AI Visibility Without Attribution Is a Vanity Metric

A new category of marketing metrics is rapidly appearing on dashboards:

AI mentions. Citation share. AI share of voice. Recommendation frequency. Prompt visibility.

These metrics answer an important question:

Are AI systems talking about us?

But businesses eventually need to answer a harder one:

What happened because they did?

That distinction will define the next phase of AI discovery measurement.

Visibility is the beginning of the journey

Imagine that ChatGPT recommends your company 30% more often this month.

That sounds positive.

But several very different things could be happening.

The additional recommendations could generate substantial qualified traffic.

They could influence users who later search for the brand directly.

They could send visitors who immediately leave.

They could assist purchases that happen days later through another channel.

They could have almost no commercial effect at all.

The visibility metric alone cannot tell you which scenario is happening.

This does not make visibility unimportant.

It means visibility is an upstream signal.

Businesses still need to connect it to downstream behavior.

AI discovery is already becoming a measurable traffic channel

This is increasingly possible.

Adobe now provides tools for examining referral traffic from AI platforms and connecting it with engagement and conversion information through analytics integrations. Its documentation explicitly describes measuring how AI-driven discovery translates into website behavior and business outcomes.¹

Adobe also reported a roughly 60% increase in referral traffic from ChatGPT during May 2026 in data it analyzed.¹

At the same time, AI platforms are moving closer to commerce itself.

OpenAI's Agentic Commerce Protocol allows structured merchant information to participate directly in ChatGPT product discovery, while its shopping experiences increasingly connect recommendation with merchant sites and purchasing decisions.²

Google is similarly building commerce experiences into AI-powered discovery, including conversational shopping and agentic commerce infrastructure.

Egaki tracks this shift directly, connecting AI-referred sessions to the same analytics stack marketing teams already use.

The distance between AI recommendation and commercial action is shrinking.

Measurement needs to catch up.

The old funnel is becoming harder to observe

Traditional digital attribution benefited from relatively explicit paths.

An ad produced a click. The click produced a session. The session produced a conversion.

Marketers could attach identifiers to the interaction and build attribution models around them.

AI discovery creates more complicated journeys.

Consider this sequence:

A user asks an AI assistant for recommendations.

The assistant mentions three brands.

The user asks follow-up questions.

One brand remains in the final shortlist.

The user visits the company's website. They read a comparison page.

Two days later they return through Google.

A week later they purchase after opening an email.

Which channel deserves credit?

The final email clearly mattered. Google mattered. The website mattered.

But the original AI interaction may have introduced the brand and shaped the shortlist.

Last-click attribution will not capture that.

Egaki is built to reconstruct that fuller path instead of crediting only the last click.

Egaki separates three different signals

A useful AI discovery measurement system distinguishes at least three layers.

1. Visibility

Did the brand appear? This includes mentions, citations, recommendation frequency, competitive share, position within answers, and sentiment and framing.

Visibility tells you what is happening inside the AI environment.

2. Engagement

What happened after exposure? This can include AI-referred sessions, pages viewed, scroll depth, time and engagement, internal navigation, product exploration, and outbound interactions.

Engagement tells you whether the discovery generated meaningful interest.

3. Outcomes

Did the journey create business value? Depending on the company, that could mean lead submission, account creation, demo request, checkout, purchase, subscription, qualified pipeline, or revenue.

Outcomes tell you whether AI discovery actually matters commercially.

The most valuable analysis connects all three.

Visibility → Engagement → Outcome

Citations and referrals are not the same thing

This distinction is easy to overlook.

An AI system can use information from a company without generating a website visit.

It might cite the company. It might mention the brand without a link. It might synthesize information and influence the customer's decision entirely inside the conversation.

This is becoming particularly important as search behavior shifts toward AI-generated answers.

In a 2025 analysis of Google search behavior, Pew Research found that users encountering AI summaries clicked traditional search results less frequently than users who did not encounter one. Direct clicks on sources inside the AI summaries were uncommon.³

That means traffic alone can understate influence.

A company could become more important to a customer's decision while receiving fewer observable visits.

Marketers therefore need both on-platform visibility measurement and off-platform behavior measurement.

Neither alone gives the complete picture.

Egaki measures both: on-platform visibility across AI answers, and off-platform behavior once a visitor arrives.

The unit of attribution should become the journey

AI discovery also makes session-based analytics increasingly limiting.

A single customer journey may span:

ChatGPT → brand website → Google → brand website → email → purchase.

If every session is analyzed independently, the causal story fragments.

A better system attempts to reconstruct the journey.

What introduced the brand? What increased confidence? What information did the visitor consume? Which interactions moved them toward conversion? Where did they ultimately act?

The objective is not to assign 100% of a purchase to one touchpoint. That usually creates false precision.

The objective is to understand the role each touchpoint played.

Content itself becomes measurable

Once AI discovery and downstream behavior are connected, another opportunity appears.

Brands can begin evaluating the economic value of individual pieces of content.

Suppose an article is frequently cited by AI systems.

Those citations generate visits.

Visitors who enter through that content navigate disproportionately toward two product pages.

Those sessions show higher purchase intent.

Eventually, a portion converts.

Now the article is no longer simply "content."

It is an asset participating in a measurable customer journey.

This creates the foundation for content valuation.

Instead of evaluating an article only through pageviews or rankings, companies can ask:

• How often does AI use this page?
• For which personas and intents?
• Which customer journeys does it initiate or assist?
• Which products does it influence?
• How much commercial value is associated with those journeys?

That begins to connect content strategy directly to business outcomes.

This is Egaki's content valuation model in practice, and how we help clients decide what to publish next.

Attribution also changes optimization

Once these signals connect, AI discovery becomes a feedback system.

Suppose your brand rarely appears for a valuable customer segment.

You identify the missing intent. You publish stronger evidence addressing that decision context. AI recommendation frequency increases. Relevant referral traffic increases. Those visitors convert well.

You now have evidence that the intervention mattered.

The process becomes:

Understand demand → Improve content → Measure visibility → Observe journeys → Measure outcomes → Improve again

This is much more useful than optimizing toward a visibility score in isolation.

It tells you where to invest.

Egaki runs this loop continuously for the clients we work with.

Not every AI mention is equally valuable

This may ultimately be the biggest change.

Two AI recommendations should not necessarily carry equal weight.

One might occur for a low-value informational question. Another might occur when a customer is actively choosing between vendors.

One might reach a user with little purchase intent. Another might place the brand into the final consideration set for a high-value transaction.

A mature measurement system considers both:

How often the brand appears

and

How commercially meaningful the underlying opportunity is.

This is why Egaki's platform connects visibility to engagement and revenue in one view instead of stopping at a mention count. Pro-tier customers get performance attribution built in, tying content and positioning changes to the AI-referred sessions and conversions they produce.

The long-term measurement problem spans an entire chain:

Persona → Intent → AI Recommendation → Citation → Journey → Conversion → Revenue

Every step provides information about the next one.

The metric that matters is business impact

AI visibility metrics are valuable because they reveal a part of the customer journey that previously did not exist.

But they should not become the destination.

Marketing has seen this pattern before.

Impressions mattered. Then clicks mattered. Then conversions mattered. Eventually businesses demanded revenue attribution.

AI discovery will follow the same progression.

Today, many companies are asking:

"Are we showing up in ChatGPT?"

Soon the more important question will be:

"Which customers are discovering us through AI, why are we being recommended, and what business value does that discovery create?"

That is the measurement problem worth solving.