---
title: "The Prompt Sheet: 60 Buyer Questions That Show Where AI Search Leaves You Out"
description: "Copy 60 buyer-prompt templates across category, modifier, competitor, problem and trust intents. Then log what ChatGPT, Gemini and Google AI Overviews actually say."
url: "https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p"
image: "https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/6c01d16c-ef3f-464d-b64c-7343364ea313.png"
published: "2026-10-09T05:29:25.727Z"
modified: "2026-10-09T05:29:25.789Z"
---

October 9, 2026 · 10 min read

# The Prompt Sheet: 60 Buyer Questions That Show Where AI Search Leaves You Out

[Alexander PerelmanHead Of Product @ groas](https://groas.com/author/alexander-perelman)[LinkedIn](https://www.linkedin.com/in/alexander-433793253/)

![A tiny business owner with a clipboard plants flags on a huge ledger sheet laid out like farmland, most rows still empty: the gaps they haven't checked.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/6c01d16c-ef3f-464d-b64c-7343364ea313.png)

In this article

1. [Swipe 1: Find out whether you make the category shortlist](#swipe-1-find-out-whether-you-make-the-category-shortlist)
2. [Swipe 2: Test the constraints that decide the purchase](#swipe-2-test-the-constraints-that-decide-the-purchase)
3. [Swipe 3: See who wins when a buyer wants out](#swipe-3-see-who-wins-when-a-buyer-wants-out)
4. [Swipe 4: Catch buyers who never say your category name](#swipe-4-catch-buyers-who-never-say-your-category-name)
5. [Swipe 5: Find the holes in your buyer’s due diligence](#swipe-5-find-the-holes-in-your-buyers-due-diligence)
6. [Artefact 6: Log the answer before you forget what you asked](#artefact-6-log-the-answer-before-you-forget-what-you-asked)
7. [Artefact 7: Fix the missing prompts in purchase order](#artefact-7-fix-the-missing-prompts-in-purchase-order)
8. [Artefact 8: Know when the sheet has done its job](#artefact-8-know-when-the-sheet-has-done-its-job)
9. [The artefact I would copy first: Swipe 2](#the-artefact-i-would-copy-first-swipe-2)

Most businesses test the 10 AI prompts they want to win, not the 60 their buyers might ask. You do not need a tool on day one to find the gaps: build a prompt sheet the way I used to build a keyword list, with head terms, modifiers, competitor terms, problems and trust checks.

## Swipe 1: Find out whether you make the category shortlist

**When to use it:** Start here when you need an unassisted baseline: does an engine name your brand when the buyer knows the category but has not chosen a vendor? [Radyant’s prompt tracking research](https://www.radyant.io/guides/prompt-tracking-ai-search-visibility) found a 13.3% mention rate for non-branded commercial prompts, compared with 99.7% for brand queries. Asking about yourself is not much of a test.

```text
1. What are the best [category] platforms for [buyer_type] in [industry]?
2. Top-rated [category] providers that serve [industry] companies in [location].
3. Give me a shortlist of the leading [category] tools for a team of [team_size].
4. Which [category] services are recommended for [buyer_type] with an annual budget of [budget_range]?
5. What are the most reliable [category] solutions currently on the market for [use_case]?
6. Recommend three independent [category] firms specializing in [target_market].
7. What is the standard [category] software stack used by [buyer_type] teams today?
8. Who are the recognized market leaders in [category] for mid-market [industry] companies?
9. Which commercial [category] vendors offer dedicated account management rather than self-serve tickets?
10. Best local [category] contractors near [location] with verified commercial credentials.
11. What [category] platforms do [industry] executives recommend for improving [primary_metric]?
12. A breakdown of vetted [category] solutions for teams transitioning away from [legacy_system].
```

**The adjustment that matters:** Keep the filled-in prompt identical across repeated runs. A [SparkToro and Gumshoe study of 2,961 prompts](https://www.searchenginejournal.com/ai-recommendations-change-with-nearly-every-query-sparktoro/566242/) found that recommended brand lists change frequently between runs. Test in ChatGPT and Gemini, and use the same query in Google Search when it produces an AI Overview. **One omission is an observation, not a pattern.**

## Swipe 2: Test the constraints that decide the purchase

**When to use it:** Use modifier prompts when buyers have a budget, team size, contract requirement or technical constraint. This is where a broad recommendation starts turning into a workable shortlist. In Google Ads, I used to separate those modifiers from the head term; lumping them together here hides the same distinction.

![Risograph illustration of a sorting sieve separating detailed prompts about price and operating constraints from generic keywords.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/a862ae80-c8a0-43ee-91d6-e45ecfcb7141.png)

```text
13. Which [category] tools cost less than [dollar_amount] per month for a [team_size] person team?
14. What are the best [category] options for [buyer_type] that do not require an annual contract?
15. Recommend [category] platforms that integrate natively with [software_tool] without third-party middleware.
16. Which [category] providers support [compliance_standard] out of the box?
17. Best [category] solutions for companies with under [revenue_or_headcount] employees.
18. What [category] software offers self-serve onboarding without a mandatory sales call?
19. Which [category] services offer transparent flat monthly pricing instead of percentage-of-spend fees?
20. Top [category] platforms for [industry] that can be deployed in under [timeframe].
21. What are the best [category] tools for Mac environments with zero browser extensions required?
22. Which [category] agencies or software guarantee [specific_deliverable] within [timeframe]?
23. Find [category] providers that handle high-volume [workload_metric] without charging overage penalties.
24. Which [category] tools allow full data export to CSV or warehouse without paid API add-ons?
```

**The adjustment that matters:** Replace adjectives with **boundaries the buyer can verify**. “Affordable” leaves an engine to guess; “under \[dollar_amount] per month with \[compliance_standard] and \[software_tool] integration” gives it criteria. If your brand disappears, check whether your public pages actually state the relevant pricing, integration and contract details before deciding the engine has a grudge against you.

## Swipe 3: See who wins when a buyer wants out

**When to use it:** Run these when buyers know an incumbent but are looking for an alternative. They are not asking for another category definition. They want the structural difference: pricing model, execution, support or migration burden.

```text
25. What are the best alternatives to [competitor_brand] for [use_case]?
26. Why do [buyer_type] teams switch away from [competitor_brand] in [current_year]?
27. How does [my_brand] compare against [competitor_brand] on [critical_capability] and pricing structure?
28. Which providers offer similar capabilities to [competitor_brand] without [common_frustration]?
29. A candid comparison between [competitor_1], [competitor_2], and [my_brand] for mid-market [industry].
30. What [category] platform is easiest to migrate to from [competitor_brand] with minimal downtime?
31. Is [competitor_brand] still worth the cost for a [team_size] company, or are there leaner modern alternatives?
32. What are the primary operational trade-offs of choosing [competitor_brand] versus newer solutions?
33. Which [category] platforms offer hands-on execution and senior support rather than shared support queues?
34. Who are the top competitors to [competitor_brand] offering guaranteed turnaround times?
35. What newer tools are displacing [competitor_brand] in the [niche_segment] market today?
36. Are there any [category] providers that match [competitor_feature] at a predictable flat monthly rate?
```

**The adjustment that matters:** Inspect the links beneath the answer, not just the names in it. An [Ahrefs analysis of 4 million AI Overview URLs](https://ahrefs.com/blog/ai-overview-citations-top-10/) found that many cited pages did not rank in the top 10 organic results. Your organic position alone does not settle this contest. If your comparison pages never spell out your actual terms and capabilities, the engine has to assemble the difference from whatever else it can find. **Give it a concrete comparison to work with.**

## Swipe 4: Catch buyers who never say your category name

**When to use it:** Use problem-first prompts when a buyer can describe the broken workflow but would never type your category noun. I used to find these symptom queries in search-term reports. They were easy to miss if the account structure started and ended with product names.

```text
37. How can a [buyer_type] fix [specific_operational_bottleneck] without adding headcount?
38. What steps or tools reduce [painful_failure] by [percentage]%?
39. How do companies handle [complex_compliance_or_technical_task] when their in-house team lacks bandwidth?
40. Why is our [core_metric] climbing despite increasing spend, and how do we resolve it?
41. What is the most reliable way to automate [repetitive_manual_task] across [platform_1] and [platform_2]?
42. How to diagnose why [critical_business_asset] tanks after a major update?
43. What solutions exist for [buyer_type] struggling with [specific_data_silo_or_reporting_nightmare]?
44. How can mid-market teams scale [output_volume] without doubling their operational costs?
45. Best approaches to eliminate [vendor_friction] in [department].
46. What workflow architecture prevents [common_disaster] during a platform migration?
47. How to manage [high_velocity_channel] around the clock when the internal team only works business hours?
48. What are companies doing to replace manual [process_name] with automated continuous execution?
```

**The adjustment that matters:** Read the answer for its diagnosis, not merely its vendor list. Does it describe the problem your product solves and still leave you out? A page that explains the failure, the checks and the remedy gives an engine more usable material than one that says “streamlined solutions” six times. **Write the fix in the buyer’s language**, even when that language contains none of your category keywords.

## Swipe 5: Find the holes in your buyer’s due diligence

**When to use it:** Run trust prompts once your brand could plausibly make a shortlist. Buyers checking fees, cancellation terms, proof and accountability are looking for reasons to proceed or stop. Your own reassuring adjective will not answer a question about contract terms.

```text
49. What are the common complaints and customer criticisms regarding [my_brand]?
50. Is [my_brand] legitimate and reliable for enterprise-level [industry] deployment?
51. How transparent is the pricing model of [my_brand], and are there hidden fees or usage caps?
52. What verified results or case studies exist proving [my_brand] delivers [specific_outcome]?
53. What happens if a customer wants to cancel their contract with [my_brand]?
54. Does [my_brand] offer a free trial, pilot program, or performance guarantee before signing?
55. How responsive is the customer support and strategic account management at [my_brand]?
56. What security audits, certifications, or data privacy protocols does [my_brand] maintain?
57. What do former customers on review platforms say about working with [my_brand]?
58. Who runs the leadership team at [my_brand], and what is their background in [industry]?
59. How does [my_brand] handle accountability when performance targets or campaign goals are missed?
60. What are the definitive operational trade-offs of hiring [my_brand] versus an in-house hire or traditional agency?
```

![Diagram showing a brand cited without a recommendation and recommended without a source link.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/b2f2566b-2bec-4326-b162-1cb0076ad641.png)

**The adjustment that matters:** Log _cited but not mentioned_ separately from _mentioned but not cited_, two failure modes described in [AuthorityRadar’s brand monitoring research](https://authorityradar.com/blog/monitor-brand-mentions-perplexity). In the first, your page helps support an answer but your brand does not make the recommendation. In the second, the brand appears without a link to your domain. Neither tells you, by itself, what to rewrite. Look at the answer and its sources before choosing a fix; that is the point of [measuring visibility with fixed buyer prompts and repeated runs](https://groas.com/post/reader-q-a-how-do-i-actually-measure-whe).

## Artefact 6: Log the answer before you forget what you asked

**When to use it:** Set up the sheet before the first run. I have seen enough shared documents full of pasted chat transcripts to know that “we’ll organize it later” means “we won’t.” Use **one row per prompt, engine and run** so an intermittent mention does not become a confident yes or no.

| Column                   | Enter                                                   | Why it matters                                      |
| :----------------------- | :------------------------------------------------------ | :-------------------------------------------------- |
| **Engine**               | ChatGPT, Gemini or AI Overviews                         | Keeps engine results separate.                      |
| **Prompt ID and bucket** | `HEAD-01`, `MOD-13`, `COMP-27`, `PROB-40` or `TRUST-49` | Connects the result to the swipe file.              |
| **Prompt string**        | Exact filled-in text                                    | Stops rephrasing from contaminating the comparison. |
| **Mentioned**            | Yes or No                                               | Records whether the brand appeared.                 |
| **Position**             | 1, 2, 3 or N/A                                          | Records its order when applicable.                  |
| **Cited URL**            | Link or None                                            | Shows whether your domain received attribution.     |
| **Competitors named**    | Comma-separated names                                   | Shows who appeared instead.                         |
| **Failure mode**         | Cited Only, Mentioned Only, Both or Omitted             | Separates two different visibility signals.         |

**The adjustment that matters:** Run the same filled-in string more than once and keep every row. The [SEO Hacker AI search tracking protocol](https://seo-hacker.com/track-prompts-ai-search/) recommends three to four consecutive runs before treating a missing mention as a structural gap. If you appear twice in four runs, **write “unstable,” not “invisible.”** A clean browser profile also helps keep your browsing history out of the test.

## Artefact 7: Fix the missing prompts in purchase order

**When to use it:** Use this list when the sheet starts filling with “No.” Do not rewrite the entire site because a broad category query omitted you once. Start where an accurate answer could change a buyer’s shortlist.

1. **Modifier gaps, prompts 13–24.** Check the price, contract, integration or company-size constraint that excluded you. If you meet it, state the relevant fact plainly on the appropriate page.
2. **Problem-first gaps, prompts 37–48.** Compare the engine’s diagnosis with your own content. If your pages skip the operational steps and jump straight to a sales pitch, document the actual workflow.
3. **Competitor gaps, prompts 25–36.** Show the structural differences a buyer can evaluate: capabilities, pricing model and terms. An attack page is less useful than a side-by-side comparison that can survive inspection.
4. **Category head gaps, prompts 1–12.** Keep tracking them, but do not let the broadest queries consume the time needed to answer specific buyer constraints.

![High-contrast graphic showing four prompt buckets ranked by purchase urgency.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/8fdd0c2f-787f-4f03-9326-2d5a3e6e2d7b.png)

**The adjustment that matters:** Before changing a page, fill in this sentence: “We meet \[buyer_constraint], but the answer omitted us because \[missing_or_unclear_public_fact].” If you cannot fill in the second bracket from the answer and your pages, investigate further. **A missing mention is a symptom, not a content brief.**

## Artefact 8: Know when the sheet has done its job

**When to use it:** Keep this checkpoint beside your logging sheet. Manual testing is a good first diagnostic; it is a poor permanent job description. Sixty prompts across three engines with three runs each means 540 queries before you have written a single fix.

```text
Keep the manual sheet when:
- You are establishing a baseline for a manageable prompt set.
- You can repeat the runs and review the cited pages yourself.
- Each finding leads to a specific page or claim to check.

Move beyond manual logging when:
- Skipped runs and stale cells make the trend unreliable.
- You need more frequent checks across many modifier prompts.
- Recording omissions takes time away from fixing the pages behind them.
```

**The adjustment that matters:** Do not buy monitoring merely to produce a prettier list of omissions. [Practitioner discussions about AI visibility tracking](https://www.reddit.com/r/SEO_LLM/comments/1wzve6q/how_do_experienced_people_build_an_ai_visibility/) describe how quickly repeated spreadsheet work becomes a burden. The reason to move from a sheet to [groas’s autonomous search execution](https://groas.com/for-businesses) is the operating model: continuous execution with a human strategist responsible for direction and accountability, rather than another dashboard you have to interpret yourself.

## The artefact I would copy first: Swipe 2

**When to use it:** If you have time for only one bucket tonight, copy **modifier prompts 13–24** and fill in the constraints your buyers actually bring to a purchase decision.

```text
First run: MOD-13 through MOD-24
Keep fixed: [category], [buyer_type], [team_size] and other shared details
Change deliberately: the price, contract, integration or operating constraint
Log: each engine, each run, each mention and each cited URL
```

**The adjustment that matters:** Use constraints you can substantiate on your own pages. Category prompts tell you whether an engine recognizes your name; modifier prompts show whether it can recommend you when a buyer states what they need. That is the gap I would want to find first.

## Frequently Asked Questions

### How many AI prompts should I test to see where my brand shows up in AI search?

Most businesses test the 10 prompts they want to win, but buyers might ask around 60. Build a prompt sheet the way a keyword list is built, covering category head terms, modifiers, competitor terms, problems and trust checks, before spending on a tool.

### Do AI answers stay the same when you ask the same question twice?

No. A SparkToro and Gumshoe study of 2,961 prompts found that recommended brand lists change frequently between runs. Keep the filled-in prompt identical across repeated runs, test in ChatGPT and Gemini, use the same query in Google Search when it produces an AI Overview, and remember one omission is an observation, not a pattern.

### Why do AI engines leave my brand out when a buyer states a budget or integration requirement?

Vague adjectives like affordable leave the engine guessing, while verifiable boundaries such as an exact monthly price, a compliance standard and a named integration give it criteria. If the brand disappears from a modifier answer, check whether the public pages actually state the relevant pricing, integration and contract details.

### Does my Google ranking determine whether AI Overviews cite my pages?

Not necessarily. An Ahrefs analysis of 4 million AI Overview URLs found many cited pages did not rank in the top 10 organic results. If comparison pages never spell out actual terms and capabilities, the engine assembles the difference from other sources, so give it a concrete side-by-side comparison to work with.

### What are problem-first prompts and why do they matter for AI search visibility?

Problem-first prompts target buyers who can describe a broken workflow but would never type the category noun, like symptom queries from search-term reports. Read the answer's diagnosis, not just its vendor list, and document the actual failure, checks and remedy in the buyer's language, even without category keywords.

### What is the difference between being cited but not mentioned and mentioned but not cited in AI answers?

In the first case a page helps support the answer but the brand does not make the recommendation; in the second the brand appears without a link to the domain. Neither tells you by itself what to rewrite, so look at the answer and its sources before choosing a fix.

### How many times should I run the same prompt before concluding my brand has a visibility gap?

The SEO Hacker AI search tracking protocol recommends three to four consecutive runs before treating a missing mention as a structural gap. Log one row per prompt, engine and run, keep every row, and if the brand appears twice in four runs, record that as unstable rather than invisible. A clean browser profile also helps.

### Which prompt gaps should I fix first when AI answers omit my brand?

Start with modifier gaps, then problem-first gaps, then competitor gaps, keeping category head prompts tracked but not time-consuming. Check the constraint that excluded you, state missing facts plainly on the right page, and fill in the sentence about which missing public fact caused the omission. A missing mention is a symptom, not a content brief.

## Pay For Results, Not For Hours

Businesses buy the outcome, agencies resell it, and groas answers for it either way.

[See If You Qualify](https://groas.typeform.com/to/xC1bQNUT)

## Structured data

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://graph.groas.com/entity/groas","name":"groas","url":"https://groas.com/","logo":{"@type":"ImageObject","url":"https://groas.com/icon-512.png","width":512,"height":512},"sameAs":["https://www.linkedin.com/company/groas/"]},{"@type":"WebSite","@id":"https://groas.com/#website","name":"groas","url":"https://groas.com/","inLanguage":"en-US","publisher":{"@id":"https://graph.groas.com/entity/groas"}},{"@type":"WebPage","@id":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p#webpage","url":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p","name":"The Prompt Sheet: 60 Buyer Questions That Show Where AI Search Leaves You Out","description":"Copy 60 buyer-prompt templates across category, modifier, competitor, problem and trust intents. Then log what ChatGPT, Gemini and Google AI Overviews actually say.","inLanguage":"en-US","isPartOf":{"@id":"https://groas.com/#website"},"about":{"@id":"https://graph.groas.com/entity/groas"},"primaryImageOfPage":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/6c01d16c-ef3f-464d-b64c-7343364ea313.png"},"breadcrumb":{"@id":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p#breadcrumb"},"dateModified":"2026-10-09T05:29:25.789Z","mainEntity":{"@id":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p#article"}},{"@type":"BreadcrumbList","@id":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://groas.com/"},{"@type":"ListItem","position":2,"name":"Blog","item":"https://groas.com/blog"},{"@type":"ListItem","position":3,"name":"The Prompt Sheet: 60 Buyer Questions That Show Where AI Search Leaves You Out","item":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p"}]},{"@type":"BlogPosting","@id":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p#article","url":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p","mainEntityOfPage":{"@id":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p#webpage"},"isPartOf":{"@id":"https://groas.com/blog#blog"},"headline":"The Prompt Sheet: 60 Buyer Questions That Show Where AI Search Leaves You Out","description":"Copy 60 buyer-prompt templates across category, modifier, competitor, problem and trust intents. Then log what ChatGPT, Gemini and Google AI Overviews actually say.","datePublished":"2026-10-09T05:29:25.727Z","dateModified":"2026-10-09T05:29:25.789Z","image":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/6c01d16c-ef3f-464d-b64c-7343364ea313.png"},"author":{"@id":"https://groas.com/author/alexander-perelman#person"},"publisher":{"@id":"https://graph.groas.com/entity/groas"},"keywords":"category, brand, buyer, competitor, prompt, use, competitor brand, prompt sheet","wordCount":2201,"inLanguage":"en-US"},{"@type":"Person","@id":"https://groas.com/author/alexander-perelman#person","name":"Alexander Perelman","jobTitle":"Head Of Product @ groas","description":"Ex Goldman Sachs and Ex Stanford Computer Science","image":"https://groas.com/media/blog/4f17cac3a81acc224cc1b2cfab8f94bbc295a3089d13e8421125771ef8d4064f.jpg","worksFor":{"@id":"https://graph.groas.com/entity/groas"},"sameAs":["https://www.linkedin.com/in/alexander-433793253/"]},{"@type":"FAQPage","@id":"https://groas.com/post/the-prompt-sheet-60-buyer-questions-to-p#faq","mainEntity":[{"@type":"Question","name":"How many AI prompts should I test to see where my brand shows up in AI search?","acceptedAnswer":{"@type":"Answer","text":"Most businesses test the 10 prompts they want to win, but buyers might ask around 60. Build a prompt sheet the way a keyword list is built, covering category head terms, modifiers, competitor terms, problems and trust checks, before spending on a tool."}},{"@type":"Question","name":"Do AI answers stay the same when you ask the same question twice?","acceptedAnswer":{"@type":"Answer","text":"No. A SparkToro and Gumshoe study of 2,961 prompts found that recommended brand lists change frequently between runs. Keep the filled-in prompt identical across repeated runs, test in ChatGPT and Gemini, use the same query in Google Search when it produces an AI Overview, and remember one omission is an observation, not a pattern."}},{"@type":"Question","name":"Why do AI engines leave my brand out when a buyer states a budget or integration requirement?","acceptedAnswer":{"@type":"Answer","text":"Vague adjectives like affordable leave the engine guessing, while verifiable boundaries such as an exact monthly price, a compliance standard and a named integration give it criteria. If the brand disappears from a modifier answer, check whether the public pages actually state the relevant pricing, integration and contract details."}},{"@type":"Question","name":"Does my Google ranking determine whether AI Overviews cite my pages?","acceptedAnswer":{"@type":"Answer","text":"Not necessarily. An Ahrefs analysis of 4 million AI Overview URLs found many cited pages did not rank in the top 10 organic results. If comparison pages never spell out actual terms and capabilities, the engine assembles the difference from other sources, so give it a concrete side-by-side comparison to work with."}},{"@type":"Question","name":"What are problem-first prompts and why do they matter for AI search visibility?","acceptedAnswer":{"@type":"Answer","text":"Problem-first prompts target buyers who can describe a broken workflow but would never type the category noun, like symptom queries from search-term reports. Read the answer's diagnosis, not just its vendor list, and document the actual failure, checks and remedy in the buyer's language, even without category keywords."}},{"@type":"Question","name":"What is the difference between being cited but not mentioned and mentioned but not cited in AI answers?","acceptedAnswer":{"@type":"Answer","text":"In the first case a page helps support the answer but the brand does not make the recommendation; in the second the brand appears without a link to the domain. Neither tells you by itself what to rewrite, so look at the answer and its sources before choosing a fix."}},{"@type":"Question","name":"How many times should I run the same prompt before concluding my brand has a visibility gap?","acceptedAnswer":{"@type":"Answer","text":"The SEO Hacker AI search tracking protocol recommends three to four consecutive runs before treating a missing mention as a structural gap. Log one row per prompt, engine and run, keep every row, and if the brand appears twice in four runs, record that as unstable rather than invisible. A clean browser profile also helps."}},{"@type":"Question","name":"Which prompt gaps should I fix first when AI answers omit my brand?","acceptedAnswer":{"@type":"Answer","text":"Start with modifier gaps, then problem-first gaps, then competitor gaps, keeping category head prompts tracked but not time-consuming. Check the constraint that excluded you, state missing facts plainly on the right page, and fill in the sentence about which missing public fact caused the omission. A missing mention is a symptom, not a content brief."}}]}]}
```