AI answer monitoring · bolddocserver

Headline

Nothing was measured this run. No MCP servers were connected, so the openrouter ask_engine tool wasn't available for ChatGPT, Gemini, Perplexity or Claude. Google AI Overviews and Microsoft Copilot have no API and need a manual check, which an unattended run can't do. All 42 prompt × engine rows are not checked. The prompt set for BoldDocServer is fixed below so the rerun can use it unchanged.

Score

n/a. This checklist has no score (overall_score: null).

Top 3 actions

1. Rerun this week with openrouter connected (see _reference/mcp-and-tools.md) to fill the ChatGPT, Gemini, Perplexity and Claude rows for P1–P7. Use the prompt set in raw/prompt-set.json word for word.
2. Check Google AI Overviews and Microsoft Copilot by hand for P1–P7 in clean, signed-out sessions, and record the country.
3. Do P5 (brand facts) and P7 (trust) first. Office Online Server support ends 2026-12-31, so buyers are asking now. Wrong licensing, security or compliance answers need same-day escalation to the product marketing lead and the COE Team.

Findings

AI answer monitoring: bolddocserver, 2026-10

Run by: Claude (unattended, via pbmaintain) | Location: n/a (no engine queried) | Date: 2026-11-12

Prompt set used (category from the checklist; competitors and the use case from products/bolddocserver/gtm-plan.md):

Summary

• Mention rate: 0 of 0 measured (no previous run for bolddocserver, so no change vs last month)
• Citation rate: 0 of 0 measured (no previous run, so no change vs last month)
• Wrong or outdated facts: none recorded (nothing measured)

Gaps and fixes

No gaps could be identified because no answers were collected.

Not checked

• All 42 rows (P1–P7 × ChatGPT, Gemini, AI Overviews, Copilot, Perplexity, Claude). No MCP servers were connected for this run, so openrouter ask_engine wasn't available. The agent didn't answer the Claude rows itself, because it isn't a clean session (it has repo context), and that would bias the result.
• AI Overviews and Copilot. These need a manual check, which an unattended run can't do.
• Phase 2 (month-over-month comparison, gap hand-off, weekly GTM deltas). There is no current data, and no previous AI monitoring run exists for bolddocserver in reports/publish/.

Choices made without asking (unattended run):

• P3 competitor: Collabora Online, the first open-source editor listed in the GTM plan. Microsoft is the incumbent being replaced, not a head-to-head choice.
• P4 competitor: Office Online Server, which matches GTM plan prompt 4 ("Office Online Server alternatives compared").
• P6 use case: GTM plan prompt 2 ("How do I keep browser editing in SharePoint on-premises after 2026?"). GTM prompt 1 overlaps P1 too closely.

Proposed changes (need approval)

None. No live system or page change is proposed from this run.

Data sources

• demand-gen/search/ai-answer-monitoring.md (checklist, fixed prompt set, category)
• products/bolddocserver/gtm-plan.md (AI-monitoring prompts, competitors)
• reports/publish/ (checked for a previous bolddocserver AI monitoring run: none found)
• raw/prompt-set.json (the exact prompts and engines for the rerun; no answers collected)
• Engines queried: none (no MCP servers connected)

Reminder: add the results to the team's shared results log once the rerun is complete.