Most lists of keyword types stop at definitions. Short-tail is short, long-tail is long, branded has a brand in it. That's fine for a glossary and nearly useless for deciding what to write on Monday morning.
This guide sorts keywords by a different test: what is the person about to do after they hit Enter? Read something, compare options, buy, call, or go to a site they already know. Once you sort by that, the old labels start to line up with real decisions. It also settles one argument for good: "LSI keywords" are not a thing Google uses, and chasing them costs you hours.
The numbers below come from the Delhi keyword study Nurotech ran on 10 November 2026. The study pulled monthly search volume for Delhi and for all of India using DataForSEO, and compared them with 90 days of Google Search Console data for nurotech.in (10 August to 8 November 2026). Treat every volume as an estimate. Keyword tools model demand; they don't count it. Still, the gaps between Delhi and India figures are large enough to teach something.
A keyword type is useful for exactly one reason: it predicts which page should answer the query. If the type doesn't change the page you build, it's trivia.
Take "types of keywords", the phrase this article targets. Per the study, it draws about 70 searches a month in Delhi and about 1,300 across India, with a keyword difficulty of 11 out of 100 and an estimated cost per click of $0.35. Almost nobody bids on it. The person typing it wants an explanation. They aren't comparing agencies and they aren't about to pay anyone. So the right page is a plain, thorough explainer, not a service page with a form.
Now compare "seo company in noida". The study put it at about 320 searches a month in Delhi against 2,400 for India. That searcher is shopping. They'll open three tabs, look at pricing, and probably call. Same field, same vocabulary, completely different next step.
That's the whole idea. The label ("informational", "commercial") is shorthand for the next step.
The idea isn't new. In 2002, Andrei Broder published a taxonomy of web searches, "A taxonomy of web search" in ACM SIGIR Forum, which split queries into navigational, informational and transactional. Nearly every modern four-way split (the one below adds "commercial investigation") descends from that paper. Google's own SEO Starter Guide makes the same practical point from the other direction: think about the words people actually use and what they want when they use them.
These four cover almost every query you'll ever plan content for.
Examples: "types of keywords", "what is a long-tail keyword", "how does local SEO work". The searcher has a gap in knowledge and wants it closed. They'll read, maybe save the page, and often leave without clicking anything else.
Volume is high and conversion is low in the short run. Informational keywords still earn their place, because they're where trust gets built months before a buying decision. A founder who learned from your explainer in March remembers your name in September.
What the page needs: a direct answer in the first 100 words, headings that mirror the questions people ask, and examples that come from a real source. What it doesn't need: a hard sales pitch in paragraph two.
Examples: "seo company in delhi", "seo agency vs freelancer", "best seo packages". Careful with that last one: the searcher uses "best", but your page shouldn't claim to be it. They want comparison material, evidence, process and pricing logic.
This group is where agencies live. Volume is smaller than the informational pool, and each visit is worth far more. A page for this type should show how the work is done, what it costs, how results are measured, and what happens in the first 30 days.
Examples: "hire seo consultant delhi", "seo audit price", "buy domain". The verb is the giveaway. The page here is short on theory and long on a clear next step: a form, a phone number, a booking link.
Transactional queries are also the easiest to misjudge. A lot of marketers treat any phrase with "services" in it as transactional. It isn't, always. "SEO services" pulls students, shoppers and competitors alike. Read the results page. If Google is showing mostly service pages, the intent is commercial or transactional. If it's showing guides and definitions, it isn't, whatever the word "services" implies.
Examples: "nurotech", "google search console login", "semrush". The searcher already knows the destination and is using search as a shortcut. You can't convert these people into customers, since they either want you or they want someone else.
You can lose them, though. If someone searches your brand name and finds a competitor's ad above you, or a thin, outdated profile page, that's a leak you caused. Own the first result for your own name and make sure it's correct.
Open the live results and look at the top five. Ask what kind of pages they are:
Google has already done the sorting for you. Your job is to read it, not to argue with it. If you want to rank a service page for a phrase where the top five are all glossaries, you're fighting the grain.
Short-tail means one or two words and a huge audience. Long-tail means three or more words, a narrower audience and a clearer need. The length is a symptom. The real difference is how much the searcher has already decided.
Here's what the study shows when you line up related phrases:
| Keyword | Delhi / month | India / month | What it tells you |
|---|---|---|---|
| types of keywords | 70 | 1,300 | Learner intent; volume sits outside Delhi |
| keyword types in seo | 40 | 590 | Same intent, longer phrasing, smaller pool |
| lsi keywords | 90 | 1,000 | Myth-driven; people repeat it from old advice |
| seo company in noida | 320 | 2,400 | Buyer intent with a place attached |
Two things jump out.
First, the Delhi figure is a small slice of the India figure for learner phrases. Roughly 5 percent for "types of keywords" (70 of 1,300) and about 7 percent for "keyword types in seo" (40 of 590). People who want definitions are spread across the country, and a Delhi agency can't build a local business on them alone. Learner keywords are for authority and reach, not for filling the calendar with Delhi enquiries.
Second, the buyer phrase behaves differently. "seo company in noida" is about 13 percent Delhi (320 of 2,400). The place name does the filtering. These searchers are looking for someone nearby, and a regional modifier raises the share of local demand a lot.
So long-tail isn't automatically better. A long phrase with a vague need ("how to do keyword research for a small business website step by step") can convert worse than a short phrase with a buyer's need ("seo audit delhi"). What a long-tail keyword gives you is lower competition and a more specific page to write. Whether that page earns money depends on the intent type, which is why the length label alone can't guide a plan.
A practical rule: don't pick keywords by length. Pick by intent first, then check volume and difficulty, and only then look at how many words it has.
A local keyword carries a place: a city, a district, a neighbourhood or "near me". The noida example above shows how strong that signal is. Add a place and the buyer's mind is made up about geography.
Three patterns come up again and again.
Explicit place. "seo company in delhi", "digital marketing agency in gurgaon". The searcher names the city. You need a page that names it too, in the title, the H1 and the body, with details that prove you actually serve the area. A list of suburb names pasted into the footer isn't proof.
Implicit place. "seo company near me", "web designer near me". There's no city in the query. Google decides what "near" means from the device's location. Here the ranking depends more on your Google Business Profile, reviews and proximity than on the words on your page. Nurotech's local SEO services page covers how that side works.
Service-area spillover. Searchers in Delhi often type Noida, Gurgaon or Ghaziabad because they'll travel or because the agency's team works remotely anyway. That's why a Noida phrase can matter to a Delhi business at all.
A local modifier also changes intent. "seo" is informational for a lot of people. "seo company in delhi" is commercial. The modifier moves the same topic from learning to hiring. If you take only one thing from this section, take that.
For a business that serves a city, the sensible order is: win the explicit-place pages, fix the Business Profile for the implicit-place searches, and write informational content that links back to both. The SEO guide for Delhi businesses walks through that sequence.
A branded keyword includes a company, product or person name. A non-branded keyword doesn't. The split matters because the two groups tell you different things about your business.
Branded searches measure awareness. If branded impressions in Search Console climb month over month, more people know your name. Non-branded searches measure reach. If non-branded clicks climb, you're being found by people who didn't know you existed.
Mixing them in one chart hides problems. A site can show healthy total clicks while all of it comes from people typing the brand name, which means the SEO isn't bringing in anyone new. Always filter Search Console by a query containing your brand, then look at the rest separately.
Competitor names are a third bucket, and an easy place to waste money. Ranking organically for a rival's brand name is close to impossible and bidding on it is expensive. Write a straight comparison page if the comparison is real, and leave it there.
One more detail: branded terms are navigational by nature. Treat them as protection, not growth. Make sure your homepage, your Business Profile and your main social profiles show up for your own name and say the same thing.
The study's own data shows "lsi keywords" pulling about 90 searches a month in Delhi and 1,000 across India. People are still searching for it, which tells you the myth is alive.
Here's the short version. Latent semantic indexing is a real technique from the late 1980s for finding relationships between terms in a body of documents. The Wikipedia entry on latent semantic analysis explains the maths and the history. It was developed for small, static document collections, long before the web existed.
The SEO use of the phrase came later. Tools started selling "LSI keywords": lists of words related to your main keyword, with the claim that Google uses a hidden list of them to judge whether your page is thorough. That claim is where it falls apart.
Google's John Mueller addressed it directly on Twitter in July 2019, saying there's no such thing as LSI keywords and that anyone telling you otherwise is mistaken. Search Engine Roundtable reported the statement when it happened. He has repeated the point since, and coverage from later years reports the same answer.
So what should you do with the idea underneath it? There is a sensible point buried in the myth: a page that covers a topic well naturally uses related words. A guide to keyword types will mention intent, volume, difficulty, Search Console and so on without being told to. That isn't a special ranking input. It's what a knowledgeable writer produces by default.
Here's the practical test.
If a tool's "LSI" list includes useful related questions, use it for that. The harm isn't in the tool, it's in believing a hidden mechanism exists and writing awkward sentences to satisfy it.
A real example of the difference: a paragraph that reads "Our keyword research, keyword research tool, keyword research for SEO and keyword research service help with keyword research" is stuffing, with or without the LSI label. A paragraph that explains how to export queries from Search Console, group them by intent and pick one per page uses plenty of related words and reads like a person wrote it.
With the types sorted, the order of work becomes a question about your business stage, not about keywords.
Start with low-difficulty informational and long-tail commercial phrases. A keyword difficulty near 11, like the one attached to "types of keywords" in this study, is reachable for a site with modest authority. You won't get Delhi leads from it, but you will get a page that earns links, builds topical coverage and gives your service pages something to receive internal links from.
Look at commercial and transactional pages. Check whether your service pages target phrases with a place or a buying verb. Often the problem is that the site ranks for learner phrases and has no page built for the buyer. Fix that gap before writing anything new. Our SEO company in Delhi page is an example of the commercial type: a service page aimed at someone choosing an agency.
This is an intent mismatch. Pull the queries that brought those visitors and sort them using the four-way test. Many agencies find that a vague informational page is attracting job seekers, students or competitors. The fix is usually clearer language on the page, or moving that content under a guide URL, not deleting it.
Give each month a mix. A workable ratio is something like half informational pieces that build reach, a third commercial pieces that compare and prove, and the rest transactional or local pages that close. The exact split depends on your sales cycle, so treat it as a starting point. For a longer view of how to schedule that mix, see Nurotech's piece on building a content calendar.
That last step catches the most mistakes. A service page won't rank where Google shows guides, and a guide won't convert where Google shows pricing.
Two pages for one intent. If you publish "types of keywords" and "keyword types in seo" as separate posts, they'll compete. Search Console will show both getting impressions on the same queries and neither ranking well. Merge them.
Chasing volume. A phrase with 2,400 searches a month and heavy competition may send less useful traffic than one with 40 searches and a buyer behind each. Count what a visit is worth, not how many there are.
Ignoring the modifier. Adding "in delhi" to a phrase doesn't just narrow it; it changes who's asking. Plan the page for that person.
Treating the type as fixed. Intent shifts. A phrase that was informational in 2022 can turn commercial once Google starts showing tool pages and comparisons. Re-check the results page every few months.
Definitions only stick once you've used them, so here's a short list sorted the way you'd do it in a spreadsheet. The phrases come from the topics in this study; the labels come from reading what Google shows for each one.
| Keyword | Intent | Length | Page to build |
|---|---|---|---|
| types of keywords | Informational | Short-tail (3 words) | Explainer guide |
| keyword types in seo | Informational | Long-tail (4 words) | Same guide, not a new post |
| seo company in delhi | Commercial | Long-tail (4 words) | Service page with process and proof |
| seo company in noida | Commercial | Long-tail (4 words) | Location page, only if you serve Noida |
| seo audit delhi price | Transactional | Long-tail (4 words) | Audit page with a booking form |
| google search console | Navigational | Short-tail (3 words) | None; the tool owns this result |
Notice what the table does. "types of keywords" is three words and "seo company in delhi" is four, yet one is a lesson and the other is a purchase. Length didn't separate them. Intent did.
Notice too that two rows share one page. The most common content-planning error is treating every phrase variation as a new post. If the top results for two phrases are the same pages, Google treats them as one question, and so should you.
The last row is also worth a second look. Not every keyword deserves a page. Skip phrases where the owner of the tool or brand holds the result and you'd be a distraction.
You don't need a paid tool to start. Google Search Console already lists the queries people typed before landing on your site, and it's free.
The method takes about an hour and works for any site. Use the Search Console performance report and follow these steps.
What you're looking for is a mismatch. Say 80 percent of impressions sit in informational queries but 90 percent of your pages are service pages. Then the site is doing the right thing for the wrong audience, or the other way round. The first fix is nearly always to match page type to query type before touching titles or links.
A zero is also information. For this site, the Delhi study found no Google top-100 ranking and no Search Console impressions for "types of keywords" across the 90 days to 8 November 2026. That's not a failure of one page; it's a topic with no page at all, and a classification like this one is how you spot the hole.
Repeat the exercise each quarter. Your sorted list will shift as pages age, competitors publish and Google changes what it shows. A quarterly check takes an hour and keeps the plan honest.
FAQs
Fill out the form or call us — let’s find where your next growth comes from.
9659 634 634 [email protected]Free Growth Audit
See exactly where your traffic, leads and ROAS are leaking — and how to fix it. A senior Nurotech strategist reviews your website, SEO and ads. Free, no obligation.
No spam. We reply within 1 business day.