Most B2B SaaS teams are still Googling their own brand name in ChatGPT and calling it measurement. IAB just told them to stop. Fewer than 50 prompts "cannot meaningfully characterize a category." That's the IAB's line in the sand, published August 3rd in its guidance document, "Measuring Visibility in the AI Era." If your team's AI visibility check consists of typing your brand into ChatGPT a dozen times and eyeballing the results, you're operating below what the IAB classifies as "exploratory" (their polite term for useless). The document isn't a standard. Caroline Giegerich, IAB's VP of AI, made this clear to AdExchanger: a standard requires stability, and marketers are in a "mass transition space." It's not even officially a framework; IAB changed the name at the last minute "so we don't have 100 frameworks," Giegerich said. Fair enough.

The 4P's: A Hierarchy, Not a Checklist

What the IAB published is a metric hierarchy called the "4P's of AI Visibility": presence, prominence, portrayal, and persuasion. They build on each other, and the order matters. Presence is the baseline. Does your brand show up at all in AI-generated answers? Measured as mention rate and citation share across platforms. Simple question, but most teams can't answer it confidently across Google AI Overviews, ChatGPT, Perplexity, and Gemini simultaneously. Prominence asks where you appear in that answer. Are you the first recommendation, or buried among competitors? Share of voice within the response is the metric here. Portrayal measures sentiment and accuracy. Giegerich split accuracy into two buckets: hallucinations (declining as models improve) and factual inaccuracies from outdated or misrepresentative training data. The latter, she said, "would be the thing that would keep me up at night." A hallucination is obviously wrong; a factual inaccuracy looks plausible and quietly erodes trust. Persuasion tracks whether citations drive action (clicks, traffic, conversions). This is the hardest to measure in a zero-click world, and the IAB acknowledges it. Post-citation click-through rate is the suggested KPI, but as AI Overviews improve at encapsulating answers without requiring a click, persuasion metrics need to evolve beyond traffic.

Directional vs. Decision-Grade: The Budget Question

The sharpest operational distinction in the document is between "directional" and "decision-grade" measurement. Directional data comes from informal spot checks. It tells you something. Decision-grade data comes from larger, cross-platform query samples with enough volume and variety to support budget allocation. The IAB's guidance is clear: only decision-grade measurement should inform spending decisions. That means building a repeatable query library, segmenting by platform, and running enough prompts to generate statistically meaningful patterns. Fewer than 50 queries per category is exploratory. Most teams aren't close to this bar. Here's the uncomfortable part for ops teams: AI isn't deterministic. The same query can return different answers on different days, platforms, or sessions. You can't just run your query set once and call it a baseline. You need sampling cadence. Weekly? Biweekly? The IAB doesn't prescribe a frequency, so your team must figure it out through experimentation.

What This Means for B2B SaaS GTM

If you're running demand gen for a B2B SaaS company, the shift here is structural. Classic search measurement provided rankings, click-through rates, and traffic. AI visibility measurement provides mention rate, citation share, share of voice, and sentiment. Different data, different dashboards, different leading indicators for pipeline. Three things worth building now: The trade-off you're accepting: none of this maps cleanly to last-click attribution. Zero-click behavior is increasing, and traffic from AI citations may remain thin even as your brand's presence in AI answers grows. Measuring AI visibility as a leading indicator of qualified pipeline influence (rather than a traffic source) is the mental model shift most teams haven't made yet.

The Vendor Landscape Won't Consolidate Soon

You might expect IAB guidance to thin the herd of AEO and GEO tools. Giegerich doesn't see it. The industry is still, as she put it, fighting "to get to the cheese at the end of the tunnel." The guidance provides a shared vocabulary but doesn't pick winners. This brings us back to the 50-query threshold. The IAB drew a line between measurement that means something and measurement that feels productive. Most teams are still on the wrong side of that line, running spot checks and treating the output as signal. The 4P's give you a structure. Whether your team builds the sampling infrastructure to make it decision-grade is a different question.