The Question Map: Plan Around What Buyers and Their Agents Ask
Step 3 of 7 in The Definitive Guide to Marketing in the Age of AI (in Seven Steps). Originally published October 14, 2024. Rewritten for the AI era October 26, 2026.
An editorial calendar for the age of AI starts with questions, not dates. List the questions your buyers ask at each stage, and note whether a person, an AI assistant or an agent is asking. Give each question one owner page on your site, plus the outside sources you want answering it too. Then publish fewer, better pieces against that map.
A note on proof: Proof below predates generative AI; an AI-era result replaces it by mid-December.
Why does a calendar built on volume fail now?
Because volume is a commodity and the reader has changed. In HubSpot’s 2026 survey, 83.5% of marketers said they’re expected to produce more content (HubSpot). Every competitor can hit any publishing quota with AI. Nobody wins on quota.
Meanwhile, questions are exactly what AI answers first. Pew found that 60% of Google searches starting with “who,” “what,” “when” or “why” produced an AI summary (Pew). Seer Interactive found question-style queries triggered an AI Overview 85.9% of the time (Seer Interactive). If your buyers’ questions get answered before the click, your plan has to be built around being the source of those answers.
The old version of this guide proves the point. It was a calendar-driven series, and in 16 months the whole series earned 89 search impressions and no clicks. It answered how to run a content program. Almost nobody was asking that.
What is a question map?
It’s the persona-by-stage content matrix from the old calendar, turned on its side. Rows are buyer questions. Columns tell you who asks, where the answer should live and what proof it needs.
| Buyer question | Stage | Who asks | Owner page | Outside sources to earn | Proof needed |
|---|---|---|---|---|---|
| “How do we measure marketing without clicks?” | Problem | CMO, via AI assistant | Your measurement guide | Trade-press byline, podcast | A dated result with a number |
| “Who does this well for companies like ours?” | Shortlist | Buying group, via AI and agents | Your industry page | Analyst or association mention, client quote | Named clients in that industry |
| “What does it take to switch?” | Decision | Champion, then a rep | Your process page | Customer review sites | Timeline, team, risks addressed |
The example rows are illustrations. Build yours from real questions.
Where do the real questions come from?
From people who talk to buyers, and from the machines buyers talk to. Gartner’s survey of 645 B2B buyers found they used an average of seven information sources in a recent purchase (Gartner). Mine as many of those as you can:
- Sales calls and win/loss notes. The questions that come up right before a deal stalls are the ones to own.
- Customer support and onboarding. Post-sale questions are tomorrow’s pre-sale questions.
- Search Console queries. Long, question-shaped queries are a direct line to what people type.
- AI spot-checks. Ask your buyers’ questions in four engines and note what’s missing or wrong.
- RFPs and procurement forms. They tell you what a buying group must document.
Why does each question need one owner page?
Because AI quotes one passage, and Google picks one page. Two of your pages competing for the same question split whatever authority you have, and often the wrong one wins.
Give every question a single canonical page. Other pages can mention the topic, but they link to the owner instead of trying to rank for it. When you add a new piece, check the map first. If a page already owns that question, improve it instead of writing another one.
How do you plan for agents, not just people?
Assume part of the research will be done by software acting for the buyer. Google announced information agents that monitor the web for a user, and agentic booking for local services, in May 2026 (Google). An agent building a shortlist doesn’t read your story. It looks for facts it can match to its instructions.
So put the facts an agent needs on every shortlist-stage owner page, in plain text. Say what you do, for whom, where and how you work. Say what it takes to start and what proof you have. If an agent can’t find a fact, it can’t put you on the list.
What cadence works?
A rhythm tied to questions and proof, not to the calendar. One strong owner page a week or two, each improving or replacing a weaker page. One earned placement a month. A monthly spot-check to see which questions you now own. Fewer pieces, each with a job.
Checklist: build your question map
- 20–40 real buyer questions, collected from sales, support, search data and AI spot-checks
- Each tagged by stage and by who asks (person, AI assistant, agent)
- One owner page per question, with no two pages competing
- Agent-ready facts on every shortlist-stage page
- An outside source planned for each of your top ten questions
- A cadence you can sustain with quality, not a quota
Where to go next
Owning the question is half the job. The page also has to do the work the buyer came for. Chapter 4 covers content that does the buyer’s work, and their agent’s. If you want an outside read on which questions AI already answers for your buyers, and whether it names you, ask about the measurement and AI-visibility audit: work with me.
The calendar asks, “What should we publish next?” The map asks, “What will our buyer ask next, and who will answer?”
