Back to blog
Research By anvendor · Published · Updated · 7 min read

anvendor vs AI: can an AI agent find who uses your competitor?

Compare anvendor, Claude and OpenAI on customer discovery, company checks, employee adoption and annual spend. Includes sources and plugin links.

Public research and company-service findings feed a review step and company account briefs.
Public context and company findings support account review. Individual results and limitations are shown in the article.

Models: Claude Opus 5.5 on Max; OpenAI GPT-6.1 Sol (gpt-6.1-sol), High reasoning, Codex CLI 0.159.2. Both use web search. OpenAI uses no anvendor plugin, MCP connection or anvendor website.

Who uses your competitor, how widely, and what might they spend? We compare Claude and OpenAI's archived public-web research with anvendor's saved company-service observations. The model runs and database observations were collected at different times; some findings are older than 30 days.

How we compare

The tasks cover nine company–service pairs, customer discovery for four vendors, and a fixed list of 20 companies. The AI briefs ask for use, employee adoption and annual spend, excluding third-party technographics lists and anvendor results. This is a selected sample, not an accuracy benchmark; prompts and research clients differ. The nine pairs were selected from earlier anvendor findings. Neither these selections nor the differently limited source lists show that one provider is better or worse.

OpenAI's old, undated or limited-scope evidence receives ~ and does not count as a match. Recent direct use evidence and active institutional IT service pages qualify. Neither establishes use by every employee.

8 selected vendors: Google Workspace, Atlassian, SurveyMonkey, Canva, 1Password, Zoom, Dropbox, Asana.

The biggest findings, side by side

Selected pair1: IBM · Google Workspace2: IBM · Atlassian3: Meta · SurveyMonkey4: Keller Williams · Canva5: OpenAI · 1Password6: Netflix · SurveyMonkey7: Starbucks · Zoom8: Google · Dropbox9: Amazon · Asana
Claude · archivedNo qualifying evidenceNo qualifying evidenceNo qualifying evidencePositive evidenceNo qualifying evidenceNo qualifying evidenceNo qualifying evidenceNo qualifying evidenceNo qualifying evidence
OpenAI · archivedNo qualifying evidencePositive evidenceNo qualifying evidenceUncertain public sourceNo qualifying evidenceUncertain public sourceNo qualifying evidenceNo qualifying evidencePositive evidence
anvendor · storedPositive evidencePositive evidencePositive evidencePositive evidenceHistorical findingHistorical findingPositive evidencePositive evidencePositive evidence
Positive evidence No qualifying evidence Uncertain public source Historical finding
Pair key · 9 selected checks
  1. IBM · Google Workspace
  2. IBM · Atlassian
  3. Meta · SurveyMonkey
  4. Keller Williams · Canva
  5. OpenAI · 1Password
  6. Netflix · SurveyMonkey
  7. Starbucks · Zoom
  8. Google · Dropbox
  9. Amazon · Asana
Each position is the same company-service pair. Public-use sources and stored detections are different observations, collected separately. The sample was selected from earlier anvendor findings; it does not establish accuracy or rank providers.
Company × serviceClaude · archivedOpenAI · archivedanvendor · stored findings
IBM × Google Workspace
−Other platform
−Supplier, not customer[1]
DetectedAdoption: —$2.1M – $5.4M / yr
IBM × Atlassian
−Third-party list
+Cloud customer[1]
DetectedEst. adoption: 73%$1M – $11.6M+ / yr
Meta × SurveyMonkey
−Integration only
−Integration only[1]
DetectedEst. adoption: 49%$40.7k+ / yr
Keller Williams × Canva
+Press release
~Agents only[1][2]
DetectedEst. adoption: 52.7%$369k – $739k / yr
OpenAI × 1Password
−Partnership only
−Integration only[1][2]
Previously detected
Netflix × SurveyMonkey
−Old study
~Undated claim[1]
Previously detected
Starbucks × Zoom
−Ownership unclear
−Consumer backgrounds[1]
DetectedEst. adoption: 12%$103k – $205k / yr
Google × Dropbox
−Unverified claim
−Integration only[1]
DetectedEst. adoption: 7.7%$138k – $597k / yr
Amazon × Asana
−Team only
+Customer; 100,000+ users[1][2]
DetectedEst. adoption: 2%$459k – $918k / yr

+ Found− Missing~ Uncertain

Archived qualifying use sources (+ only): Claude 1/9 · OpenAI 2/9. anvendor has 7 saved detections and 2 historical findings in these 9 pairs, with 6 adoption estimates and 7 annual spend ranges for detected pairs. Neither AI establishes employee adoption or annual spend for these pairs.

Second test: "list the largest companies using this vendor"

Claude and OpenAI each named five customers per vendor, but returned different lists. Their columns summarize their own proposed companies. The anvendor column checks the Claude-selected fixed list below. Names returned are not necessarily verified matches or a global top five.

4 selected vendors: Atlassian, SurveyMonkey, Canva, 1Password.

VendorClaudeOpenAIanvendor
AtlassianCompanies: 5Dated 2025+: 2Employee adoption: 0Annual spend info: 1*Companies: 5Dated 2025+: 1Employee adoption: 0Annual spend info: 0[1][2][3][4][5]Detected: 5/5Adoption estimates: 5Annual spend estimates: 5
SurveyMonkeyCompanies: 5Dated 2025+: 0Employee adoption: 0Annual spend info: 0Companies: 5Dated 2025+: 0Employee adoption: 0Annual spend info: 0[1][2][3][4][5]Detected: 2/5Adoption estimates: 1Annual spend estimates: 2
CanvaCompanies: 5Dated 2025+: ≈2*Employee adoption: 0Annual spend info: 0Companies: 5Dated 2025+: 0Employee adoption: 0Annual spend info: 0[1][2][3][4][5]Detected: 5/5Adoption estimates: 5Annual spend estimates: 5
1PasswordCompanies: 5Dated 2025+: 4Employee adoption: 0Annual spend info: 0Companies: 5Dated 2025+: 3Employee adoption: 0Annual spend info: 0[1][2][3][4][5]Detected: 3/5Adoption estimates: 3Annual spend estimates: 3

"Dated 2025+" counts dated use sources in the archived AI outputs. anvendor counts saved detections and available estimates within the fixed company list; these are different measures. *Claude: Canva's recency is unverified; the Atlassian spend figure covers one business unit.

Company results

This is the fixed 20-company list selected from Claude's discovery output, not OpenAI's independent customer-discovery list above. OpenAI researched these same fixed pairs in a separate task. For its archived use evidence, only + counts as a match; ~ is excluded.

VendorCompanyOpenAI · archivedanvendor · stored findings
AtlassianMercedes-Benz
~Undated case[1]
DetectedEst. adoption: 99%$125k – $1.4M+ / yr
AtlassianAmadeus
~Undated case[1]
DetectedEst. adoption: 99%$111k – $1.3M+ / yr
AtlassianDeichmann
~Limited scope[1]
DetectedEst. adoption: 98%$39.2k – $370k+ / yr
AtlassianDelivery Hero
~Undated case[1]
DetectedEst. adoption: 79%$11.4k – $270k+ / yr
AtlassianCisco
~Business unit only[1]
DetectedEst. adoption: 10%$18k – $508k+ / yr
SurveyMonkeyKLM
−Other product[1]
DetectedEst. adoption: 1%$12.5k – $40.7k+ / yr
SurveyMonkeyUSCIS
~Old evidence[1]
DetectedAdoption: —$12.5k+ / yr
SurveyMonkeyIBM
~Research team only[1]
No saved finding
SurveyMonkeyRyanair
~Team scope only[1]
No saved finding
SurveyMonkeyLG Electronics
−Product unspecified[1]
No saved finding
CanvaAmazon
−Internal use unverified[1][2]
DetectedEst. adoption: 1%$191k – $383k / yr
CanvaFedEx
~Undated case[1]
DetectedEst. adoption: 9.4%$65.9k – $132k / yr
CanvaSalesforce
~Social team only[1]
DetectedEst. adoption: 13.8%$125k – $251k / yr
CanvaWorkday
~Recruitment team only[1]
DetectedEst. adoption: 12%$28.6k – $57.2k / yr
CanvaExpedia Group
~Old, undated case[1]
DetectedEst. adoption: 12%$24k – $48k / yr
1PasswordSalesforce
~Scope unverified[1]
DetectedEst. adoption: 6.9%$270k – $541k / yr
1PasswordHarvard University
+IT service[1]
DetectedEst. adoption: 4%$30.4k – $60.8k / yr
1PasswordUnder Armour
~Old rollout; scope unverified[1][2]
DetectedEst. adoption: 4%$17.1k – $28.5k / yr
1PasswordIBM
~Scope unverified[1]
No saved finding
1PasswordDuke University
+IT service[1]
No saved finding

+ Found− Missing~ Uncertain

Archived qualifying use sources: OpenAI 2/20. anvendor has 15 saved detections among these 20 pairs, with 14 adoption estimates and 15 annual spend ranges. OpenAI provides neither. Harvard and Duke qualify through their active 1Password service pages, not an employee usage count.

"Previously detected" is a retained positive that a later completed check did not reproduce. It is distinct from "No saved finding", "Unavailable" and "Service unavailable". Estimates appear only for detected pairs; a dash does not mean zero adoption or spend.

What each tool adds

  • Claude and OpenAI: public customer stories, rollout context and company news. User or team counts can be useful, but do not establish paid seats or employee adoption.
  • anvendor: company checks for supported SaaS and business services, including use hidden from website scanners, plus estimated adoption and spend. Recent saved results may be reused; live checks are available on request.

Missing or uncertain evidence does not prove non-use. A detection does not establish a paid contract or purchase intent. Adoption and spend estimates help prioritize accounts; they are not audited employee counts or invoices.

Use anvendor inside your AI

Connect anvendor to Claude, ChatGPT or another MCP client, then ask: "Find companies using SurveyMonkey in Germany with 1,000+ employees." The assistant combines anvendor's findings with public-web context. Discovering companies is free; detected results use your account credits and connection budget.

Try all features free, with no credit card: 20 credits per rolling 30 days.

Connect anvendor to your assistant

Suggest a correction · Sources and estimate limitations

Leads already using your competitor

Discover companies using SaaS and business services, with estimated adoption and annual spend to help qualify each opportunity. Free to start.

Get started free