Hub › Lessons › Content planning, lesson 6
Check a planned title against what searchers want, before you write
You can write a good answer to a question and still choose a title that brings the wrong readers. People who type your title into Google may want something else: a how-to guide instead of an opinion, or a price list instead of advice. This lesson shows you how to check, in about 20 minutes, before you write. You look at the real results yourself, write one line about each, and use a prompt to compare them with what your post will cover. It does not predict rankings or traffic. Nobody can promise those.
What you will be able to do: say whether a planned title is a Match, Partly or Different from what searchers want, based on results you looked at yourself, and pick a clearer title that only promises what your post covers. You also get the searchers' questions your post does not answer, so you can decide whether to add them.
Step 1: Write down the title and the searches (3 minutes)
- Your planned title, as you would publish it.
- What the post will cover, in 2 or 3 lines: who it is for, your answer and your offer. If you made a brief with lesson 2, copy it from there.
- One to three searches a customer might type. Write them the way people type, in lower case and without your brand, such as "should i put prices on my cafe website menu". Your title and the search do not have to match word for word.
Step 2: Look at the real results and take notes (10 minutes)
Google says that to return relevant results it first works out "the intent behind your query". You are doing the same thing by hand.
- Search in your country's Google, such as google.com.sg, in a private or incognito window so your own history matters less. Write down the date.
- Skip ads. Results marked "Sponsored" are paid for and do not show what searchers want.
- Write one line for each of the first 8 to 10 results: the type of page (guide, discussion, product page, video), who wrote it (a company, a forum, a blogger) and its main answer or angle. Write each line as a full sentence; short notes are easy to misread later.
- Copy the "People also ask" questions and any related searches you see.
- Look for "Missing:" under a result. It means Google showed the page without one of your words, which is a sign searchers may mean something broader.
- Do not copy other pages' titles or text. You are learning what readers want, not borrowing someone's work.
Step 3: The check prompt
Fill in each line, add one line per result under the notes, then paste it into your AI tool. The AI only sees your notes, so the check is as good as they are.
Help me check whether a title I plan to write matches what people searching for it want. Use only my notes below. My planned title: What my post will cover (2 or 3 lines): The searches I tried: My notes on the search results (one line per result: type of page, who wrote it, its main answer or angle): - "People also ask" or related questions I saw: - Return these parts: 1. What searchers seem to want: 1 to 3 points. After each point, give the numbers of the result notes that support it. 2. Verdict: Match, Partly or Different, judged on the question the results answer, not on their answer. - Match: more than half of the results answer the same question my post answers. - Partly: some results answer my question, but half or fewer. - Different: no more than one result answers my question. Give one sentence explaining the verdict with result numbers. Then one sentence on how my answer compares with the answers in the results. A different answer is fine; just say so. 3. Title options: up to 3 titles for the post I described. Each must be something my post will actually cover. After each, say in a few words which searcher need it fits. 4. Questions my post does not answer: questions from my notes that searchers ask and my post does not cover. I will decide whether to add them. 5. Not checked: one line on what this check cannot tell me. Rules: - Use only my notes. Do not add results, websites, facts or numbers that are not in my notes. - Do not predict rankings, traffic, clicks, search volume or how hard it is to rank. - Do not suggest a title that promises something my post will not cover. - If I gave notes on fewer than 5 results, say that the notes are too few to judge and stop after part 1.
Step 4: Decide what to do (5 minutes)
- Match: write the post. If your answer differs from most results, say so early. A different, honest answer gives readers a reason to choose your page.
- Partly: make the title name the question you answer, so the right readers click. Keep the other question as an idea for another post, if you can answer it from your own experience.
- Different: change the title to the question your post really answers, or pick another topic for search. You can still publish the post for people you send it to directly.
- Read every title option before you use it. Delete any that promises more than your post covers.
Worked example
We checked two planned titles for the same made-up web designer from lesson 5. The business and its posts are illustrative. The search results are real: we searched google.com.sg in English on 29 October 2026 and wrote one line per result, in our own words. Results change over time and by location, so yours will differ.
Input: notes for "Should your menu page show prices?"
My planned title: Should your menu page show prices? What my post will cover (2 or 3 lines): For small cafes and home food businesses in Singapore that have, or are planning, a website menu page. My answer: yes, show prices if you can, because most people check the price online instead of calling. Price changes are included with my basic site, so changing prices is not a reason to hide them. Offer: tell me what you sell and I will tell you how I would lay out the page. The searches I tried: should i put prices on my cafe website menu (google.com.sg in English, 29 Oct 2026) My notes on the search results (one line per result: type of page, who wrote it, its main answer or angle): 1. Reddit small business discussion started by someone who builds restaurant websites: show prices or say "contact for a quote"? Top answer: the writer says they leave any website that does not list prices. 2. Guide by a menu design company about how to set cafe menu prices (Google showed "Missing: website"). 3. Quora question: why do some restaurants show menus without prices on their website? Top answer: some very high-end restaurants do this on purpose. 4. Guide by a digital menu company: cafe menu pricing, menu engineering and pricing psychology. 5. Article by a web designer: what every cafe website needs. Answer: yes, show prices; also mentions adding menu items and prices to a Google Business Profile. 6. Facebook food group discussion: why don't restaurants list prices on their websites? Top answer: no prices makes them assume it is expensive. 7. Blog by an online ordering company: how to set menu prices using food cost. 8. Blog by a restaurant booking software company: pricing psychology for menus. 9. Guide by a food business site: how to price your restaurant menu, with food cost percentages. "People also ask" or related questions I saw: - What is the ideal pricing strategy for a cafe? - What is a reasonable price to charge for a website? - How do I price my menu? - How many items should be on a cafe menu?
Output (as returned)
Shown word for word from our test with Claude; only the formatting was converted for this page.
1. What searchers seem to want
- A yes/no answer on whether a cafe or restaurant should list prices on its website. Supported by 1, 3, 5, 6.
- To know how visitors react when prices are missing. They assume it is expensive, or they leave. Supported by 1, 6. Result 3 gives the one stated exception: some very high-end restaurants hide prices on purpose.
- Help working out what to charge: pricing strategy, food cost, pricing psychology. Supported by 2, 4, 7, 8, 9. The related questions "ideal pricing strategy" and "how do I price my menu" point the same way.
2. Verdict: Partly
Only 1, 3, 5 and 6 deal with whether to show prices on a website, which is 4 of 9. Results 2, 4, 7, 8 and 9 answer how to set prices, which is a different question. Even if you count only 1 and 5 as the same question, the verdict stays Partly.
Your answer (yes, show prices) agrees with 1 and 5, and 6 backs it by saying missing prices make people assume it is expensive. Result 3 is the only one that gives a reason to hide prices, and it applies to very high-end places.
3. Title options
- Should a cafe website menu show prices? Yes, and here's why. It fits searchers who want a direct yes/no (1, 5).
- Should your menu page show prices? A guide for small cafes and home food businesses in Singapore. It fits owners who want an answer for their size of business.
- Worried about changing prices? Why that's no reason to hide them on your menu page. It fits owners who hesitate over keeping prices up to date.
4. Questions your post does not answer
- How do I price my menu?
- What is the ideal pricing strategy for a cafe?
- How many items should be on a cafe menu? Your offer covers page layout, but the post doesn't say how many items to list.
- What is a reasonable price to charge for a website? Your post says price changes are included with your basic site, but it doesn't give a website price.
- When is it fine to hide prices, or use "contact for a quote"? (From results 1 and 3.)
- Should menu items and prices also go on a Google Business Profile? (From result 5.)
5. Not checked
This check uses one search (google.com.sg, English, 29 Oct 2026) and your one-line notes on 9 results. It can't tell you what the pages actually say beyond your notes, or what other wordings of the question would show.
What we checked
- The verdict follows the rule. We counted the notes ourselves: 4 of the 9 results are about showing prices on a website, and 5 are about setting prices. Four of nine is half or fewer, so "Partly" is right.
- Every point points back to a note. A short script confirmed that every result number cited is between 1 and 9, and that the output has no number that is not in our notes.
- No ranking talk. The script found no mention of rankings, traffic or search volume in parts 1 to 4.
- The titles stay inside the post. We read all three against the post summary. Each promises only the yes answer, the audience or the price-changes point, all of which the post covers.
What we would do with this: the post answers "should prices be shown", but more than half of these results are about "how to set prices". Title option 1 or 3 makes the question clearer. "How do I price my menu?" is a different post, and the web designer should only write it if they can answer it from real experience.
The other title we checked
For "Do I need a website if I already have Instagram?" (search: "do i need a website if i have instagram", same day), all nine results we noted answered that question, and both test runs gave Match. Almost every result was written by a web design or website company, and most said yes, you need a website. The planned post says Instagram may be enough until customers keep asking the same questions. That different answer is fine; both outputs suggested a title that shows it, such as "Instagram or a website for your home bakery or cafe: when DM orders are enough".
What we changed after testing
The prompt above is the third version. We ran each version twice on both titles:
- Version 1 gave "Partly" in one run and "Match" in the other for the same Instagram notes, because it did not say what the verdicts meant. We defined them by the question the results answer, not by their answer.
- Version 2 still gave "Different" once and "Partly" once for the menu notes, which split 4 to 5. The definitions overlapped. Version 3 uses counts: more than half, half or fewer, or no more than one. All four runs of version 3 with our final notes gave the same verdicts.
- Our notes, not just the prompt. One run said our note for result 1 was unclear ("leaves any website without prices"). We rewrote it as a full sentence, which is why Step 2 asks for full sentences.
- The stop rule works. With notes on only 2 results, every version said the notes were too few and stopped after part 1.
Watch out for
- This is not keyword research. It does not tell you how many people search, or how hard a search is. Google's own starter guide says "There are no secrets here that'll automatically rank your site first in Google".
- Results change. Google notes that many searches have a local intent, so results differ by place, and they also change over time. Note the date and check again before you update an old post.
- Do not let the AI do the searching for you unless you can see the results it used. An AI can describe search results that do not exist. Your own notes are the evidence.
- Some "People also ask" questions are off-topic, such as Instagram rules in a website post. Skip them.
- Search is one reason to write, not the only one. A post you send to customers who ask the same question can be useful even if few people search for it.
Next step
Next: a 10-minute weekly content review
Lesson 7 of this series brings lessons 1 to 6 together in one short weekly check.
Go to lesson 7Turn the checked title into a brief
Lesson 2 of this series turns one topic into a one-page brief. Use your new title and the questions from part 4 that you decided to answer.
See lesson 2How we checked this lesson
- On 29 Oct 2026 (Singapore time) we searched google.com.sg in English for the two searches shown and wrote one line per result for the first nine ordinary results, plus the "People also ask" questions. The web designer, the posts and the planned titles are illustrative.
- We ran the prompt filled in with those notes using Claude Code 2.1.312 in print mode. The output is shown as returned, with formatting adjusted for this page. We ran each prompt version twice per title, and a third input with notes on only two results to test the stop rule.
- A separate script checked every output: verdict, result numbers between 1 and 9, no numbers missing from the notes, and no ranking, traffic or search-volume words before part 5. We also read every output against the notes.
- Quotes from Google are from its SEO Starter Guide (last updated 10 Dec 2025) and How Search works: ranking results, both checked on 29 Oct 2026.
- We tested one assistant on two titles. AI output varies by tool, model and input, so read the result each time.
- Last reviewed: 29 October 2026.