AI answer monitoring · bolddesk

Headline

None of the four API engines mentions BoldDesk for category, enterprise, Zendesk-alternative or tier-1 automation prompts (0 of 16 rows). Every engine names BoldDesk when the prompt names it (P3, P5, P7: 12 of 12 rows), and most of those answers cite bolddesk.com. Overall mention rate is 12 of 28 and citation rate is 11 of 28. Five answers carry wrong or outdated facts. The most serious is Gemini saying BoldDesk is ISO 27001 certified, which the security page does not claim. Several of the wrong facts trace back to our own inconsistent pages: a KB article, a blog post, and the pricing page showing two different per-agent prices. This is the first logged run (baseline), so there is no month-over-month comparison.

Score

No score. The checklist does not define one, so overall_score is null. Baseline numbers: mention rate 0.428571 (12/28), citation rate 0.392857 (11/28), wrong or outdated facts 5.

Top 3 actions

1. Fix the security claims today (P7-gemini). Find and correct any bolddesk.com page or blog that claims ISO 27001 unless a certificate exists, and reword "PCI DSS compliance" to say that Stripe handles card payments. The checklist says to tell the product marketing lead and the COE Team the same day.
2. Make our own pricing and plan facts consistent. Update support KB article 21603 so it matches the current Scale AI / Momentum AI / Enterprise AI plans and the $40 per 1,000 credits. Remove the $20 vs $35 per-agent conflict on the pricing page. Add a worked AI-credit example to the pricing FAQ, which also answers objection 2.
3. Get into category and alternatives answers. Get listed in the third-party "best AI help desk" and "Zendesk alternatives" roundups the engines cite. Refresh /enterprise-helpdesk-software and /alternative/zendesk with direct-answer tables. Publish the tier-1 automation pillar post.

Findings

AI answer monitoring: bolddesk, 2026-10

Run by: Claude | Location: n/a (API via OpenRouter; server location not reported) | Date: 2026-11-09

Prompt text used: P1 "What are the best AI help desk / AI customer support software?"; P2 "What is the best AI help desk / AI customer support software for an enterprise / midsize company?"; P3 "BoldDesk vs Zendesk: which should I choose?"; P4 "What are the best alternatives to Zendesk?"; P5 "What is BoldDesk, how is it licensed and who uses it?"; P6 "How to automate tier-1 support tickets with AI" (GTM plan AI-monitoring prompt 5); P7 "Is BoldDesk secure and compliant for enterprise use?". Position for P3/P5/P7 is 1 because BoldDesk is the subject of the answer.

Summary

• Mention rate: 12 of 28 (baseline; no previous run). By engine: ChatGPT 3/7, Gemini 3/7, Perplexity 3/7, Claude 3/7. Unprompted (P1, P2, P4, P6): 0 of 16.
• Citation rate: 11 of 28 (baseline). By engine: ChatGPT 3/7, Gemini 2/7, Perplexity 3/7, Claude 3/7.
• Wrong or outdated facts:
  • P3, Claude (Wrong): included AI credits described as a 3-month promotion, then $20 per 1,000, plus a separate $20/agent Copilot. The pricing page says $40 per 1,000, credits reset monthly and there are no automatic overages.
  • P5, Gemini (Wrong): "$99/month for up to 5 agents". The pricing page says unlimited human agents.
  • P5, Claude (Wrong): one "Enterprise (team based)" plan with Scale and Momentum closed as legacy plans, and a $20 credit pack. This comes from our own support KB article 21603.
  • P7, Gemini (Wrong): "ISO/IEC 27001-certified ISMS" and "PCI DSS Compliance". Neither is claimed on the security page.
  • P3, Gemini (Outdated): presents BoldDesk as a low-cost SMB ticketing tool and omits the AI-first plans and unlimited agents.

Gaps and fixes

Hand-off (needs a human): send this gap list to the SEO/GEO program and the blog program, post the baseline in the weekly GTM review, and add the results to the team's shared results log (location TBD).

Not checked

• Google AI Overviews, all prompts (P1–P7): there is no API and the openrouter server lists it as check-by-hand. This run was unattended. Check by hand within the week.
• Microsoft Copilot, all prompts (P1–P7): there is no public API for consumer answers, so it is check-by-hand. Check by hand within the week.
• Month-over-month comparison: there is no earlier BoldDesk run in reports/publish/, so this run is the baseline. previous and delta_pct are null.
• Repeat runs: each prompt was asked once per engine. The checklist asks for a second run when results vary; that wasn't tested.
• Country: the API calls don't report the serving location.
• Phase 2 hand-off: sending the gap list and posting the deltas are human steps. They are listed above but not done.

Proposed changes (need approval)

None of these were made. Each one changes a live website, listing or third-party relationship.

1. Correct or remove the ISO 27001 claim on any bolddesk.com page, and reword the PCI statements (P7-gemini). Urgent.
2. Update support KB article 21603 to match the current plans and credit pricing (P5-claude).
3. Fix the $20 vs $35 per-agent Help Desk price conflict on bolddesk.com/pricing and /pricing/help-desk.
4. Add a worked AI-credit explanation to the pricing page FAQ (P3-claude, objection 2).
5. Reconcile data-residency, hosting and GDPR wording across the security page, KB 15272 and the blogs (P7-chatgpt, P7-claude).
6. Refresh /alternative/zendesk, /blogs/zendesk-vs-bolddesk, /blogs/best-zendesk-alternatives and /enterprise-helpdesk-software with direct-answer tables (P2, P3-gemini, P4).
7. Publish the tier-1 automation pillar post (P6).
8. Third-party outreach: ask eesel.ai to correct its BoldDesk pricing article, make "unlimited human agents" explicit on G2, Capterra and SoftwareSuggest, and pitch inclusion to the category and Zendesk-alternative roundups the engines cite (P1, P2, P4, P5-gemini).

Data sources

• openrouter MCP, ask_engine: 28 calls (7 prompts × 4 engines), 2026-11-10 01:42–01:50 UTC (2026-11-09 local). The models are listed by list_engines: chatgpt = openai/gpt-6.1-terra, gemini = google/gemini-3.8-flash, perplexity = perplexity/sonar, claude = anthropic/claude-sonnet-5.5. Reported API cost was about $1.18 in total. list_engines listed Perplexity with web_search false, but every Perplexity response came back with web_search true and citations.
• Fact check (read-only WebFetch, 2026-11-09): https://www.bolddesk.com/pricing and https://www.bolddesk.com/legal/security. Also the repo references _reference/pricing-and-licensing.md, _reference/objections/bolddesk.md and products/bolddesk/gtm-plan.md.
• Raw files: raw/P1.json–raw/P7.json hold the answer text and cited URLs. In P5.json and P7.json the Gemini grounding-redirect tokens were dropped and only the titles kept. P7.json keeps Perplexity's key claims instead of the full answer. raw/results.csv has the results table and raw/fact-check-sources.json has the facts used to judge accuracy.
• Choices made in this unattended run: the top competitor is Zendesk (the first incumbent in the GTM plan). P6 is GTM plan AI-monitoring prompt 5. {category} was inserted verbatim, slash included, so the prompts stay comparable month to month. When a fact couldn't be confirmed or disproved from our public pages (community license, case-study names, Copilot add-on price), it was noted but not counted as Wrong.