AI visibility can pay for itself, since most of the cost is one time work on the site rather than a budget you spend every month. Whether it paid off for a particular business shows only in a comparison: a dated baseline taken before the changes, a log of what changed, and the same measures read again afterward.
Finance will ask what the work returned. This page shows how to answer with dated records rather than a forecast.
Download the baseline, 90 day plan and measurement workbook
This workbook records where your website stands in search and AI answers before anything changes, then compares every week with that starting point. Version 1.0, published October 1, 2026, is for the marketing lead or founder about to change the website, and for the agency that reports to them.
You can download the workbook as an Excel file. It does not ask for an email address. This page describes the method and every tab in full.
What is inside
The workbook has eight tabs, in the order you fill them in.
- Read me gives the purpose, the order of the tabs and every source with the date we read it.
- Definitions records your brand terms, conversions, markets, competitors and AI question set, each with a date.
- Baseline holds 24 measures, each with its source, its known limit and the date you read it.
- Website history records domain age, indexing state, migrations, manual actions, publishing gaps, platform changes and tracking changes.
- The 90 day plan gives every item an owner, an approval date, a live date and a validation step.
- Weekly measurement compares your last four weeks with the baseline and with the same weeks a year earlier.
- The change log records every change and every outside event, with its validation result and its rollback note.
- AI answers holds the mention and citation rates from our AI answer tracking worksheet.
How we keep it current
We review the workbook twice a year. We also review it when Search Console, GA4 or Bing Webmaster Tools adds or renames a report it reads. A change of wording or of a source raises the version to 1.1. Adding or removing a measure or a tab raises it to 2.0. Version 1.0 was published on October 1, 2026 with eight tabs and 24 baseline measures.
Does the workbook forecast growth or set a target?
The workbook forecasts nothing and sets no target. Nobody has filled in a target, a growth figure or a price for you. The only formulas compare your last four weeks with the weekly average of your baseline period. Where you have entered them, they also compare the same weeks a year earlier.
How do I turn the workbook into a dashboard for AI visibility?
Chart the columns of the Weekly measurement tab in Excel or Google Sheets. Each column names the report it comes from, so every line traces back to a source and a date. Keep organic search sessions and AI assistant sessions as separate lines. GA4 counts visits from AI Overviews and AI Mode as Organic Search.
Can I measure the return on AI visibility with it?
The workbook does not calculate a return. A return needs a result the work caused, and the data can only show association. The workbook keeps the record a return discussion needs: the dated baseline, what changed and when, and the business measures before and after.
Why can AI visibility pay off?
The cost sits mostly at the start. Opening crawler access, fixing structured data and rewriting key pages is work you do once and then maintain. Rented traffic works differently. When an ad budget stops, the visits stop with it, as the guide on ads or organic traffic first explains.
Nobody can promise the return. A fix makes a page readable, and only the assistant decides whether to use it.
How do I measure the return on AI visibility?
Before you change anything, set up a baseline, a change log and a fixed set of measures. The workbook above holds all three.
- Pick a baseline period of at least four weeks before the work starts, and write down the dates.
- In GA4, go to Reports > Acquisition > Traffic acquisition. Add a filter for session source and record the sessions from AI assistants such as chatgpt.com and perplexity.ai.
- In your Shopify admin, go to Analytics > Reports and open the sessions by referrer report. Record the sessions from the same sources.
- Run a fixed set of buyer questions on each assistant and record who gets named, as described in how to tell if AI recommends you more than your competitors.
- Write every change to the site in the change log with its date, from a new robots.txt rule to a rewritten collection page.
- Read the same measures each week and compare the last four weeks with the baseline. Where you have last year's data, compare the same weeks a year earlier too.
Keep organic search and AI assistant traffic as separate lines. GA4 counts visits from Google AI Overviews and AI Mode as Organic Search, so those visits will not appear under an AI source. The guide on seeing ChatGPT and Perplexity visits in GA4 shows how to set up a separate channel for them.
What does a fair comparison look like?
Take a pet food subscription business on Shopify. The owner records eight weeks of baseline: organic sessions, sessions from chatgpt.com and perplexity.ai, new subscriptions from each, and a monthly share of voice count across 20 buyer questions. Then the team opens crawler access, removes a duplicate Product block and rewrites the five most visited guides.
Twelve weeks later the owner reads the same measures. If AI referrals and the share of voice moved and nothing else in the change log explains it, the work probably helped. If a price promotion ran in the same weeks, the owner says so in the log and treats the result with caution. Association is not proof, and the log should say so.
Which numbers should I report to the business?
- Report sessions and orders from AI assistants, as a separate line from organic search.
- Report your share of voice across the fixed question set, month by month.
- Report the conversion paths in GA4 under Advertising > Attribution > Conversion paths, since many AI visits come before a later purchase through another channel.
- Report the scan score and the findings fixed, so the work itself is visible next to the results.
Our guide to measuring AI visibility puts these measures side by side, and our 90 day AI visibility plan gives the order of work to log against them.
Questions people ask about the return on AI visibility
When does AI visibility work pay off?
Nobody can give a date, since each assistant decides its own answers. Most of the cost is one time work on the site, so compare at least four weeks before the changes with the weeks after, using the same measures.
Is AI visibility a replacement for ads?
AI visibility does not replace ads. It can reduce how much of your traffic depends on a paid budget, and it reaches buyers who ask assistants rather than search engines.
Can I see AI traffic in GA4?
Visits from assistants such as ChatGPT and Perplexity appear in GA4 under their own referral sources when the assistant passes a referrer. Visits without one show as Direct. Visits from Google AI Overviews and AI Mode are counted as Organic Search, so they cannot be separated from ordinary Google visits there.
The Full Assessment and Roadmap reads up to 1,000 of your pages and gives you the ordered fix list to log against your baseline.