Phase 1 · StrategyModule 3 of 15
Pick keywords you can actually win
The number in your keyword tool is not a measurement. Once you know that, the job changes from finding big terms to finding winnable ones, which is a different search entirely.
What this actually costs you
- Hours to learn
- 8
- Hours / month
- 2 to 3
- Tools
- $29 to $139/mo
- Difficulty
- Medium
What this looks like when it is done badly. Someone exports two thousand keywords, sorts by search volume, and writes ten posts targeting the terms at the top. None of them rank, because the top of a volume-sorted list is by definition the part everyone else also found first.
Hours and tool costs are RedSEO’s own estimates from client work and from teaching this material, not industry survey data. They assume a business owner doing the work themselves on a site under about 50 pages.
Keyword research has a reputation as the analytical part of SEO, which is probably why so many people enjoy doing it and so few people finish. It is mostly a filtering exercise, and the filter that matters is not the one everybody uses.
The number in your keyword tool is not a measurement
Start here, because it changes how much weight you put on everything downstream. Search volume is not counted, it is estimated and then rounded. Google’s own documentation for Keyword Planner says it plainly: “Your search volume statistics are rounded. This means that when you get keyword ideas for multiple locations, the search volumes might not add up as you’d expect.” The figure is an average over twelve months, for exact matches only.
Every third-party tool builds its own estimate on top of that, or on clickstream data, or both. So when one tool says 1,300 and another says 880, neither is lying and neither is right. They are estimates of a quantity nobody publishes.
The practical consequence: treat volume as an order of magnitude, not a number. The difference between 50 and 5,000 is real and worth acting on. The difference between 880 and 1,300 is noise, and choosing between two terms on that basis is a decision made on nothing.
Volume is the least important of the three numbers
Every term has three properties worth knowing, and most people rank them in exactly the wrong order.
Intent is first, because it decides what kind of page can win and therefore whether you are even in the right competition. Difficulty is second, because it decides whether you can win at your current size. Volume is third, because it only tells you the size of a prize you may not be eligible for.
Sorting a list by volume, which is what every export defaults to, puts the terms you are least likely to win at the top. That single default is responsible for more wasted content than any other thing in this curriculum.
Intent decides the format before you write a word
Search intent is what the person actually wants, and Google has already worked it out. You do not need to guess: the current page one is Google’s published answer to what satisfies that query.
So search the term and look at the ten results. If they are all product or category pages, the intent is transactional and an article will not win, no matter how good the article is. If they are all guides, a service page will not win. If they are a mix, the query is ambiguous and there is room for either.
This is the check that takes thirty seconds and the one people skip because it feels too simple to be the answer. It is the answer. Every large content failure I have been asked to diagnose has the same root: forty posts written against terms whose page one was entirely commercial.
Why the long tail is where a small site starts
The shape of search demand is not intuitive. Ahrefs, working from their US keyword database, report that keywords with fewer than 10 searches a month account for almost 93 percent of it, some 2.3 billion terms, while just under 18,000 keywords have volumes above 100,000 a month.
Keywords under 10 searches a month are 2.3 billion of the terms in the database. Fewer than 18,000 sit above 100,000 a month. The winnable list looks disappointing precisely because the unwinnable one is so small.
That distribution is why “target long tail keywords” is such tired advice and also why it is correct. Nearly all of the available demand is in terms too small to appear on anybody’s priority list, which is precisely what makes them winnable. Twenty terms at 40 searches a month is 800 searches you can realistically rank for, against one term at 5,000 that you cannot.
There is a second reason, and it is the one I find more persuasive. Ahrefs also cite research that around 15 percent of daily Google searches are new and have never been searched before, a figure Google reconfirmed in 2025. You cannot do keyword research on a query that has never been typed. What you can do is cover a topic properly enough that you turn up for the versions of it nobody predicted, and that only happens if you are writing for a subject rather than for a string.
Have us do the research and hand you the map
Two failures worth naming
Chasing zero-volume terms because they are easy. The long tail argument gets stretched until people target things nobody searches at all. A zero-volume keyword is worth targeting only when you have a specific reason to think the tool is underreporting, which is common for genuinely local and genuinely new phrasings and rare otherwise.
Building two pages for one intent. If two of your pages could reasonably answer the same query, Google picks one and often picks badly, and both underperform what one good page would have done. That is keyword cannibalisation, it is created at this stage, and it is fixed in module 4. Grouping terms into one row per page before you build anything is what prevents it.
How I actually pick
For a small site with little authority, in this order. Terms where I already get impressions on page two, from the baseline in module 2, because Google has already told me I am close. Then terms where page one contains at least one site roughly my size, which is the only difficulty score I fully trust. Then terms where the intent matches a page I already have, because updating beats publishing.
Only after all three of those do I look at anything purely because the volume is attractive. By then the list is usually full, which is the point.
Do it on your own terms
Three tools and a spreadsheet. The template is the one I actually use, with the intent and grouping columns already in it.
The tools for this module
Free, no signup. Open each in a new tab and work through them in this order.
- Keyword Research AssistantExpands a seed term into the phrasings you would not have thought of yourself.
- Keyword Opportunity ScoreScores a term on volume against difficulty, which is the sort order that actually matters.
- Keyword Research TemplateThe spreadsheet, with intent, grouping and cannibalisation columns already set up.
Build a keyword list you can actually act on
The output is a spreadsheet with one row per page you intend to build, not a list of two thousand terms. If a row does not name a page, it is not finished.
Write down what you sell, in the words customers use
Not your internal product names. What somebody would type who does not know your company exists. Ask three recent customers what they searched for. It is the cheapest research in this module and almost nobody does it.
Expand each of those into a list with a tool
Any keyword tool will do this. You are looking for the phrasings you would not have thought of, not for the biggest number on the screen.
Search each candidate yourself and look at what ranks
This is the step people skip and it is the one that decides everything. If page one is all product pages, a blog post cannot win it. If page one is all guides, a service page cannot win it.
Label every term with its intent before you look at volume
Informational, commercial, transactional, or navigational. Doing this before you see the volume stops the big number talking you into the wrong content type.
Cut anything where page one is entirely major brands
Not because you should never compete, but because it should not be in your first twenty pages. Come back in a year when the site has some authority behind it.
Group what survives into one row per page
Terms that would be answered by the same page belong on the same row. This is the step that prevents two of your pages competing for one term later, which is the most common self-inflicted wound in this curriculum.
Assign each row a page that exists or a page you will build
A URL. Real or planned. A keyword with no page attached is a note, not a plan, and it is where keyword research quietly turns into procrastination.
Sort by winnable rather than by volume, and stop at twenty
Twenty rows is more than most small businesses will get through in a year. A list of four hundred is a way of avoiding starting.
Can you do this yourself?
You can do this yourself if
- You can look at a page of Google results and describe what all ten have in common. That is the entire skill, and it is learnable in an afternoon of actually doing it.
- You sell something specific enough to have specific searches attached to it. Plumbers, dentists, accountants and manufacturers all qualify.
- You are willing to target terms that look too small. The winnable list almost always looks disappointing next to the list you wanted.
You cannot do this yourself if
- You cannot tell commercial intent from informational intent, and will write blog posts for terms that needed a service page. That mistake is invisible for about forty posts and then it is a year gone.
- Your market is genuinely new and the searches do not exist yet. SEO is demand capture. If nobody is searching, this is the wrong channel and no amount of research fixes it.
- You need volume at national scale in a market where page one is Amazon, Wikipedia and three companies with funding. Not impossible, but not a first-year DIY project.
Questions people ask about this
- How do I choose keywords I can actually rank for?
- Sort by winnable, not by volume. Start with terms you already get page-two impressions for, then terms where page one contains at least one site roughly your size. Search every candidate yourself before committing: the current page one is Google's published answer to what satisfies that query, and if none of the ten results look like the page you intend to build, you are in the wrong competition.
- Is search volume data accurate?
- No, and it is not meant to be. Google's Keyword Planner documentation says search volume statistics "are rounded", averaged over twelve months and exact-match only. Third-party tools estimate on top of that. Treat volume as an order of magnitude: 50 versus 5,000 is meaningful, 880 versus 1,300 is noise.
- What are long tail keywords and should I target them?
- Terms with low individual search volume, usually longer and more specific. Ahrefs report that keywords under 10 searches a month make up almost 93% of their US database, around 2.3 billion terms, against just under 18,000 above 100,000 a month. Nearly all available demand sits in terms too small for anyone to prioritise, which is exactly what makes them winnable for a small site.
- How do I tell what search intent a keyword has?
- Search it and look at what already ranks. If page one is all product or category pages, the intent is transactional and an article cannot win it. If it is all guides, a service page cannot win it. This takes thirty seconds and it is the check that decides whether the rest of the work is aimed at the right target.
- How many keywords should I target?
- Twenty rows, where each row is one page. That is more than most small businesses get through in a year. A list of four hundred keywords is usually a way of avoiding starting, and it guarantees that terms which should share a page end up on separate ones.
- What is keyword cannibalisation?
- Two or more of your own pages competing for the same query, so Google picks one and both underperform. It is created during keyword research, when terms that should have shared a page get separate rows, and it is fixed in module 4. Grouping terms by the page that would answer them prevents it.