This article is part of our AI visibility benchmarks for 2026.
Your CFO looks at the line in the budget and asks a fair question: what does this bring us?
You could answer with a formula. One extra client a month, times the value of a client, against the cost of the tool. It looks clean on a slide. It is also built on a number nobody can measure, and a good CFO will spot that in thirty seconds.
There is a better answer. It is less spectacular and much harder to knock down: here is what we measure, here is what moved after each action, here is what we still cannot know, and here is what waiting would cost us.
This article walks through that answer, for a marketing team without a dedicated SEO that has to defend an AI visibility budget in front of its leadership.
Why is the revenue impact hard to put in a spreadsheet?
Because the decisive moment happens where you cannot see it.
A buyer asks ChatGPT for the best options in your category. The answer names three brands. Your buyer reads it, maybe asks a follow-up question, then closes the window. Two days later, they type a brand name into Google, or go straight to its site. In your analytics, that visit is a branded search or direct traffic. The conversation that created the shortlist left no trace.
ChatGPT now reaches a mainstream audience, and Perplexity, Gemini and Google's AI answers add their own. Part of those users ask exactly that kind of question: which tool, which provider, which brand. Some of those answers send a click. Many do not, because the answer itself was the point.
So the clean formula breaks on its first term. You do not know how many buyers asked, and you do not know how many of them came to you because of what an AI said. Anyone who gives you a precise revenue figure for AI visibility is extrapolating.
That does not make the budget indefensible. It means the defense has to rest on what you can observe, in the right order: are you cited, does it send traffic, did your actions change anything.
What can you measure honestly today?
Three things, and each one answers a question your leadership will ask.
Are we cited, and where?
This is the base of everything. Mentionable asks the questions your buyers ask every day to the AI engines you choose among ChatGPT, Claude, Perplexity, Gemini, Google AI Mode and Google AI Overview, and records whether your brand is named and whether your site is cited as a source.
You read it by question, by AI engine and by country. Countries matter more than most teams expect: each question is asked from its country, in its language, and the brands recommended in France are not the ones recommended in Germany. Next to your own results, you see the competitors cited in your place and your share of voice against them.
For a leadership meeting, that gives a sentence anyone understands: on the questions that matter to us in France, we are named in this share of answers, our main competitor in that share, and here are the questions where we are absent.
Does it send traffic?
Visibility is a leading indicator. Traffic is the next link in the chain.
The problem is that AI traffic hides. Visits from AI engines often arrive without a referrer, so Google Analytics files them as direct traffic, mixed with bookmarks and typed URLs. Mentionable connects to your Google Analytics 4 property and separates the visits AI engines send you: sessions and users, their share of your total traffic, the trend over time, the AI engine behind each visit and the pages they land on.
That changes the conversation. "Direct traffic is up" is a shrug. "ChatGPT and Perplexity sent this many visits last month, mostly to our comparison pages" is a fact your leadership can follow month after month.
Did our actions change anything?
This is the question that follows any budget, and the hardest one to answer without a method. AI answers move on their own: the sites cited on a question often change without anyone doing anything. If your visibility goes up after you rewrite a page, was it the page or the market?
Mentionable's Progress page answers it action by action. Each page fix and each action marked done is measured on the questions it targets, over the 4 weeks before and the 4 weeks after. Then the same change is measured on the questions you did not touch, over the same dates, and subtracted. If everything rose by 6 points, an action that rose by 6 points did nothing on its own. If its questions rose by 14 while the others rose by 4, the action gets credit for 10.
No figure appears before 3 scans on each side, and an action is only marked "no visible effect" once its window is over. That is the kind of rigor that survives a skeptical finance director, because the method is on the page.
What stays uncertain?
Saying it out loud makes the rest of your case stronger.
Revenue attribution stays partial. Google Analytics 4 sees the visits that come from a click in an AI answer. It does not see the buyer who read ChatGPT's answer, closed it and searched your brand name a week later. A rise in branded searches after your visibility grows is a reasonable signal, not a proof. Mentionable measures visibility, citations and AI traffic. It does not measure your revenue, and it does not claim to.
The questions you track are your choice. They represent what your buyers ask, but real conversations are private and phrased in a thousand ways. Pick questions close to purchase decisions, review them as your offer changes, and say so in your reporting.
Timing varies by AI engine. Engines do not pick up a change at the same speed. That is why Progress shows a badge per engine, and why a single week of data is never a verdict.
Some effects take longer than a reporting cycle. A mention on a comparison site that AI engines read can take weeks to show up in answers. Plan your reviews monthly, not weekly.
None of this is a reason to wait. It is the frame in which you present your numbers, so nobody reads more into them than they say.
What does doing nothing cost?
You will not get an honest euro figure here either. But the cost of inaction is concrete, and you can describe it without inventing anything.
The shortlist forms without you. Every buyer who asks an AI for options in your category gets an answer today. If your brand is not in it, your competitors take that first slot of attention, and you never see the lead you lost. No referral, no bounce, no line in any report.
The sources keep repeating your competitors. AI engines build their answers from a short list of comparison sites, media, directories and forums. Each month those sites name your competitors and not you, the gap holds. Getting into those pages takes time, and the brands that start earlier are the ones the next answers repeat.
You lose your baseline. This one is underestimated. To show your leadership what an action changed, you need a "before". A team that starts measuring the day it starts acting has no reference point. Measuring a few weeks earlier, even before any action, gives you the before that makes every later result credible.
You keep guessing. Without measurement, the team keeps doing what it believes works and drops what it cannot prove. That has a cost too, in time spent on content that never gets cited.
How do you build the case for your leadership?
A one-page structure works better than a deck. Four blocks, updated each month.
Where we stand. Visibility on your key questions, by country, next to your two or three main competitors. The weekly email already gives the visibility rate, its trend and the mentions won and lost, question by question and engine by engine. It reaches the account owner, and can be forwarded as is.
What we did. The actions of the month: pages published or rewritten, fixes applied, sources obtained. The History tab of the Progress page keeps that log, including what you add by hand, like a press article or a partnership.
What moved. The actions marked "it worked" on the Progress page, with their effect in points, and the AI traffic trend from Google Analytics 4.
What comes next. The next actions of the Action Plan, each with its reason taken from your data: the AI engines that never cite you, the competitors cited instead, the search demand behind the page.
If your leadership lives in Looker Studio, the Mentionable connector sends mention rate, citation rate, share of voice and sentiment into the same reports as Google Analytics and Search Console. Nobody has to learn a new tool to follow the numbers.
What does it take to measure and act?
Measuring alone gives you the diagnosis. The budget conversation gets easier when the same tool also moves the numbers, because the before and after come from one place.
Mentionable covers the three levers of AI visibility. Measure: visibility by country, by question and by AI engine, against your competitors. On-site: a monthly Action Plan, a crawl of your site with each fix to apply, and pages to create delivered as drafts in your brand voice. Off-site: the sites AI engines cite, and from the Influence plan, outreach campaigns run for you to get mentioned there.
| Plan | Price per month | Questions tracked daily | AI engines | Countries | Articles per month | Outreach |
|---|---|---|---|---|---|---|
| Presence | 249 € | 50 | up to 3 | 1 | 4 | Not included |
| Influence | 449 € | 100 | up to 4 | 2 | 8 | 200 sites per month |
| Reference | 799 € | 150 | up to 6 | 3 | 15 | 500 sites per month |
Prices exclude VAT and users are unlimited on every plan.
If your team has no capacity to run the work, Mentionable Agency goes further than the tool: its team builds the strategy, writes and earns the links with its own methods, starting with a measured GEO audit. Mentionable then measures your visibility before, during and after.
The simplest way to start the conversation with your leadership is to come with a baseline. The 7-day trial needs no credit card: enter your URL, keep the questions that matter, and bring your first numbers to the next budget review.
Related articles
- You Ran an AI Visibility Check. Now What?: interpreting your results and taking action
- AI Visibility Benchmarks 2026: what "good" looks like by industry
- 15 Best AI Visibility Tools in 2026 (Pricing, LLMs, Use Cases): how the tracking tools compare
