How to Do Keyword Research for Your Business, Using Free Google Tools

A chartered accountant in Kharghar types “CA for GST registration Kharghar” into a keyword tool and gets back a volume of zero. He concludes nobody searches for it and decides to write about “income tax” instead, a phrase with huge volume where he will compete with government portals and national publications. That is the most common keyword research mistake I see, and it comes from trusting the tool more than the customer.

Keyword research is finding the words people type when they need what you sell, so you can build pages that match. In India that mostly means Google: Statcounter puts its share of search at 96.63% in September 2026. The good tools for Google are mostly free and mostly from Google. So this guide answers a practical question: how does a small business owner find the keywords that bring enquiries, without paying for software or being misled by volume numbers?

1. Start with your customers’ words, not a tool

Your best keyword list is already sitting in your phone. Customers describe their problem in their own words when they call, message or walk in, and those words are often different from how you describe your service. A pest control company calls it “integrated pest management”; customers search “cockroach treatment for flat”. An interior designer says “turnkey residential projects”; customers search “2BHK interior cost”.

In Mumbai and Navi Mumbai you will also hear area names, building-society terms and mixed Hindi-English phrasing. In my experience these show up in Search Console far more than in keyword tools, and they are worth noting exactly as customers say them, spelling variations included.

Write down three groups to begin with: your services as customers name them, the problems that make people look for you, and the areas you serve. Combine them and you have a seed list of 20 to 50 phrases before you open any tool.

2. Expand with Google’s own free suggestions

Type each seed phrase into Google and watch the autocomplete predictions. Google’s autocomplete help page says predictions “reflect real searches that have been done on Google”, adjusted for language, location and trending interest. That makes them a direct, free view of how people in your area phrase things.

Then look further down the results page. The “People also ask” box shows questions related to your search, and the related searches at the bottom show other phrasings. A useful trick is to add letters after your phrase, such as “root canal a”, “root canal b”, to surface more predictions.

Two cautions. Predictions are personalised by location and your own past searches, so check in a private browser window. And Google says predictions are not statements of fact, so a prediction tells you what people type, not what is true or popular in absolute terms.

3. Read the intent behind every keyword

Every search has a purpose behind it. Getting that purpose right decides whether a page ranks and whether the visitor becomes a customer. Look at the words, then confirm by searching the phrase and seeing what kind of pages Google already shows.

Illustration based on my own client work; check intent for any phrase by looking at what Google currently ranks for it.

The Google results page is the best intent check there is. If the top results for a phrase are all how-to articles, a sales page will struggle there. If they are all business listings and service pages, a blog post will struggle. Match what Google already rewards rather than arguing with it.

For “near me” and area searches, remember that the map results are a separate system. Google’s local ranking help page says local results are based mainly on relevance, distance and prominence. Your website helps with relevance and prominence, but your Google Business Profile does much of the work.

See also: For the map side of local searches, see our Google Business Profile guide and local SEO for Navi Mumbai.

4. Treat search volume numbers with suspicion

Every keyword tool shows a monthly search number. It is useful for comparing phrases against each other and misleading as an absolute count, especially for local businesses. Here is what the main free tools actually tell you, in their makers’ own words.

Sources: Google Ads Help, About Keyword Planner; Google Trends Help, How Trends data is adjusted; Search Console Help, Performance report data; Google Search Help, autocomplete.

The key line is in Google’s own Keyword Planner help: “Keywords with very low search volumes, or keywords that are considered sensitive aren’t discoverable or forecastable.” A phrase like “CA for GST registration Kharghar” may be searched a handful of times a month by people ready to hire. A tool that cannot see it will report it as nothing.

This is not a quirk of Google’s tool. Ahrefs reports that keywords with fewer than 10 searches per month make up almost 93% of its US keyword database. That is US data from a paid tool’s own dataset, so I would not apply the exact figure to India, but the shape is the same everywhere: most searches are rare, specific phrases. Ahrefs defines long-tail keywords by this low volume, not by word count.

One more caution on Google Trends. Its help page says each data point is divided by total searches for that place and time, then scaled from 0 to 100. A score of 100 in Navi Mumbai and 100 in Delhi do not mean the same number of searches. Use Trends for seasonality, such as when interior or wedding searches peak, not for sizing a market.

5. Why small, specific keywords are the opportunity

Compare two searches. “Interior designer” could come from a student, a job seeker, someone in another city or a person browsing for ideas. “2BHK interior designer Kharghar budget” almost always comes from someone in or near Kharghar with a flat and a budget. The second phrase is searched far less, and is far more likely to turn into a site visit.

Specific phrases also face fewer competitors. National brands and directories target the big head terms. A local business can realistically rank for the long tail of service-plus-area-plus-detail phrases, and twenty of those together often bring more enquiries than one broad term you will never rank for.

There is a limit to this. If you build one thin page for every possible area and phrase combination, with only the place name swapped, you create the kind of low-value pages Google’s helpful content guidance warns about. Group close variations onto one strong page instead.

6. Check who you are up against

Search each target phrase and look honestly at page one. You do not need a paid “difficulty score” for this. Ask three questions: what kind of sites rank, how good are their pages, and is there room for a local specialist?

• If page one is government portals, Wikipedia and national media, the phrase is mainly informational and hard to win. Look for a more specific angle.
• If page one is directories and marketplaces, you can still compete, but your Google Business Profile and listings on those directories matter as much as your site.
• If page one has small business sites with thin, generic pages, that is your opening. A page that actually answers the question in detail can win.

Look at the map pack too, for local phrases. If competitors there have many more reviews than you, the gap is in reviews rather than keywords.

7. Mine the data you already own: Search Console

If your site has been live for a few months, Google Search Console holds the most valuable keyword data you will ever get: the actual searches your pages appeared for. The Performance report shows clicks, impressions, click-through rate and average position for each query.

The most useful filter is simple: sort by impressions and look for queries where you appear often but rank on page two, roughly positions 8 to 20. Those are phrases Google already associates with your site. Improving the relevant page, or creating one if none fits, is usually the fastest win available.

Google notes that the report does not show every query, to protect user privacy, so totals will not add up perfectly. It is still real data from real searches, which no third-party estimate can match. If you run Google Ads, the search terms report gives the same kind of real-query data for paid clicks.

See also: If you are deciding between paid and organic for these keywords, read Google Ads vs SEO and our Google Ads cost guide for India.

8. Map every keyword group to one page

The last step turns a list into a plan. Group phrases that mean the same thing, then assign each group to one page, existing or new. Service and hire-intent groups go to service or location pages. Learn and compare groups go to blog guides that link to the relevant service page.

Illustration only; the phrases and pages are examples of the method, not research data.

The map also stops cannibalisation, where two of your own pages target the same phrase and split Google’s attention. If the map shows two pages for one group, merge them or give each a clearly different job.

See also: Once a page has its keyword group, use our on-page SEO checklist to build or rewrite it.

Where this advice breaks down

This method is built for local and niche service businesses. If you sell products nationally online, volume data matters more, competition is fiercer, and a paid tool such as Ahrefs or Semrush earns its cost for the competitor and backlink data alone. Free tools will still get you started, but they will not show you everything.

Keyword research also cannot rescue a page that is weak. Google’s SEO Starter Guide says keyword stuffing is against its spam policies, and the helpful content guidance asks whether a page exists mainly to help people or to attract search traffic. Knowing the phrase only gets you into the room; the page still has to be the most useful answer.

And searches change. AI Overviews, voice search and new phrasings shift how people ask, and seasonal businesses see big swings. Revisit your list every six months, starting from fresh Search Console data and fresh customer conversations, not last year’s spreadsheet.

Your keyword research checklist

The Kharghar CA did not need a bigger keyword. He needed to trust the phrase his clients were already using, and build one good page around it.