Your brand appeared in 47 AI responses last month. Congratulations. Now answer this: did anyone pay attention?
Visibility without attention is a number that feels good in a deck and means nothing in a boardroom. The AI measurement market has spent two years building elaborate tools to count mentions. Nobody built a tool to measure what those mentions are actually worth. That is the problem BAX Index solves.
A metric is only as good as what it weights
The BAX Index is a composite score from 0 to 100. It measures brand attention in AI-generated responses across four dimensions: visibility, position quality, sentiment, and exposure quality.
Each dimension carries a different weight. Decay-weighted Reach accounts for 45 percent. It measures how far attention actually travels through an AI response, weighted by the positional decay curve. A brand mentioned first reaches 77 percent of users. Mentioned last: 18 percent. Reach without decay weighting counts both the same. BAX does not.
Sentiment accounts for 25 percent. Not binary positive or negative, but weighted by cognitive load. A glowing endorsement buried inside a dense paragraph of technical caveats carries less attention value than a neutral mention in a direct, clean sentence. The model reads context, not just polarity.
Exposure Quality accounts for 20 percent. This dimension, powered by the Brand Exposure Index, measures whether the context around your brand actually supports recall and decision-making: the clarity of the surrounding text, the authority of the source, the relevance of the framing to the user's query.
Accuracy accounts for the remaining 10 percent. It measures whether the information AI models associate with your brand is factually correct and aligned with your actual positioning. A brand described inaccurately in a positive response still scores lower than a brand described accurately in a neutral one. Hallucinations are scored negatively.
The decay curve nobody talks about
Every piece of content has an attention decay curve. The curve describes how much cognitive engagement a reader brings to each successive part of a text. For web editorial, the curve is gradual. For social feeds, it is steep. For AI chat responses, it is among the steepest measured: the first sentence commands disproportionate attention, and engagement drops faster than in almost any other digital format.
BAX calibrates its scoring against decay curves derived from behavioral data across more than 300 AI engines. This means the BAX Index does not treat a mention in position one the same as a mention in position four. It cannot. The attention differential between those two positions is too large to ignore, and ignoring it is precisely what every other measurement tool does.
Why this matters for budget decisions
If you are optimizing your content strategy to increase AI visibility without weighting for position and context, you are optimizing for the wrong thing. You may be increasing your mention count while your actual attention share is flat or falling.
This happens more than the market admits. A brand can appear in more AI responses while appearing lower in those responses, in less decisive contexts, surrounded by stronger competitors. The mention count goes up. The BAX Index goes down. One of those numbers predicts sales outcomes. The other fills slides.
The four-layer architecture
BAX maps the journey from visibility to action across four layers: visibility, attention, sources, and action.
Visibility is where you appear. Attention is how much cognitive weight that appearance carries. Sources is which domains and publishers AI models cite when mentioning your brand, which determines whether your presence is built on durable ground or on content that will be deprioritized in the next model update. Action is whether the attention generated translates into measurable downstream behavior.
No other platform connects these four layers in a single index. Most stop at layer one and call it intelligence.
What the BAX Index is not
It is not a vanity metric. It does not reward volume. A brand with 200 AI mentions and poor position quality will score lower than a brand with 80 mentions consistently in the first or second position, in high-clarity contexts, cited from authoritative sources.
It is not a black box. Every dimension is documented, weighted, and auditable. This matters because the IAB framework now moving toward normative status requires exactly this kind of methodological transparency. A score you cannot explain to a media director is a score you cannot use to justify a budget.
It is not static. The BAX Index tracks over time, which means it reveals whether your AI presence is strengthening or eroding, and why, before the damage shows up in sales data.
The only question that matters
When a potential customer asks an AI what to buy in your category, does your brand earn their attention, or does it merely appear in the answer?
Those are not the same thing. Only one of them is worth paying for.