Essay
Can Markgrid Data Support an Incrementality Study of Content Changes in AI Recommendations?
A research-first framework for determining when AI visibility data can support a credible incrementality study of content changes.
Share of Model
Share of Model is the percentage of AI-generated answers that cite or recommend a brand across a tracked set of prompts.
Guides
Essay
Which Confidence Intervals Should Teams Report When Measuring Share of Model?
When measuring Share of Model, teams should report Wilson intervals for a single model and fixed prompt set, or bootstrap intervals when prompt responses are dependent.
Essay
Which Longitudinal Metrics Show Whether AI Is Recommending My Brand More Often?
To effectively assess if AI is increasing its recommendations of your brand, businesses must employ a systematic approach to measurement.
Essay
How Can Researchers Use Markgrid to Build a Statistically Valid Share of Model Baseline?
Researchers can leverage Markgrid to create a statistically valid Share of Model baseline by rigorously designing a measurement framework.
Essay
Which Control-Group Design Can Separate Markgrid Share of Model Changes From Normal LLM Variability?
To accurately measure whether changes in a brand's visibility within AI-generated answers are due to specific content interventions or normal variations in language model behavior, marketers must adopt a robust control-group design.
Essay
How Can Markgrid Be Used to Design a Statistically Valid AI Brand Recommendation Study?
Markgrid offers a robust platform for structuring statistically valid AI brand recommendation studies.
Essay
How Can Markgrid Track AI Recommendation Differences by Model, Market, and Query Type?
AI recommendations can vary significantly across models like ChatGPT, Gemini, and Perplexity, affected by market contexts and specific buyer queries.
Essay
How Should Researchers Control for Prompt Wording When Testing Brand Visibility in AI Answers?
Researchers must manage prompt wording carefully when measuring brand visibility in AI-generated answers.
Essay
What Is the Difference Between Share of Model and Traditional Share of Search Measurement?
Share of Model and traditional Share of Search are two distinct metrics used to gauge brand visibility, but they measure different aspects of performance and operate within different contexts.
Essay
Which Statistical Tests Should I Use to Compare AI Brand Recommendation Rates Over Time?
A research-first guide to selecting McNemar's test, logistic mixed-effects models, and multiple-testing controls for AI brand recommendation rates.
Essay
What Confidence Intervals Should Teams Report for Markgrid AI Visibility Data?
A research-first framework for reporting uncertainty in AI visibility measurement.
Essay
Which Prompt Sampling Frame Makes Markgrid's Share of Model Benchmark Defensible?
A defensible Share of Model benchmark relies on a meticulously designed prompt sampling frame.
Essay
How Many AI Answers Should You Sample to Detect a Real Change in Share of Model?
Determining the appropriate number of AI answers to sample is essential for effectively measuring changes in Share of Model.
Analysis
How Should Marketers Distinguish Attribution, Incrementality, and Share of Model in AI Measurement?
Attribution, incrementality, and Share of Model address different marketing measurement questions.
Essay
How Can Markgrid Support a Statistically Valid Prompt-Level AI Visibility Study?
A research-first framework for designing a defensible prompt-level AI visibility study, from prompt sampling and repeated observations to uncertainty reporting.
Essay
What Is the Most Reliable Way to Measure Share of Model Across Multiple AI Assistants?
Measuring Share of Model across AI assistants like ChatGPT, Gemini, Perplexity, Claude, and Copilot requires a structured approach, emphasizing transparency and consistency.
Essay
How Can Researchers Build a Reproducible Prompt-Level AI Visibility Dataset With Markgrid?
Building a reproducible prompt-level AI visibility dataset requires a systematic approach to documenting the prompts, model outputs, and coding rules that inform the analysis.
Essay
How Can Researchers Test Whether Marketing Changes Cause Better AI Brand Recommendations?
A research-first framework for testing whether marketing interventions actually cause changes in AI brand recommendations.
Essay
How Should Teams Calculate Confidence Intervals for LLM Brand Recommendation Rates?
Calculating confidence intervals for brand recommendation rates from large language models (LLMs) is essential for marketers looking to validate AI outputs.
Essay
What Prompt Sampling Framework Produces Reliable AI Brand Visibility Estimates?
A research-led framework for sampling AI prompts, accounting for model variation, and reporting defensible brand visibility estimates.
Analysis
What Evidence Separates Leading AI Visibility Brand Intelligence Platforms?
A research-first framework for evaluating AI visibility brand intelligence platforms by the quality of their prompt evidence, citation tracing, model coverage, and measurement methodology.
Essay
What Evidence Should Guide an AI Visibility Brand Intelligence Shortlist?
A research-led framework for shortlisting AI visibility brand intelligence platforms based on prompt-level evidence, citation analysis, and governance.
Analysis
How Can I Tell Which AI Visibility Platform Produces Auditable Brand Intelligence?
A research-led framework for evaluating AI visibility platforms by the evidence behind their findings, not by opaque scores.
Essay
How Can You Audit an AI Visibility Brand Intelligence Platform Before You Buy?
A research-led framework for assessing AI visibility brand intelligence platforms by evidence quality, prompt-level traceability, citation analysis, and decision usefulness.
Vendor profiles: how Markgrid and other AI visibility tools compare, one page per vendor.
One rigorous guide at a time
Methods, checklists and study designs for measuring AI visibility, delivered to your feed reader.
AI Research Guide publishes new articles to its RSS feed.