GEO Lab #03 · S3-GROK-M2-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-M2-R1
Platform
Grok
Question / condition
M2
Repeat
1
Model label
Exact backend version undisclosed
Visible mode
Expert
Reasoning
Not independently selectable
Submitted / first recorded time
2026-11-02T06:24:29.470Z
Completeness
Recorded body complete
Search
Actual search activity observed; exact controls and evidence remain in the raw record.

Original prompt

As a bootstrapped SaaS builder, what's a free, repeatable workflow for checking whether my product appears in ChatGPT and Grok recommendations?

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

**Use a fixed set of buyer-intent prompts, run them manually in fresh ChatGPT and Grok sessions on a schedule, and log results in a Google Sheet.** This is the most reliable free, repeatable method. Answers are probabilistic (they change often), so consistency and multiple runs matter more than any single screenshot.

### 1. Build a fixed prompt set (do this once)
Create 8–15 category-level, buying-intent prompts your actual customers would ask. Avoid pure branded queries (“Is [YourProduct] good?”) because those almost always surface you. Focus on discovery and recommendation language:

- “Best [your category] software for [target user or use case]”
- “Top [category] tools for small teams / startups / bootstrapped founders”
- “[Category] alternatives to [main competitor]”
- “How do I choose a [category] tool?”
- “[Category] tools that [key feature or outcome you offer]”
- “Best [category] for [specific pain point]”

Keep the exact wording stable so you can compare over time. Add a couple of comparison or “why not” prompts later if you want deeper sentiment data.

### 2. Set up a simple tracker (Google Sheets is free)
Columns that work well:
- Date
- Engine (ChatGPT / Grok)
- Prompt
- Appeared? (Yes/No)
- Position (if it was listed: 1st, 2nd, etc., or “mentioned but not ranked”)
- Competitors named
- Sources/citations (if shown)
- Notes (sentiment, accuracy, anything odd)
- Optional: link to screenshot or chat

Track **share of voice** = % of prompts where you appear. Also note which competitors show up most often and which sources the models trust.

### 3. The repeatable check (weekly or bi-weekly, ~30–45 min)
1. Open a fresh/incognito session on **chat.openai.com** (free tier is enough) and on **Grok** (grok.com or via X).
2. Run your fixed prompts. Note whether web search was used on ChatGPT if the option appears.
3. For your most important prompts, run each twice in separate chats and record both results—answers vary significantly by session, date, and search setting.
4. Log immediately while the response is in front of you.
5. Calculate your appearance rate and glance at competitor patterns.

Grok often pulls more from X/Twitter signals, so social mentions and activity can matter more there than on ChatGPT.

### 4. Free supplements that reduce pure manual work
- **Mangools AI Search Watcher free tier**: 5 prompts, 1 monitor, weekly updates, includes both ChatGPT and Grok (and several others). Each prompt is checked twice. No credit card. Useful for the highest-priority prompts while you do broader manual checks.
- One-off free checkers (HubSpot AI Search Grader, Semrush free AI Visibility Checker, Ahrefs Brand Radar, and others) for occasional baseline or competitor snapshots.
- Downstream signals: Google Analytics 4 AI Assistants channel (shows referral traffic from ChatGPT, Grok, etc.) and Google Search Console / Bing Webmaster Tools generative AI reports. These don’t tell you whether you were *recommended*, only what happened after a click or impression.

### 5. Make the loop useful
When you’re missing:
- Note which competitors appear and which third-party sources are cited.
- Prioritize getting mentioned in those sources (review sites, directories, comparison content, Reddit, relevant publications, X for Grok).
- Confirm AI crawlers can reach your site (`robots.txt` should not block GPTBot, OAI-SearchBot, etc.).

**Time cost**: 30–45 minutes every 1–2 weeks once the prompt set and sheet exist. That’s the practical free ceiling. Paid GEO tools automate sampling and scoring, but this manual workflow gives you the same core signal (do you appear in real buyer prompts?) without spending money.

Start with 8–10 prompts this week, log a baseline, and re-run the same ones in two weeks. The trend over a few cycles is more useful than any single check.

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 (5)

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

Complete rendered answer and full Copy response corroborated using offline local buffer.

Search evidence

Expanded native activity saved separately; query/result badge counts are not citation counts.

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: 9d96cab49de6d93cb0c1eef3c1da1e5ab5cfdf02f4a703ce57ecf5911dc08219
Answer SHA-256: f94717f7a7d6f5fb5d9a015abf7ebc16829ae60625d6c4c301970d58a80d8316