Playbook April 29, 2026 4 min read
Buyer queries you should run every Monday, a 15-prompt sanity check
Most teams run AEO audits quarterly and miss the weekly drift. Here's a 15-prompt Monday-morning routine that surfaces problems in 20 minutes a week.
Why Monday morning, why 15 prompts
LLM citation behavior drifts week to week, not just at major model releases. By the time you notice a problem in your quarterly audit, you’ve been losing citations for 8-10 weeks.
A weekly 20-minute scan catches the drift early and turns AEO from a project into a maintenance practice. Monday morning works because:
- The week’s pipeline meeting hasn’t happened yet, you have time to act on what you find.
- Most LLM model updates roll out late in the prior week, Monday catches the first observable behavior change.
- It anchors AEO ownership to a specific time, which is what gets things actually maintained.
The 15 prompts
Five categories, three prompts each, all written from the perspective of your actual buyer.
Category fit (3 prompts):
- “What’s the best [your category] for [your ICP company-stage]?”
- “I’m comparing [your product] vs [top competitor]. Which is better for [primary use case]?”
- “What does [your category] cost in 2026?”
Buyer skepticism (3 prompts):
- “Is [your product] worth it for a [your ICP role]?”
- “What are the downsides of [your category]?”
- “When should I NOT use [your category]?”
Workflow questions (3 prompts):
- “How do I [primary workflow your product solves] in 2026?”
- “What’s the standard process for [adjacent workflow]?”
- “Tools and templates for [primary workflow]?”
Authority queries (3 prompts):
- “Who’s the best resource on [your category’s underlying discipline]?”
- “Where can I learn [your category’s hardest topic]?”
- “What’s the latest research on [your category]?”
Alternatives queries (3 prompts):
- “[Top competitor] alternatives 2026”
- “Cheaper alternatives to [adjacent category leader]”
- “Open-source [your category] options”
Run them in ChatGPT, Claude, and Perplexity. That’s 45 queries. Twenty minutes if you don’t rabbit-hole.
What to look for
For each query, three things matter:
- Are you cited? Yes/no.
- What sources are cited around you? This is the cite-graph for that query. Note any source you don’t appear on.
- What’s the framing? If you’re cited but as “the more expensive option” or “limited to enterprise,” that’s a positioning problem the platform learned from somewhere specific. Trace it.
What to do with the findings
Three buckets:
- Missing citation, source is one you control. Audit the source. Fresh date, third-party corroboration, clean schema. Republish within the week.
- Missing citation, source is one you don’t control. Either earn presence (review, Reddit, analyst outreach) or accept that query as out of reach and reprioritize.
- Cited but framed badly. Hunt for the original wording the platform learned from. Often it’s a single G2 review or Reddit comment from years ago. Counter with a fresher, more accurate source, the platform will reweight within 2-3 weeks.
The discipline that compounds
The teams who do this every Monday outperform teams who do quarterly audits by a factor of 3-4 on citation share within a year. Not because they’re smarter, because they’re noticing problems while they’re small.
Twenty minutes a week. Pick a Monday.