Keyword Research for Beginners
Find and choose keywords with free methods and a transparent scoring formula, and learn why difficulty and relevance matter as much as volume.
In this lesson you will
- Find keyword ideas for free
- Weigh demand, difficulty, relevance
- Score ideas with a formula
- Give each page one main phrase
About 16 minutes
Priya plans a blog
Priya wants to start a blog about cheap meal prep. She opens a notes file and writes ‘meal prep’ at the top. Then she freezes. Is that the right phrase? Should she write about containers, recipes or weekly plans? Thousands of people write about meal prep already. Keyword research is the habit of choosing your topics with evidence instead of hope. It does not need expensive software, and it saves you from writing pages nobody searches for.
What a keyword is
A keyword is a word or phrase that people type into a search engine. It can be several words long, and the name is slightly misleading. You are not trying to collect words. You are trying to understand the questions and needs of real people and decide which ones your site can answer well.
Where to find ideas for free
- Your own knowledge. Start with the questions customers or friends ask you. Write down real wording.
- Search suggestions. Type a phrase into a search box and look at the suggestions that appear as you type. They are drawn from what people search. They vary by place and over time.
- Related questions and related searches. Results pages often show questions people also ask and searches related to yours. Treat them as prompts for pages.
- Search Console. Once your site has some history, it lists the actual searches that showed your pages. This is often the best source of all.
- Free or paid keyword tools. Many show estimated monthly searches and a difficulty score. Each tool has its own scale and its numbers are estimates, so compare ideas inside one tool and do not trust the exact figures.
The four things to weigh
- Demand: roughly how many people search this? (A tool estimate.)
- Difficulty: how hard is it to compete? (A tool score, plus your own look at the results.)
- Relevance: can my site genuinely help with this? (Your honest judgement.)
- Intent: does the searcher want what I offer? (Lesson 2.)
Demand alone misleads. A phrase searched nine thousand times is no use if every result is from a giant brand and your site is new. A phrase searched four hundred times that matches your offer exactly can bring a steady stream of the right people.
Score ideas with a transparent formula
Below is a practice list for Priya. The numbers are made-up sample data so you can see the method. They are not real search figures. The code gives each keyword a score using a simple formula: searches multiplied by relevance, divided by difficulty plus ten. Adding ten stops the score from exploding when difficulty is very low. This is a teaching formula that Skills in Minutes made up for practice, not an industry standard. Its value is that you can see exactly what it does.
# SAMPLE numbers for practice. They are NOT real search data.
# keyword, monthly searches, difficulty 0-100, relevance 1-5, intent
rows = [
("meal prep ideas", 9000, 70, 3, "informational"),
("cheap meal prep for one", 1300, 25, 5, "informational"),
("meal prep containers", 6000, 65, 2, "commercial"),
("budget meal prep for a week", 900, 20, 5, "informational"),
("how to meal prep chicken", 2400, 40, 4, "informational"),
("meal prep delivery", 3000, 80, 1, "transactional"),
("student meal prep on a budget", 400, 15, 5, "informational"),
("meal prep recipes", 7000, 75, 3, "informational"),
]
def score(volume, difficulty, relevance):
return volume * relevance / (difficulty + 10)
ranked = sorted(rows, key=lambda r: -score(r[1], r[2], r[3]))
print("All keywords:")
for kw, vol, diff, rel, intent in ranked:
print(f" {score(vol, diff, rel):6.1f} {kw} (difficulty {diff})")
# a brand-new site cannot win hard keywords yet
easy = [r for r in ranked if r[2] <= 40]
print("New-site shortlist (difficulty 40 or less):")
for kw, vol, diff, rel, intent in easy:
print(f" {score(vol, diff, rel):6.1f} {kw}")
All keywords:
337.5 meal prep ideas (difficulty 70)
247.1 meal prep recipes (difficulty 75)
192.0 how to meal prep chicken (difficulty 40)
185.7 cheap meal prep for one (difficulty 25)
160.0 meal prep containers (difficulty 65)
150.0 budget meal prep for a week (difficulty 20)
80.0 student meal prep on a budget (difficulty 15)
33.3 meal prep delivery (difficulty 80)
New-site shortlist (difficulty 40 or less):
192.0 how to meal prep chicken
185.7 cheap meal prep for one
150.0 budget meal prep for a week
80.0 student meal prep on a budget
Look at the first list. The top score goes to meal prep ideas at 337.5, mainly because of its large search number. But its difficulty is 70 out of 100. For a brand-new blog that is a long shot. So the second list keeps only phrases with difficulty of 40 or less. Now how to meal prep chicken leads with 192.0, followed by the cheap and budget ideas that suit Priya’s theme exactly.
This is the real skill. The formula did not make the decision. It organised the evidence, and Priya’s judgement (difficulty cut-off, relevance scores) shaped the result. Change the cut-off and the shortlist changes. That is fine. The aim is to make your choices visible and easy to revisit.
One main keyword per page
Give each page one main topic and one main phrase. If two pages on your site compete for the same phrase, search engines have to choose between them, and neither may do well. This problem is often called keyword cannibalisation. A simple map helps: a table with one row per page and its main phrase. If you find a phrase appearing twice, merge the pages or give one a different focus.
Picking keywords only by volume. The biggest numbers usually belong to the hardest competition and to searches that are less specific. Weigh difficulty and relevance too.
Search volume and difficulty figures are estimates and differ between tools. Never quote them as facts, and never promise a client or yourself that a keyword will bring a set number of visitors.
Practice for this lesson
Check what you learned
Five quick questions. Pick an answer and you will see straight away whether it is right and why.
What is a keyword in SEO?
- A secret code for Google
- A phrase people type into search
- A word hidden in a page
- A paid advert
A keyword is the wording of a real search. The job is to understand the need behind it.
Why can volume alone mislead you?
- Big numbers are always wrong
- Volume cannot be measured
- Low volume is always better
- High volume often means tough competition
Popular phrases usually have strong competition, which a new site may struggle to beat.
In Priya’s sample data, why filter out difficulty above 40?
- Hard terms are banned
- A new site cannot win hard terms yet
- The tool requires it
- Easy terms pay more
The cut-off reflects Priya’s judgement that a brand-new blog should start with winnable phrases.
What are the sample keyword numbers in this lesson?
- Real figures from Google
- Official difficulty scores
- Live search results
- Made-up practice data
They were invented to teach the method and must not be treated as real search data.
What is keyword cannibalisation?
- Two pages competing for one phrase
- Using too many images
- Deleting old keywords
- Copying a rival page
When two of your pages target the same phrase, they compete with each other instead of working together.
Summary
- A keyword is the wording of a real search and the need behind it.
- Weigh demand, difficulty, relevance and intent together.
- New sites should start with winnable, specific phrases.
- Give each page one main phrase to avoid cannibalisation.
Questions
Do I need a paid keyword tool?
No. Search suggestions, related questions, your own knowledge and Search Console are enough to begin. Paid tools save time later, but their numbers are estimates.
How many keywords should one page target?
One main topic and phrase, plus the natural variations people use. Do not force a list of unrelated phrases into one page.
What is a good difficulty score?
There is no universal answer, because every tool uses its own scale. Compare ideas within the same tool, and look at the real results to judge whether you can compete.
