How to Do Keyword Research Without Guessing
Keyword research goes wrong in a predictable way: people pick the terms with the biggest numbers and lose to sites forty times their size. Volume is the least useful column.
Useful keyword research picks terms by intent and winnability rather than by search volume. Start from what customers actually ask, group terms by the intent behind them, check who currently ranks to judge whether you can compete, then map one intent to one page so your own pages don't compete with each other.
Why Volume Is the Least Useful Number
Search volume tells you how many people search a term. It tells you nothing about whether those people want to buy, whether you could ever rank, or whether the traffic would be worth anything if you did. Sorting by it is how small sites end up targeting terms they cannot win.
A term with a thousand monthly searches that you will never rank for is worth zero. A term with thirty searches a month, where the searcher is ready to buy and the current results are weak, can be worth more than everything else on your list combined. The arithmetic is unintuitive only until you have watched it happen.
Volume figures are also estimates, rounded into buckets, and frequently wrong at the low end — which is exactly where a small site's opportunities are. Treating them as precise is a false confidence.
The two columns that actually matter are intent and winnability. Intent decides whether the traffic converts. Winnability decides whether you get the traffic at all. Volume only matters once both of those are satisfied, as a tiebreaker between options you could genuinely rank for.
Read the Intent Before Anything Else
Every query has a job the searcher is trying to do. Match the page to the job and ranking gets dramatically easier; mismatch it and no amount of optimisation helps, because Google is ranking pages that do a different job.
Informational — they want to understand something
"how long does SEO take", "what is schema markup". These want a guide or an explanation. A service page targeting them will not rank, because Google can see that searchers want to learn rather than to buy.
Commercial — they are comparing options
"wordpress vs custom website", "best CRM for small business". These want a comparison with the trade-offs left in. Enormously valuable because the searcher is close to deciding, and reliably under-served because most businesses will not write honestly about alternatives.
Transactional — they want to buy or hire now
"web development company in Rohini", "emergency plumber near me". These want a service page with proof, contact details and location. Highest value, highest competition, and the terms most worth building landing pages for.
Navigational — they want a specific site
Your brand name, or a competitor's. Worth ensuring you own your own; rarely worth chasing anyone else's.
A Method You Can Run This Week
This does not need paid tools to start. The first three steps use free sources and produce a better list than most tool-driven research, because they start from real language rather than from a database.
- 01
Write down what customers actually ask
Go through your enquiry emails, WhatsApp messages and call notes from the last three months and record the questions in the customer's own words. This is the highest-quality keyword source you have and almost nobody uses it. It is also the only source that reflects your specific market rather than a national average.
- 02
Mine Google's own suggestions
Type your core terms into Google and read the autocomplete, the People Also Ask box and the related searches at the bottom. These come from real query data. People Also Ask is particularly valuable because each question is a heading you could answer directly.
- 03
Check what competitors already rank for
Look at the pages ranking for your core terms and read their headings. You are looking for the sub-topics they cover, which tells you what Google currently considers a complete answer on the subject.
- 04
Judge winnability by reading the results
For each candidate, search it and look at the top five. Are they national brands or businesses your size? Are the pages thorough or thin? This manual read is more reliable than any difficulty score, because it tells you what you would actually have to beat.
- 05
Group by intent, then map one group to one page
Terms that want the same answer belong on the same page. "seo cost", "seo pricing" and "how much does seo cost" are one page, not three. This grouping is what prevents cannibalisation later, and it is easier to do now than to untangle afterwards.
One Intent, One Page
Keyword cannibalisation is what happens when several of your pages target the same intent. Google has to choose between them, often picks the weakest, and splits the internal links and external signals that should have concentrated on one strong page.
It usually develops innocently. A blog post covers a topic, then a service page covers it again, then someone writes a second post from a slightly different angle. Each was reasonable in isolation; together they compete. The symptom is a term where your ranking page keeps changing, or where a post outranks the service page you actually wanted to rank.
The fix is consolidation rather than deletion. Pick the page that should own the intent, merge the useful content from the others into it, and redirect the rest to it permanently. The consolidated page usually outranks all the originals within weeks, because everything that was split is now pointing at one target.
Preventing it is simply a discipline: before writing anything new, check whether you already have a page for that intent. If you do, improve that page instead. This single habit removes most cannibalisation before it starts.
Common Questions
What comes up when people start doing this properly.
How many keywords should one page target?
One intent, which usually means one primary term and a handful of natural variations. Pages written to hit a list of loosely-related terms read badly and rank for none of them well. Google matches on meaning, so covering the topic thoroughly captures the variations without forcing them in.
Are long-tail keywords still worth it?
For a small or new site they are usually the only realistic starting point, and they convert better because they are more specific. The strategy is to win a cluster of specific terms first, which builds the topical authority that eventually makes the broader terms reachable.
Should I target the same keywords as my competitors?
Study them, but do not copy the list. A larger competitor's terms may be unwinnable for you, and their list reflects their business rather than yours. The useful signal is the sub-topics they cover, which tells you what a complete answer looks like.
How often should keyword research be redone?
A proper pass once or twice a year, plus ongoing capture of real customer questions as they arrive. Search behaviour shifts slowly in most categories, so the bigger risk is a stale list nobody revisits rather than the language changing underneath you.