 1,000 to 1,200 word blogs simply do not perform.
• Interlink everything. No standalone pages. A service page for a procedure should link to the blogs about that procedure, and back again.
• Get the technical hygiene right. Clean heading structure, proper meta titles and descriptions, and above all, make sure pages are actually getting indexed. If a page is not indexed, ranking it is impossible.
• Add detailed FAQs with schema. They help both traditional rankings and citations in AI answers.

And expect turbulence. Down months are normal, not failure. A site can slip from 500 to 300 visits in a month and still be on the right long-term path. Even the biggest agencies test constantly. Judge results fairly over a year, not a single report. If you are wondering whether the whole effort pays off, we made the case in is SEO worth it.

For local businesses, keyword research also has to bend toward location. You add the terms people search near you, and your Google Business Profile does much of the heavy lifting. We cover that in our guide to local SEO for doctors, where the same "win small local terms first" logic applies.

Frequently asked questions

How many keywords should I research for a new site?

Do not stop at 50. Mine your competitors' full footprint, which can be 10,000 to 20,000 keywords, then narrow down. For actual targeting in the first two months, plan for 150 to 200 keywords so you can find where you genuinely perform, then double down on the winners.

Should I target high-volume or low-volume keywords first?

Low volume, low competition first, often in the 50 to 100 search range. They rank faster, build authority, and compound. High-volume head terms are worth chasing only once your domain has the authority to compete for them.

Which keyword research tools do you actually use?

A mix, because no single tool is complete. We use Ahrefs, Ubersuggest, and SEMrush, plus local keyword analyzers to catch what is trending in a specific city. For finding FAQ-style queries we also lean on tools like AnswerThePublic.

How do I judge keyword difficulty without paid tools?

Search the term and study who ranks. If the first page is full of big, established brands with deep content, difficulty is high. If you see thin pages, forums, or weak local results, there is room. Paid tools give you a difficulty score, but reading the actual results teaches you the most.

How long before keyword research turns into traffic?

Month one is mostly setup: getting the site right, checking Search Console, and confirming pages index. Real movement usually starts within one to two months once good content is going live, and it compounds from there. It is a consistency game, not an overnight one.

Is keyword volume data always accurate?

No. Tool volumes are estimates and they differ between tools, which is one reason we cross-check across several. Treat volume as a rough signal, not a promise, and weight intent and winnability just as heavily.

What does professional keyword research cost in India?

SEO retainers that include this depth of research typically run from around 40,000 to 1,50,000 rupees a month, depending on your niche and whether you are targeting local or international audiences. The research itself is only part of the retainer, which also covers content, technical work, and ongoing iteration.

Let us find your winnable keywords

If this sounds like more than a spreadsheet exercise, that is because it is. The real edge is in mining thousands of competitor keywords, filtering for the ones you can actually win, and staying consistent long enough for the small wins to compound. That is the work we do every day.

If you would rather have a team run this end to end, from research to published, indexed, interlinked pages, take a look at our SEO and content strategy service. We will map the small keywords worth winning first and build the pages that turn them into real leads.