GEO Lab #03 · S3-GROK-M1-R1 · Original response

This page presents one selected response from GEO Lab #03. It adds a smaller reading and citation entry point; it does not replace the full archive or collect a new answer.

The OpenAI arm uses the Responses API with web search and does not represent consumer ChatGPT Search. Grok records identify their actual consumer-interface mode and available model label.

Recorded conditions

Record ID
S3-GROK-M1-R1
Platform
Grok
Question / condition
M1
Repeat
1
Model label
Exact backend version undisclosed
Visible mode
Expert
Reasoning
Not independently selectable
Submitted / first recorded time
2026-11-02T06:29:21.318Z
Completeness
Recorded body complete
Search
Actual search activity observed; exact controls and evidence remain in the raw record.

Original prompt

I'm a solo SaaS founder with no GEO budget. How can I reliably measure whether ChatGPT and Grok recommend my product to potential buyers?

The agreed prompt is reproduced exactly. Any recorded difference in the actual submitted text or search UI is retained in the raw record and full appendix.

Original answer

The most reliable no-budget method is a fixed set of buyer-intent prompts that you run regularly on ChatGPT and Grok, logging results in a simple spreadsheet. This is the standard approach recommended across GEO (Generative Engine Optimization) measurement guides for solo founders and small teams. Paid trackers exist, but manual testing with consistency gives you actionable data without cost.⁠Shmeer⁠Gomega

1. Build a fixed prompt set (your “keyword list”)

Create 10–20 unbranded questions that a real potential buyer would ask before choosing a tool in your category. Keep the wording identical every time so changes reflect real shifts, not prompt variation.

Include a mix of:

Category / “best of” queries (“best [category] tools for [use case]”)
Problem-solving queries (“how to [solve specific problem]”)
Comparison or alternative queries (“alternatives to [competitor]”)
Use-case specific queries

Examples of structure (adapt to your SaaS):

“What are the best [category] tools for [specific use case]?”
“Which [category] software helps with [problem] for small teams?”
“Top tools for [job to be done]”
2. Run them consistently and log results

Query the exact same prompts on:

ChatGPT (chatgpt.com, preferably with web search/browsing enabled for more current results)
Grok (grok.com or via X)

Frequency: Weekly or bi-weekly is practical for a solo founder. Monthly still works for trend tracking.

Logging (spreadsheet columns):

Date
Engine (ChatGPT / Grok)
Prompt
Mentioned? (Yes/No)
Position (e.g., 1st, 3rd, or not listed)
Recommended? (Yes if positively suggested as a good option)
Sentiment (positive / neutral / negative)
Your site cited as a source? (Yes/No + URL if present)
Competitors mentioned
Notes (any quirks)

Key metrics to calculate:

Mention rate = prompts where your product appears ÷ total prompts
Recommendation rate = prompts where it is positively recommended ÷ total prompts
Share of voice vs. top competitors

Run each high-priority prompt 2–3 times per session if you want to smooth out AI non-determinism (answers vary). Trends over weeks/months matter more than any single run.⁠Geoaura

Tips for reliability:

Use a consistent browser setup (incognito or same account).
Note the approximate date and whether web search was on.
Focus first on commercial/high-intent prompts rather than pure informational ones.
3. Complement with free lagging indicators

Manual testing shows whether recommendations happen. These show whether they drive any results:

Google Analytics 4: Look for sessions with source chatgpt.com (OpenAI adds utm_source=chatgpt.com to many referral links) or the AI Assistant channel group. Track landing pages and conversions from those sessions. Note that not every recommendation produces a click, and tracking is imperfect (especially from mobile apps).⁠Goodatmarketing
Google Search Console: Check generative AI / AI Overviews performance reports if available for your property (impressions data).
Bing Webmaster Tools: AI Performance reports for Copilot-related signals.
4. Free readiness checks (do these once or quarterly)

These don’t measure ongoing recommendations, but they show whether your site is set up for AI systems to discover and cite it:

HubSpot AEO / AI Search Grader (brand perception snapshot across models)
Mangools AI Search Grader
geo-tool.com free GEO check (crawler access, citability, structure)
Ensure AI crawlers are not blocked in robots.txt, key pages have clear structure, and you have relevant structured data
Realistic expectations for a solo founder

A full monthly cycle of 15 prompts × 2 engines can take 30–90 minutes. Start with 8–10 high-intent prompts. The value comes from the gap list: prompts where competitors appear and you don’t. Use those to guide content (answer the exact question clearly, add evidence/stats/quotes, get mentioned on third-party sites that AI systems already cite).

This method is the foundation used by almost every free/low-cost GEO guide. Paid tools mainly automate the same process at higher volume. Consistency with fixed prompts is what makes the measurement reliable enough to act on.

Recorded sources

These are the source roles recorded in the public dataset. A retrieval or source-panel item is not automatically an answer-body citation. A URL supplied in the prompt and repeated by an answer does not prove content access. No destination has been repaired or independently validated here.

Recorded answer-body links (4)

Evidence and qualifications

This is the existing redacted public record. It has not been rewritten for this page. Earlier privacy edits and collection qualifications continue to apply. Editorial text is English; a few original captures retain non-English UI badges or source-panel date labels. An answer’s empirical claims are platform output, not verified findings of this archive.

Capture method

Rendered Grok article innerText through browser connection; UI heading, thought-duration label and source badge removed. Complete opening and concluding paragraph reviewed.

Record notes

[
"Browser DOM access became available during collection. Earlier native UI capture remains preserved.",
"Clipboard read through browser returned empty; rendered article supplied complete text. No question was resubmitted."
]

Existing public-text privacy policy

{
"body_changed": false,
"policy": "Only actual URL tracking, signed/session URL parameters and necessary identifying information are removed; original wording, including analytics parameter examples outside URLs, remains.",
"annotation_positions": "Original position indexes omitted because URL/privacy edits can change character locations."
}

Prompt SHA-256: 709fd6fc82fe3f87b623fb3dd8b2adecdd6655c3508115f8b40c573d2397776d
Answer SHA-256: d2f48dae81455820600309fb30d0e88a485b52d5b3c3b7b94a32fb531963f1b3