Choose blog topics by starting with a question your intended customers need answered, deciding what useful evidence you can add, and checking whether the answer deserves a new article. Search data can help you investigate demand, but it shouldn't replace a clear reason for someone to read the page.
A topic is ready when you can name the reader, the decision they're trying to make, and the contribution your article will provide. If one of those is missing, spend the next hour clarifying the idea before adding it to the publishing calendar.
This guide gives you a way to make that choice, including a hypothetical furniture retailer's five-topic shortlist.
Define the reader's next decision
Start with a sentence: “This article will help [a specific reader] decide or do [a specific thing].”
“Office furniture” is a subject. “Help someone check the delivery route before ordering a desk” is a useful article brief. The second tells you what the reader needs to leave with: a process, the relevant measurements, and a way to resolve uncertainty.
Be specific about the situation without narrowing it arbitrarily. A person furnishing a small home office may need help choosing a desk, understanding product dimensions, or arranging delivery. Those tasks share a setting, but they don't necessarily need the same page.
Then write down how the question relates to your business. Can you answer it responsibly? Do you have access to relevant specifications, examples, subject knowledge, or someone who can verify the explanation? Would your intended audience still value the article if they found it through an email or a direct link?
Google's people-first content guidance asks publishers to consider their intended audience, the site's focus, and whether readers can achieve their goal. That gives you a useful editorial test before you look at a keyword list.
Find questions and keep track of where they came from
Keep each question together with the situation that made the answer necessary.
Collect questions from places where your audience is already asking for help: sales conversations, support requests, enquiries, site search, and feedback on existing pages. Preserve the wording and context in a private planning record; remove identifying details from anything you publish.
Record the source, date, situation, and what the person still couldn't decide. “Someone asked about delivery” is less useful than “A prospective buyer couldn't tell whether the desk arrives assembled or in separate packages.”
Keep observations separate from suggestions. A question raised by your team may be worth investigating, but label it as a hypothesis until you find support. Also distinguish one conversation from several independent conversations. Repeated messages in the same thread aren't evidence of several different people having the problem.
Search Console adds another view. Its Performance report shows clicks, impressions, queries, and pages for your site's appearances in Google Search. Start with a relevant page, examine the queries associated with it, and look for a question that the page doesn't fully answer.
Those queries describe your site's recorded visibility. They don't measure every search in the market. Google also omits anonymized queries and limits the query rows it shows, so an absent phrase doesn't establish that nobody needs the answer.
For seasonal ideas, Google Trends can help you compare relative search interest within a chosen time and location. Its values are normalized rather than raw search counts, and low-volume terms can appear as zero. Use it to investigate timing, alongside other evidence.
When the evidence is thin, the next task can be a small piece of research: ask a relevant colleague what makes the question difficult, review recent enquiries, or examine whether an existing page leaves the answer unclear. You don't need to turn every uncertainty into a published article immediately.
Check the need, the evidence, and the destination
Check the reader need, available evidence, and correct destination before ranking ideas.
Before ranking your ideas, give each one three checks. The checks below are an editorial decision method, not a search-engine scoring formula.
Is there a clear need?
Name the question and the consequence of leaving it unanswered. The consequence might be choosing an unsuitable option, abandoning a task, misunderstanding a limit, or repeatedly asking for the same clarification.
A broad trend can be relevant, but popularity alone doesn't describe a need. “People are discussing home offices” still leaves you to explain who needs help and with what.
If you can't identify the reader's task, hold the idea for further research. If you can, record the evidence and its limits. A few actual enquiries may justify a useful explanation without proving a large market.
Can you add evidence or useful analysis?
Write down the specific contribution you will provide: a documented example, a verified comparison, an annotated diagram, a worked calculation with stated assumptions, or an explanation of tradeoffs that generic summaries leave unclear.
“We'll research it” is a plan. “We'll show the relevant package dimensions and explain which delivery questions require confirmation” is a proposed contribution. You still need to gather and verify that material, but you know what the article depends on.
If the idea requires product testing, expert review, or data you don't have, put that requirement in the brief. Narrow the promise or gather the missing evidence before writing conclusions. Our guide to what makes a page worth citing explains how a clear claim and inspectable support work together.
Where should the answer live?
Check your existing pages before creating a URL. A question about one product's dimensions may belong on that product page. A delivery policy belongs where people can find the current policy. A reusable process may deserve a guide, with relevant product pages linking to it.
Two proposed titles can be different while promising the same answer. Compare the reader's task and the planned content, rather than just the wording.
Where an existing article already serves the task, improve it. Use our guide to choosing which blog posts to update first if your shortlist contains several existing pages needing attention.
A new article earns its place when it serves a distinct question and you can explain how someone would reach it from the rest of the site.
Work through a five-topic shortlist
A conceptual illustration for the hypothetical furniture retailer example.
Imagine a small furniture retailer planning its next article. Everything in this example—the business, questions, existing pages, and available materials—is hypothetical. It illustrates editorial choices; it doesn't report customer research or performance results.
Assume the retailer has documented product and package dimensions, an existing small-room desk guide, and a delivery policy. It hasn't carried out a material-comparison test.
| Proposed topic | Assumed clue or constraint | Editorial choice |
|---|---|---|
| How to check a desk's delivery route before ordering | Several separate enquiries ask how to evaluate doorways and turns; the retailer can supply verified package dimensions and obtain delivery-team review | Prepare a focused guide once the explanation and diagrams are reviewed |
| Does this desk arrive assembled? | The answer depends on the individual model and is missing from its product page | Add the verified answer to the product page |
| What happens if I need to change my delivery date? | Buyers need the current procedure, and the delivery policy leaves it unclear | Clarify the policy page |
| Is solid wood better than veneer? | A relevant comparison, but the team hasn't defined the materials, use conditions, or evidence for its claims | Research and narrow the comparison before committing to it |
| How to choose a desk for a small room | An existing guide already answers the intended task, with no specific new gap identified | Keep the existing guide; create a revision task if a gap emerges |
Only one candidate currently points toward a new article. Two require changes elsewhere, one needs evidence, and one is already covered. That is a productive content decision even though it produces fewer new posts.
For the delivery-route guide, the intended reader is someone considering an order. The article's contribution would be a diagram showing what information to collect, a distinction between product and package dimensions, and a clear explanation of when to contact the delivery team. It shouldn't promise that a single measurement proves an item can be delivered.
The existing small-room guide concerns choosing furniture for a usable workspace. The proposed delivery guide concerns getting an order into the building. Those are distinct tasks, provided the new article stays focused on the latter.
Nothing here establishes that the delivery topic has the largest search audience. Its advantage in this shortlist is a clear unresolved task, an appropriate relationship to the business, and a contribution the team could verify.
Turn the selected idea into a brief
Before drafting, save a short brief that answers these questions:
- Reader and task: Who needs this, and what should they be able to decide or do afterward?
- Evidence of need: Where did the question come from, and what remains uncertain?
- Direct answer: What can you responsibly tell the reader near the beginning?
- Contribution: What example, evidence, or analysis will make this page useful?
- Limits: Which claims need verification, and what will the article leave unresolved?
- Destination: Is this a new guide, an existing-page update, or a product or policy clarification?
- Connections: Which relevant pages should help readers reach the answer and continue their task?
For the fictional retailer, the brief would explicitly require delivery-team review before publication. Writing a confident first draft wouldn't satisfy that requirement.
Once the brief is sound, use an answer-first structure to deliver the promised help promptly. Plan the illustrations around explanations they can clarify, and use internal links that help readers continue the same task.
Keep the same standard for AI-search topics
If an idea begins with “we want AI to mention us,” translate it into the question you can help a person answer. Then apply the same need, evidence, and destination checks.
Google says its existing SEO practices remain relevant to AI Overviews and AI Mode, with no special optimization required for those features. It also makes clear that satisfying the requirements doesn't guarantee inclusion. This guidance concerns Google's search features; it doesn't establish how every other answer system selects sources.
A precise question and well-supported answer give the article a clear purpose. They don't entitle it to rankings or citations.
Choose what to publish next
Among ideas that pass all three checks, prioritize the unresolved task with the strongest connection to your audience and the most useful contribution you can deliver. Consider whether the answer is time-sensitive, how consequential a misunderstanding could be, and what evidence is ready.
Avoid turning that judgment into a precise score unless you have a reason for the weights. A high estimate of audience size shouldn't conceal a missing evidence requirement or an answer that belongs on an existing page.
After publication, return to the brief. Check whether the article actually answers the question, whether people still need the same clarification, and what relevant search queries it begins to receive. Treat these as different observations. A page that explains a delivery process clearly may be useful even while its search visibility is limited.
Keep a backlog of questions, their evidence, and the action each needs. That gives you a practical reason to choose the next article—and a reason to update a page or investigate a question when that would help more.
References
- Google Search Central: Creating helpful, reliable, people-first content.
- Google Search Console Help: Performance report overview and basic setup.
- Google Search Console Help: Dimensions and data groupings, including query limitations.
- Google Trends Help: FAQ about Google Trends data.
- Google Search Central: AI features and your website.