Dive Brief:
- The Interactive Advertising Bureau is addressing the fragmented state of measurement for artificial intelligence-powered discovery platforms with a new set of standardized guidelines, per a press release from the trade group.
- The IAB identified more than 20 vendors offering AI visibility measurement tools but saw little consistency between their methodologies or results. Its framework centers on four Ps of visibility — presence, prominence, portrayal and persuasion — as well as two tiers of measurement based on data quality and rigor.
- The guidelines arrive as core channels like search are further reshaped around AI, putting more pressure on marketers and publishers to optimize for Generative Engine Optimization, Answer Engine Optimization and prompt engineering.
Dive Insight:
The IAB pointed to the widespread popularity of OpenAI’s ChatGPT and Google’s AI Overviews as indicative of how the shift to AI-powered discovery has achieved a scale that requires a more standardized measurement framework for marketers and publishers alike. The stakes are high for brands that fail to adapt for AI-powered discovery, as McKinsey estimates that laggards could see traffic declines as high as 50% from traditional search. Only 16% of brands currently track their AI search performance in a systematic way, McKinsey found in CMO surveys.
The IAB’s guidelines are not designed to rate the growing crop of AI measurement providers, an area where the group sees little consistency. Instead, the IAB wants to introduce shared vocabulary, quality criteria and disclosure requirements that enable marketers to make apples-to-apples comparisons and measurement vendors to “differentiate on rigor rather than claims,” the guidelines state. The IAB is also initially focused on organic AI visibility measurement, not paid advertising or commerce attribution. Paid measurement was described as an “adjacent priority” since “organic and paid visibility increasingly appear on the same response surface,” the IAB explained.
Getting the industry to agree on common vernacular and best practices could be difficult, especially given that AI-powered discovery is still in its early days and rapidly evolving. The IAB’s argument is that advertisers are ready to spend more on these channels assuming they have a solid basis to evaluate the efficacy of their buys.
The IAB’s framework centers on four “P” principles for brands that sit in a hierarchy: at the top layer is presence, embodied in mention rate, citation rate, share of voice and “visibility momentum.” Below that is prominence, which is largely based on placement and ranking order, and then portrayal, an evaluation of sentiment, framing and hallucination and factual inaccuracy rates. Last is persuasion, a gauge of recommendation strength and post-citation click-through rate. The same Ps apply to publishers, albeit with different metrics attached to each principle.
In addition, the IAB is proposing two different measurement tiers for AI-powered discovery. The first, directional measurement, is focused on early signal detection, internal briefings and competitive awareness, but does not bring enough rigor to influence ad spending or other strategic decisions. The second, decision-grade measurement, lives up to its name by ingesting sample size, query volume, prompt type coverage, testing cadence, reproducibility and data validation, among other factors. That latter could guide strategy on everything from agency performance reviews to budget allocations, the IAB said.
AI-powered discovery presents a distinct set of challenges for marketers to tackle compared to traditional search, which had a simpler journey of user query to results page to click. Generative AI platforms instead work by scraping and condensing an array of existing sources of information into a single answer, and still can produce inaccurate or misleading results. Asking the same question twice may show two different responses, amplifying issues around consistency. Similarly, over 40% of brand citations that appear in organic search results do not appear in AI overviews for the same query, per recent LQ research.
The IAB’s working group that collaborated on the AI Visibility Measurement Framework is representative of stakeholders from across marketing and advertising. Members include measurement experts from Walmart, Acxiom, Microsoft, WPP Media, EMaketer and Tinuiti.
