IntermediateLesson 12 of 305 min read

Keyword and Search Intent Research for Bloggers

Find the questions people really search for, match each to a search intent, and score ideas by volume, difficulty and fit with a short program on made-up sample data.

In this lesson you will

  • Name four kinds of search intent
  • Judge volume, difficulty and fit
  • Score ideas with a program
  • Find ideas without paid tools

About 24 minutes

A question people really ask

Priya notices that her post about seedlings falling over gets more visitors than her long essay on the history of allotments. Nobody told her why. The reason is simple: people type questions into search engines, and her post answers one. Keyword research means finding the words people really type, so you can write posts that answer them. It sounds technical, but at blog scale it is mostly careful thinking about your readers.

What a search phrase tells you

Every phrase carries an intent, which is what the person wants. Four kinds are worth knowing:

  • Learn: ‘how to start seeds indoors’. They want an explanation. Blog posts suit this best.
  • Compare: ‘compost bin vs heap’. They are choosing between options.
  • Shop: ‘buy tomato seeds’. They want to buy, and shops usually win these.
  • Broad: ‘allotment’. It is unclear what they want, and the competition is huge.

Match your post to the intent. A reader who typed a how-to question will not stay for a sales page, and a shopper will not read a long story. Matching intent is the single most useful idea in this lesson.

Three numbers per idea

For each idea, note three things. Searches a month is roughly how many people look for it. Free tools and the search engines’ own suggestions can give a rough sense, but treat any figure as an estimate, since tools disagree. Difficulty is how hard it is to appear, which depends on who already ranks. Judge it by looking at the first page of results yourself: if it is all large, famous sites, a new blog will struggle. Fit is how well the topic matches what you know and what your blog is about.

Score the ideas

The program below holds seven made-up ideas with those three numbers. It guesses the intent from the first words, then scores each idea as searches times fit, divided by difficulty squared. Squaring the difficulty is my choice to punish hard topics, and it is an opinion. You can change the formula, but write down what you chose and why.

# MADE-UP search ideas: (phrase, searches a month, difficulty 1-5, fit 1-5)
ideas = [
    ("how to start seeds indoors", 900, 3, 5),
    ("best spade for allotment", 300, 4, 3),
    ("allotment", 5000, 5, 2),
    ("compost bin vs heap", 250, 2, 5),
    ("buy tomato seeds", 700, 4, 2),
    ("why do seedlings fall over", 200, 1, 5),
    ("raised bed soil mix", 400, 3, 4),
]

def intent(phrase):
    words = phrase.split()
    if words[0] in ("how", "why", "what"):
        return "learn"
    if "vs" in words:
        return "compare"
    if "buy" in words or "best" in words:
        return "shop"
    return "broad"

rows = []
for phrase, searches, difficulty, fit in ideas:
    score = round(searches * fit / difficulty ** 2)
    rows.append((score, phrase, intent(phrase)))
rows.sort(reverse=True)
for score, phrase, kind in rows:
    print(f"{score:5}  {kind:8}{phrase}")

by_intent = {}
for score, phrase, kind in rows:
    by_intent.setdefault(kind, []).append(phrase)
print("learn ideas:", len(by_intent["learn"]))
print("write first:", rows[0][1])
print("biggest raw volume:", max(ideas, key=lambda r: r[1])[0])
 1000  learn   why do seedlings fall over
  500  learn   how to start seeds indoors
  400  broad   allotment
  312  compare compost bin vs heap
  178  broad   raised bed soil mix
   88  shop    buy tomato seeds
   56  shop    best spade for allotment
learn ideas: 2
write first: why do seedlings fall over
biggest raw volume: allotment

Look at the order. ‘Allotment’ has by far the biggest raw volume, 5,000 a month, yet it scores only 400, because it is the hardest and the least relevant to what Priya knows. The winner is ‘why do seedlings fall over’ on 1000: few people search it, but it is easy to rank for and a perfect fit for her. ‘How to start seeds indoors’ follows on 500. The shop phrases score lowest, 56 for the spade, since Priya is not a shop and the difficulty is high.

Two lessons follow. First, the biggest number is rarely the best target. A new blog does better with specific, low-competition questions it can answer well. Second, the score only ranks your ideas. It does not predict traffic, and the volumes here are invented.

Find ideas without a tool

  • Type the start of a question into a search engine and read the suggestions it offers.
  • Look at the ‘people also ask’ boxes and the related searches at the bottom of results.
  • Read the comments, emails and questions your readers already send you.
  • Browse forums and groups in your subject and note what people keep asking.
  • Check the search report of your own site once it has data, which lesson 17 explains.

Use one main phrase per post

Give each post one main phrase and a few close variations, and use them naturally in the title, the first paragraph and a sub-heading. Do not repeat the phrase mechanically, which reads badly. Also avoid writing two posts for the same phrase, since they compete with each other and confuse readers. If a new idea overlaps an old post, improve the old post instead.

Common mistake: aiming too big

Chasing high-volume phrases you cannot win. A new blog fighting for ‘allotment’ against established sites will wait a very long time. Start with narrow questions and widen as your blog earns trust.

Estimates, not promises

All volumes, difficulties and scores here are made up for teaching. Real tools give estimates that differ between tools and over time, and no research can guarantee that a post will rank.

List seven ideas for your blog. Give each searches, difficulty and fit, run the scoring, and pick your top two to write next.

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 search intent?

  1. How long the phrase is
  2. What the person typing the phrase wants
  3. How many ads show
  4. The site’s age

Intent is the goal behind the search, such as learning or buying.

Which intent suits a how-to blog post best?

  1. Shop
  2. Broad
  3. Learn
  4. None

People who want an explanation are happy to read a good guide.

Why did the biggest phrase not win the score?

  1. It was hard to rank for and a poor fit
  2. It had no searches
  3. It was too short
  4. It was misspelt

Difficulty and fit pulled its score down despite the volume.

Why is the score only a ranking of ideas?

  1. It is always exact
  2. It uses real data
  3. It does not predict real traffic
  4. It guarantees rankings

The numbers are estimates or invented, so the score only orders your options.

What should you do about two posts for one phrase?

  1. Improve one post instead of splitting
  2. Publish both
  3. Delete the site
  4. Copy the text

Two posts for one phrase compete with each other.

Summary

  • Match each post to the intent behind the phrase.
  • Score ideas on volume, difficulty and fit.
  • The biggest number is rarely the best target for a new blog.
  • Use one main phrase per post and avoid competing posts.

Next lesson

Lesson 13 turns a chosen phrase into a well-built post.

Questions

Do I need a paid keyword tool?

No. Search suggestions, related searches, reader questions and your own site data are enough to start. Paid tools help later, and their numbers are estimates too.

How accurate are search volumes?

Not very. Different tools disagree and the figures change. Use them to compare ideas with each other, not as exact counts.

Is the scoring formula official?

No. It is a simple model I chose to show the thinking. You can change the formula as long as you stay consistent.

Progress is saved in this browser.