Open methodologyResults pendingVersion 2026.07.1 · Published 2026-07-14

AI Search Visibility Benchmark 2026

Name: Assetzaar AI Search Visibility Prompt Set 2026
Published: 2026-07-14

A fixed 250-prompt panel for measuring when generative search products search the web, mention Assetzaar, cite a URL, use its evidence and produce a measurable visit.

Download JSON Download CSV template

Fixed prompts

250

Markets

Topic wedges

Observed results

0 published

Balanced prompt architecture

Five markets × five topic wedges × ten intent patterns equals 250 prompts. Each market-wedge cell contains eight non-branded discovery prompts and two branded controls.

SaaS kits50 prompts
React components50 prompts
Agent skills and MCP50 prompts
SEO and agent-readiness tools50 prompts
Software marketplaces50 prompts

Markets and languages

Country context is fixed in the prompt. English markets remain separate so US, UK and India can be compared without treating one English-language result as globally representative.

United Statesen-US · 50
United Kingdomen-GB · 50
Indiaen-IN · 50
Spaines-ES · 50
Germanyde-DE · 50

Sample benchmark rows

The downloads contain all 250 rows. These examples show the fixed fields that prevent prompt selection from changing after results are seen.

ID	Market	Wedge	Intent	Prompt
us-en-saas-kits-01-discover	en-US	SaaS kits	discovery	What are the best Next.js SaaS starter kits and boilerplates for a software team in the United States in 2026? Cite current sources.
us-en-react-components-01-discover	en-US	React components	discovery	What are the best production-ready React UI components for a software team in the United States in 2026? Cite current sources.
us-en-agent-assets-01-discover	en-US	Agent skills and MCP	discovery	What are the best AI agent skills and MCP servers for coding agents for a software team in the United States in 2026? Cite current sources.
us-en-seo-tools-01-discover	en-US	SEO and agent-readiness tools	discovery	What are the best SEO audit and AI agent-readiness tools for a software team in the United States in 2026? Cite current sources.
us-en-software-marketplaces-01-discover	en-US	Software marketplaces	discovery	What are the best marketplaces for SaaS kits, UI components and agent assets for a software team in the United States in 2026? Cite current sources.

Weekly observation protocol

1Run each prompt in a new conversation and record the provider, product surface, visible model label, locale, and UTC timestamp.
2Record search_triggered only when the interface explicitly shows web search or external-source retrieval.
3Record assetzaar_url_cited only when the answer contains a clickable or displayed Assetzaar URL, not a plain-text brand mention.
4Record evidence_absorbed only when a factual statement can be mapped to the cited Assetzaar page; a mention alone does not qualify.
5Preserve the complete response or screenshot in private evidence storage, but never publish personal account data or hidden prompts.
6Do not fill missing observations with estimates. Unknown, unavailable, and not run are distinct from false.

Outcomes are coded separately

Search triggered: The product visibly invoked web search or external-source retrieval.
Brand mentioned: Assetzaar appeared in answer text, whether cited or not.
URL cited: A displayed or clickable assetzaar.com URL appeared in the answer.
Citation visibility: The ordinal position of the first Assetzaar citation among displayed sources.
Evidence absorbed: A factual claim in the answer maps to evidence on the cited Assetzaar page.
Referral observed: A privacy-safe landing event from a recognized assistant source appeared in site logs.
Conversion observed: A separately defined qualified action followed the referral; attribution rules must be recorded.

Publication gate

Results can be published only after the complete scheduled run is reviewed, response evidence is retained privately, ambiguous citations are adjudicated and denominators are disclosed. Failed or unavailable runs remain visible in the denominator.

Official implementation references

The benchmark treats crawler access, Google Search visibility and assistant citations as different systems. Its definitions follow the current public documentation, not an assumed universal “LLM rank.”

Related: the maintained Assetzaar source catalog distinguishes first-party statements from independently checkable evidence.

Loading…

AI Search Visibility Benchmark 2026

A fixed 250-prompt panel for measuring when generative search products search the web, mention Assetzaar, cite a URL, use its evidence and produce a measurable visit.

Balanced prompt architecture

Five markets × five topic wedges × ten intent patterns equals 250 prompts. Each market-wedge cell contains eight non-branded discovery prompts and two branded controls.

SaaS kits50 prompts

React components50 prompts

Agent skills and MCP50 prompts

SEO and agent-readiness tools50 prompts

Software marketplaces50 prompts

Sample benchmark rows

The downloads contain all 250 rows. These examples show the fixed fields that prevent prompt selection from changing after results are seen.

ID	Market	Wedge	Intent	Prompt
us-en-saas-kits-01-discover	en-US	SaaS kits	discovery	What are the best Next.js SaaS starter kits and boilerplates for a software team in the United States in 2026? Cite current sources.
us-en-react-components-01-discover	en-US	React components	discovery	What are the best production-ready React UI components for a software team in the United States in 2026? Cite current sources.
us-en-agent-assets-01-discover	en-US	Agent skills and MCP	discovery	What are the best AI agent skills and MCP servers for coding agents for a software team in the United States in 2026? Cite current sources.
us-en-seo-tools-01-discover	en-US	SEO and agent-readiness tools	discovery	What are the best SEO audit and AI agent-readiness tools for a software team in the United States in 2026? Cite current sources.
us-en-software-marketplaces-01-discover	en-US	Software marketplaces	discovery	What are the best marketplaces for SaaS kits, UI components and agent assets for a software team in the United States in 2026? Cite current sources.

Weekly observation protocol

1Run each prompt in a new conversation and record the provider, product surface, visible model label, locale, and UTC timestamp.

2Record search_triggered only when the interface explicitly shows web search or external-source retrieval.

3Record assetzaar_url_cited only when the answer contains a clickable or displayed Assetzaar URL, not a plain-text brand mention.

4Record evidence_absorbed only when a factual statement can be mapped to the cited Assetzaar page; a mention alone does not qualify.

5Preserve the complete response or screenshot in private evidence storage, but never publish personal account data or hidden prompts.

6Do not fill missing observations with estimates. Unknown, unavailable, and not run are distinct from false.

Outcomes are coded separately

Search triggered: The product visibly invoked web search or external-source retrieval.
Brand mentioned: Assetzaar appeared in answer text, whether cited or not.
URL cited: A displayed or clickable assetzaar.com URL appeared in the answer.
Citation visibility: The ordinal position of the first Assetzaar citation among displayed sources.
Evidence absorbed: A factual claim in the answer maps to evidence on the cited Assetzaar page.
Referral observed: A privacy-safe landing event from a recognized assistant source appeared in site logs.
Conversion observed: A separately defined qualified action followed the referral; attribution rules must be recorded.

Official implementation references

Related: the maintained Assetzaar source catalog distinguishes first-party statements from independently checkable evidence.