AI Visibility Tracking Template

A copy-ready template to track AI search visibility across Google AI Overviews, Gemini, ChatGPT Search, and Perplexity. Includes daily logging, weekly rollups, and anomaly flags.

This template supports the workflow described in How to Track AI Search Visibility. Copy it into Google Sheets, Airtable, or Notion and start logging today. Each row captures one prompt × platform observation. The weekly rollup section below explains how to aggregate daily data into actionable trends.

Column reference

Column	What to enter
Date	YYYY-MM-DD of the observation.
Platform	Google AIO / Gemini grounded / ChatGPT Search / Perplexity.
Prompt	The exact prompt text you ran.
ICP / Funnel stage	Problem / Evaluation / Implementation (maps to buyer journey).
Brand mentioned?	Y or N — did the AI name your brand in the answer?
Competitors mentioned	Comma-separated list of competitor brands mentioned.
Citations present?	Y or N — did the answer include any source citations?
Your domain cited?	Y or N — was your domain among the cited sources?
Cited URL(s)	Exact URLs from your domain that were cited.
Top third-party cited sources	Domains (not full URLs) of non-your-brand sources cited.
Notes	Free text: anything notable about the answer.
Anomaly flag	Y or N — did this observation deviate sharply from recent pattern?
Anomaly reason	Short description of why you flagged it (e.g., "lost citation to competitor docs").

Sample data (copy-ready)

Date	Platform	Prompt	Funnel	Mentioned?	Competitors	Citations?	Your cited?	Cited URLs	3rd-party sources	Notes	Anomaly	Why
2026-02-19	Google AIO	best tools for X in scenario Y	Evaluation	Y	Competitor A, B	Y	Y	example.com/comparisons	competitor-a.com	First AIO appearance	N
2026-02-19	ChatGPT Search	how to configure X for compliance	Implementation	Y		Y	N		docs.competitor-a.com	Competitor docs cited	Y	Lost citation
2026-02-19	Perplexity	X vs Y for enterprise	Evaluation	Y	Competitor A	Y	Y	example.com/pricing	g2.com; trustradius.com	Pricing page cited	N
2026-02-19	Gemini grounded	what is X and how does it work	Problem	N	Competitor A, C	Y	N		competitor-a.com; wikipedia.org	Brand absent	Y	Not mentioned

CSV (paste into your tool)

CSV

Date,Platform,Prompt,ICP / Funnel stage,Brand mentioned?,Competitors mentioned,Citations present?,Your domain cited?,Cited URL(s),Top third-party cited sources (domains),Notes,Anomaly flag,Anomaly reason
2026-02-19,Google AIO,"best tools for X in scenario Y",Evaluation,Y,"Competitor A, Competitor B",Y,Y,https://example.com/comparisons,competitor-a.com; competitor-b.com,First appearance in AIO for this prompt,N,
2026-02-19,ChatGPT Search,"how to configure X for compliance",Implementation,Y,,Y,N,,docs.competitor-a.com,Competitor docs cited instead of ours,Y,Lost citation to competitor docs
2026-02-19,Perplexity,"X vs Y for enterprise",Evaluation,Y,"Competitor A",Y,Y,https://example.com/pricing,g2.com; trustradius.com,Pricing page cited directly,N,
2026-02-19,Gemini grounded,"what is X and how does it work",Problem,N,"Competitor A, Competitor C",Y,N,,competitor-a.com; wikipedia.org,Not mentioned at all — investigate,Y,Brand absent from answer

Weekly rollup

Daily rows give you anomaly detection. Weekly rollups give you trend. Here's how to aggregate.

Compute the weekly median

For each metric (mention rate, citation rate, coverage), take the median of the 7 daily values rather than the mean. The median resists outliers, so a single spike or drop on one day won't distort the weekly number.

In Google Sheets: =MEDIAN(range_of_7_daily_values)

Compute the 25th–75th percentile band

This tells you how volatile the week was. A narrow band means stable visibility. A wide band means something is shifting.

In Google Sheets: =PERCENTILE(range, 0.25) and =PERCENTILE(range, 0.75)

Flag anomalies vs the 7-day median

Any daily value that falls outside 1.5× the interquartile range (IQR = 75th − 25th percentile) from the median gets an anomaly flag. This is a simple, robust method that prevents you from chasing noise.

In Google Sheets:

IQR = PERCENTILE(range, 0.75) - PERCENTILE(range, 0.25)
Anomaly if: value < median - 1.5 * IQR or value > median + 1.5 * IQR

When you spot an anomaly, write a short "why" in the anomaly reason column. Over weeks, these notes become your changelog for what actually moved the needle.