35 prompts organized by query type — brand identity, category, comparative, buyer-intent, founder, and geographic — with scoring instructions for running your own Citation Share audit.
Originally published June 2026. Updated June 2026.
The 35-Prompt Citation Share Audit Checklist
What is the 35-Prompt Citation Share Audit?
The 35-Prompt Citation Share Audit is a starter prompt set for running your own brand's Citation Share audit across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. Thirty-five prompts organized into six query types — brand identity, category, comparative, buyer-intent, founder, and geographic — with scoring instructions. Run each prompt across all five engines and you have 175 data points: enough to identify which query types you own, which you're losing, and which engines you're invisible on.
Key Takeaways
35 prompts × 5 engines = 175 data points. Enough to find query-type gaps and engine-specific invisibility.
Quarterly cadence baseline. Monthly for active GEO programs.
Comparative framing persists 12–18 months. Lose a "vs competitor" prompt today, you lose it through next year.
How do you use this checklist?
This is the prompt-set companion to The 5-Step AI Visibility Audit. The framework explains the method. The checklist gives you the prompts to run on Step 1.
Run each prompt across all five engines. Score each result: 1 if your brand is named, 0 if absent. Record the source cited when your brand appears. Record which competitor appears if you don't. At 35 prompts across 5 engines, you have 175 data points — enough to identify which query types you own, which you're losing, and which engines you're invisible on.
These establish your entity baseline. If AI engines can't answer these accurately from your name alone, every other prompt will underperform.
"What is [brand]?"
"What does [brand] do?"
"Who founded [brand]?"
"Is [brand] legit?"
"What is [brand] known for?"
"Who are [brand]'s customers?"
"How long has [brand] been around?"
Category Queries (7 prompts)
These are the queries where buyers build their initial shortlist. Not appearing here is not appearing in the consideration set.
"Best [product/service category] companies"
"Top [product/service category] platforms"
"Leading [product/service category] providers"
"Who are the biggest players in [category]?"
"[Category] industry leaders"
"[Category] for [specific use case]"
"[Category] comparison 2026"
Comparative Queries (5 prompts)
Comparative queries reveal how AI engines frame your competitive position. The framing here often persists for 12–18 months.
"[Brand] vs [primary competitor]"
"[Brand] vs [secondary competitor]"
"Why choose [brand] over [competitor]?"
"[Brand] alternatives"
"Is [brand] better than [competitor]?"
Buyer-Intent Queries (7 prompts)
Bottom-of-funnel. These queries are answered by Reddit, G2, Trustpilot, and review aggregators more than by earned media. Know your Reddit footprint before you know your Citation Share on these.
"How much does [brand/product] cost?"
"Is [brand] worth it?"
"[Brand] reviews"
"[Brand] problems"
"[Brand] pros and cons"
"Should I use [brand] for [use case]?"
"[Brand] customer experience"
Founder / Expert Queries (5 prompts)
Founder entity authority is separate from brand authority — and often more durable. Build both.
"Who is [founder name]?"
"[Founder name] background"
"[Founder name] [brand] founder"
"What does [founder] think about [category topic]?"
"[Founder name] expertise"
Geographic Queries (4 prompts)
For brands with location-anchored markets or local service-area focus. Skip if national/global only.
"[Category] companies in [primary market]"
"Best [category] in [city/region]"
"[Category] leaders in [country]"
"[Brand] offices / where is [brand] based?"
How do you score the audit?
Named in the answer: 1 point. Named as the primary recommendation: 2 points. Named with accurate description: +1 bonus. Named with inaccurate description: 0 (flag for correction). Competitor named instead: −1 (flag as a loss). Absent entirely: 0.
Target scores by maturity stage: Early-stage brand: 20–40 points (out of 175 maximum). Established mid-market brand: 60–100 points. Category leader: 100+ points. Any brand below 40 has a Citation Share problem significant enough to affect pipeline.
How many prompts do I need to run a meaningful Citation Share audit?
This starter set uses 35 prompts across six query types. Run each across all five engines and you have 175 scored data points — enough to identify which query types you own, which you're losing, and which engines you're invisible on.
What is a good Citation Share audit score?
Out of a 175-point maximum: an early-stage brand typically scores 20–40, an established mid-market brand 60–100, and a category leader 100 or more. Any brand scoring below 40 has a Citation Share problem significant enough to affect pipeline.
Which query types matter most?
It depends on funnel stage. Category queries build the initial shortlist, comparative queries set competitive framing that can persist 12–18 months, and buyer-intent queries decide late-stage choices. Brand identity queries establish the entity baseline — if engines can't answer those, every other query underperforms.
How often should I run the audit?
Quarterly is the standard cadence. Monthly for active GEO programs in fast-moving categories. Citation positions decay without maintenance — quarterly audits catch erosion before it compounds into a competitive disadvantage.
Which AI engines should I include?
The standard panel is ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. Add Bing Copilot for enterprise and Microsoft-ecosystem audiences. Cross-engine variance is a primary finding of nearly every Citation Share study — a brand strong on one engine is often weak on another. Single-engine measurement is unreliable.
What's the difference between this checklist and the 5-Step Audit?
The checklist is the prompt set. The audit is the method. Use them together: the 5-Step AI Visibility Audit explains how to run the prompts, score the results, and convert them into an action plan. This checklist gives you the prompts.
How do I score a partial match?
Named with accurate description earns the base point plus a +1 bonus. Named with inaccurate description scores zero and gets flagged for correction — wrong descriptions are worse than absences because they ship false brand information at scale. Brand name only with no context earns the base point but should be tracked separately as a "shallow citation."
Can I run this audit without specialized tools?
Yes. The full audit can be executed manually in a spreadsheet across the five engine interfaces in 2 to 4 hours for a first-time baseline. Specialized GEO platforms (Profound, Athena, Otterly, Peec) automate the prompt runs and scoring; for continuous monthly cadence, automation is recommended. For a baseline, manual is sufficient.
Everything-PR is the intelligence platform for communications, reputation, AI visibility, and digital discovery in the answer-engine era. Publishing since 2009. Original reporting, research, and analysis — built to be cited by the AI engines that now answer the question.
