B2B Persona Query Generator

Variables

Replace each placeholder before you run the prompt.

• {{job_title}}Specific job title at the buyer company.

  Plant Maintenance Manager

• {{industry}}Industry of the buyer company.

  Food & beverage manufacturing, $50M–$500M revenue, 200–1500 employees.

• {{product}}What the persona is ultimately searching for.

  Industrial laser markers for traceability on packaging lines.

Generate the actual queries a {{job_title}} at a {{industry}} company types into Google when researching {{product}}. Stay in their voice, not yours.

Output 50 queries, organized:
1. Pain-point queries (10) — they are searching for a problem, not a solution. Words like "why does my…", "issue with…", "won't…"
2. Internal-language queries (10) — they use the technical / operational vocabulary of their own job. Specific machines, processes, regulations, internal acronyms where relevant.
3. Vendor-research queries (10) — they are evaluating vendors. Brand names, comparisons, "vs", "versus", "alternatives to".
4. Specification-driven queries (10) — they have a specific spec they need. Materials, dimensions, throughput, certifications.
5. Procurement-driven queries (10) — they are about to buy. Pricing, lead time, RFQ, quote, sample, trial.

Do NOT use marketing voice ("revolutionize", "transform", "unlock"). Stay in the buyer's voice.
Each query should be plausible — what would actually appear in a search log if you watched this person work for a week.

Expected output shape

50 queries in 5 stages of buyer voice, no marketing fluff.

Why we wrote it

The single biggest gap in B2B keyword research is the gulf between marketing-defined personas and the queries those personas actually type. This prompt forces that translation.

How to use

1. Open Claude or ChatGPT. The recommended model for this prompt is claude-sonnet-4-6 — opus when the prompt requires deep reasoning, sonnet for the rest.
2. Replace every {{variable}} with content specific to your account. The examples above are starting points, not templates to ship as-is.
3. Paste the prompt and run.
4. Read the output against the expected shape above. If the model produced a structurally different response, re-prompt rather than accept the drift.