B2B content strategy: map content to buying decisions
Build a B2B content plan from the questions that slow buying decisions: problem recognition, option evaluation, internal approval and implementation. A worked content map and a blank template.
The short version
- A calendar schedules work; a strategy decides which work deserves scheduling, and what you will not publish.
- Collect real buyer questions from sales, support and search, and label the ones you inferred.
- Group questions by the decision they support. Internal approval and implementation are often under-served.
- Use AI to cluster questions, then separate different intents by hand. Every piece needs evidence, a next step and a signal to watch.
On this page
A content calendar can be full for a year and change nothing about how customers buy. That usually happens when the plan starts from a publishing rhythm (“two posts a week”) or a keyword export, and only later asks who the pieces are for and what they should help that person decide.
A B2B content strategy works the other way round. Start from the questions that slow a buying decision, the ones sales hears on calls and buyers ask each other before they ever speak to you. Decide which of those questions your content should answer, in what format, with what evidence, and where each piece should lead. Search data helps you find and phrase the questions. AI helps you organise them. Neither decides what is worth publishing.
A strategy is not a calendar
A content calendar schedules work. A content strategy decides what work deserves scheduling. For a B2B team, a strategy should settle:
- The audience: which accounts and roles, ideally from your ideal customer profile.
- The business objective: more qualified enquiries, shorter sales cycles, better-informed prospects, fewer repeated questions for sales.
- The role of content: what content does that other activities cannot, such as answering a question at 11pm, or giving a champion something to forward to finance.
- What you will not publish: topics outside your positioning, formats you cannot sustain, and articles whose only reason to exist is a keyword.
The last point is the easiest to skip. Saying no is what makes a content plan a strategy.
Collect the questions that delay decisions
The raw material for the plan is a list of real questions. Collect them from as many places as you can reach:
- sales call notes and recordings, especially the questions asked more than once
- discovery and proposal documents, including deals you lost
- support tickets and onboarding questions from new customers
- your site search, if you have one
- Search Console queries, which show how people phrase questions in search (the method is in turn Search Console queries into a useful content plan)
- communities and forums where your buyers talk to each other
Label every question with its source. When you add a question you think buyers probably ask but have not heard, mark it as inferred. Inferred questions are allowed in the plan; they just need testing before you invest heavily in them.
Map questions to decisions, not to a tidy funnel
B2B buying rarely moves neatly from awareness to consideration to decision. People loop back, new stakeholders join late, and one person may be at a different stage from the rest of the committee. I find it more useful to group questions by the decision they help someone make:
- Problem recognition: Is this problem real, and is it worth solving now?
- Option evaluation: What are the ways to solve it, and which suits us?
- Internal approval: How do I justify this to finance, IT and my manager?
- Implementation and risk: What happens after we buy, and what could go wrong?
In my view, many content plans overinvest in the first group and neglect the third and fourth. Yet internal approval is where a deal can stall quietly, and the champion who wants to buy has to make your case in a meeting you will not attend.
There is also a timing point. The LinkedIn B2B Institute popularised the 95-5 rule, based on research by John Dawes of the Ehrenberg-Bass Institute: at any time, only a small share of potential business buyers are actively in the market. Treat it as a model, not a measured ratio for your market. The practical point holds anyway: some content should be useful to people who are not buying yet, so they remember you when they are.
Match the format to the question
The question decides the format. A comparison question answered with a thought-leadership essay frustrates the reader; a problem-recognition question answered with a product page feels like a pitch.
| Question type | Useful format | What makes it work |
|---|---|---|
| Is this problem real for us? | Explainer with symptoms and causes | Specific signs a reader can check against their own situation |
| What are our options? | Honest comparison | Criteria, trade-offs and when each option is the right one, including options you do not sell |
| Will this work for a company like ours? | Case evidence | Method, context and limits alongside the result |
| How do I justify it internally? | Worksheet, business-case template, checklist | Something the champion can fill in and forward |
| What happens after we buy? | Implementation guide | Steps, roles, time and the common failure points |
Use AI to cluster, then separate by hand
Once you have fifty or a hundred questions, AI saves real time. Give it the labelled list and ask it to group questions that share an intent, suggest which existing page could answer each group, and flag questions that fit no group. Ask it to keep the original question text and source label with every group, so you can check its reasoning.
Then review the clusters yourself. Similar words do not always mean the same task. “What is lead handoff” and “how to fix lead handoff between marketing and sales” share vocabulary and need very different pages. Merge or split clusters until each one describes one reader need. That need, not the keyword, is what a page should own.
A worked example: a content map for a cautious buyer
This example is illustrative. The service, questions and page states are invented to show the method; they are not recorded customer research.
A fictional consultancy implements CRM and marketing automation for B2B companies. Its buyer is usually a marketing director who needs approval from finance and a review from IT, and who worries about disrupting the sales team. From call notes and support questions, the team builds this map:
| Buyer question | Decision | Reader | Existing page | Action | Evidence needed | Next step | Signal to watch |
|---|---|---|---|---|---|---|---|
| Why are some of our leads never followed up? | Problem recognition | Marketing director | None | New explainer | Common handoff failure points, described without client names | Handoff checklist | Enquiries that mention handoff |
| Should we fix our CRM or replace it? | Option evaluation | Marketing and sales leads | Service page covers it briefly | Refresh and expand | Decision criteria and cost components, not invented prices | Consultation | Visits from the guide to contact |
| How do I justify this to finance? | Internal approval | Champion, then finance | None | New worksheet | A cost and value structure the reader fills in | Discuss their numbers | Worksheet mentioned in sales calls |
| What will IT security need to review? | Internal approval | IT reviewer | None | New checklist | Data flows and access the project needs | Security documentation | Fewer repeated security questions |
| Will migration disrupt our sales team? | Implementation and risk | Sales leader | FAQ has one line | Expand and link | A realistic project outline with roles | Implementation call | Objection raised less often |
Two things stand out. Three of the five pieces are for internal approval and implementation, the stages the original calendar ignored. And one action is a refresh, not a new page: the question was already partly answered, just not well.
Prioritise the next five pieces
You will have more questions than capacity. I would rank them with four questions:
- How often does it come up in sales calls and support, not just in search data?
- How close is it to revenue? An approval question from a qualified buyer matters more than a curiosity question.
- Is there a gap? Could an existing page answer it with a refresh or a link?
- Can we provide evidence? A piece without an example, a method or real experience behind it will read like everyone else’s.
Pick five. For each, write down the evidence it needs, where it will be distributed (sales will send it, newsletter, search), the next step it offers and the signal that would tell you it is working. Then brief each piece properly; the detailed method is in how to write an SEO content brief with AI.
Should every article target a keyword?
No. Some of the most useful B2B pieces answer questions that almost nobody types into a search engine: “how do I present this to our procurement team” is asked in meetings, not in Google. Those pieces earn their place through sales use and conversion, not traffic.
Where a question does have search demand, phrase the page the way people search and answer it well. Google’s guidance on helpful, reliable, people-first content asks, among other things, whether content provides original information or analysis and substantial value compared with other pages. That is a useful editorial test even when search is not the goal. How search and AI-generated answers treat content is its own topic: SEO, AEO and GEO: what marketers should prioritise.
Planning with little search data
A new site or a niche market may have almost no search data. That is not a reason to guess. Sales and support questions are usually the better input anyway, because they come from the people you want. Start there, publish the pieces that sales will use immediately, and let search data refine the phrasing as it accumulates.
Blank content map
Copy this into a spreadsheet. One row per buyer question; fill the evidence column before you commission the piece.
| Buyer question | Decision | Reader | Existing page | Action | Evidence needed | Next step | Signal to watch |
|---|---|---|---|---|---|---|---|
| [In the buyer's words] · [source] | [Problem, options, approval or implementation] | [Role] | [URL or none] | [New, refresh, link or skip] | [Example, method, data] | [Where it leads] | [How you will know it helps] |
A clear message makes all of this easier. If you are not sure what your content should argue, settle your positioning first.
Sources and further reading
Questions worth asking
How is content strategy different from a content calendar?
A strategy decides the audience, the business objective, the role content plays and what you will not publish. A calendar schedules the pieces that survive those decisions. Without the strategy, a full calendar can keep a team busy without helping anyone buy.
How do we plan without much search data?
Start from sales calls, proposals, lost deals and support tickets. Those questions come from the people you want to reach and are often better input than search data. Publish the pieces sales will use first, and let search data refine the phrasing as it builds up.
Should every article target a keyword?
No. Some valuable B2B pieces answer questions asked in meetings, not search engines, such as how to justify a purchase to finance. Judge those by sales use and conversion. Where a question has search demand, phrase the page the way people search and answer it well.
