Ask a founder how they come up with blog topics and you'll hear some version of the same ritual: open a doc, stare at it, list ten things about the industry that seem interesting, pick the least intimidating one. It feels like strategy. It's actually guessing — and it's why so many startup blogs read like a diary of what the founder found fascinating that month.
The alternative isn't a $100-a-month keyword tool. It's noticing that topic research is mostly already done for you. Your buyers announce what they're searching for every time they ask a question — on a demo, in a support ticket, in a Reddit thread. Your job isn't to invent topics. It's to collect them.
Why does brainstorming blog topics fail?
Brainstorming produces topics that interest you; buyers search for whatever is blocking them, and the overlap between those two lists is smaller than it feels. You generate ideas from inside your product — its architecture, its roadmap, your hot takes on the industry. A buyer at the top of their research isn't there yet. They're typing the problem, in their own clumsy words, and the post that wins is the one shaped like their question, not your enthusiasm.
This is the quiet first cause behind most dead startup blogs. When we broke down why content marketing fails, cause one was publishing without a keyword strategy — writing what you wanted to say instead of what anyone was asking. This post is the fix for that cause: a repeatable way to source topics with proof of demand baked in, using nothing you don't already have.
The five free sources, ranked by intent
Where do the best blog topics come from? From places where customers already ask questions in their own words: your sales calls, your support inbox, your niche's communities, autocomplete, and competitors' thin pages. Ranked from highest buying intent to lowest:
Sales calls and demos
The questions and objections prospects raise before buying are the exact things other prospects type into Google and ChatGPT while deciding. We've published a full workflow for turning sales calls into content — record every call, extract each question and objection, and answer every one as its own post. It's the single richest source on this list; start there.
Support and onboarding questions
Post-purchase questions predict pre-purchase searches. Whatever confuses your customers in week one — "how does X integrate with Y," "can I migrate from Z" — is what your next customer is researching right now, before they've talked to anyone. Skim your last thirty tickets and onboarding emails; every recurring question is a topic with demand already proven.
Community threads in your niche
Reddit, industry Slacks, niche forums. The gift here is phrasing: a thread titled "is it worth paying for a CRM at 5 employees" isn't near a keyword, it is the keyword, verbatim, in the language buyers actually use. Bonus: the threads you answer well get pulled into AI search results — we've covered how Reddit answers become AI citations separately.
Autocomplete and People Also Ask
The free fifteen-minute method. Type a seed term — your category, your buyer's problem — into Google and note every autocomplete suggestion; add "how," "vs," "best," and "for" after it and note the new ones. Search the term, expand every People Also Ask box (each click loads more). Then ask ChatGPT what questions people in your buyer's role ask about the topic. You'll leave with twenty real queries.
Competitor content gaps
Read your competitors' blogs the way a skeptical buyer would. Wherever a page is clearly ranking — it's linked from their nav, it keeps showing up in your searches — but the content is thin, outdated, or generic, that's a validated topic with a low bar. You don't need their tool stack to see the gap; you need to be the better answer to a question they've already proven people ask.
Notice the ranking logic: the closer the source sits to a purchase decision, the better the topic. A sales-call objection comes from someone with budget in hand. An autocomplete suggestion might come from a student writing an essay. Use all five, but when the queue gets long, the top of this list wins.
The three-question filter
How do you know a topic is worth writing? Run it through three questions: would a buyer actually type this, can you say something non-generic, and does answering it lead anywhere commercially? A topic needs all three yeses.
- Would a buyer type this? Not "is this interesting" — would a real person, mid-problem, put words like these into a search box or an AI chat? Topics from sources one to three pass automatically, because someone already did. Brainstormed topics usually fail right here.
- Can we say something non-generic? If your answer would be identical coming from any company in your space, skip it or wait until you have a real opinion. Generic answers don't rank, don't get cited, and don't get remembered.
- Does it lead anywhere commercially? The reader who finishes the post should plausibly be a future customer. "How to fix the problem my product fixes" leads somewhere; adjacent trivia that happens to share your industry doesn't.
How do you turn one topic into a cluster?
Every good question implies a family: the question itself, the follow-ups a buyer asks next, and the comparison the answer eventually forces. One qualified topic is rarely one post — it's the seed of three to six.
Take "is it worth paying for a CRM at 5 employees." The core post answers it. The follow-ups are obvious once you look: what does a CRM cost at that size, when do spreadsheets stop being enough, what's the migration effort. And every "should I buy" question eventually becomes "which one" — a comparison post between the two or three options your buyers actually shortlist. Write the family together and each post reinforces the others: internal links flow naturally, and you cover the buyer's whole journey on one topic instead of scattering single posts across ten.
Keep a queue, not a list
Collection without discipline produces a junk drawer. Two habits turn these sources into a system. First, rank by recurrence: a question you've heard five times this quarter outranks a clever idea you've heard once, every time. Frequency is your free proxy for search volume — if your small sample of buyers keeps asking it, the much larger population of searchers is asking it too. Second, review monthly: thirty minutes to add what you've collected, promote what's recurring, and cut anything that no longer passes the three-question filter.
The ranked queue then feeds your publishing schedule directly — next month's calendar is just the top of the queue, in order. We've written a full guide to building a content calendar that picks up exactly where this post leaves off — or see how our content strategy builder turns a sourced topic queue into pillars and cadence automatically. The topics were never the hard part. They were sitting in your call notes and your inbox all along; the system is just the habit of picking them up.