New research examines how Profound’s 2B+ corpus of real-user AI conversations feeds Prompt Volumes, Answer Engine Insights and the Profound Index.
GLASGOW, UNITED KINGDOM, September 16, 2026 /EINPresswire.com/ — QueryBurst has published the first in a planned series of reviews examining how AI visibility platforms collect, model and interpret the data behind their scores, benchmarks and market-level claims.
The publication coincided with news that New York-based AI marketing platform Profound has raised $180 million in a Series D at a $1.8 billion valuation.
QueryBurst’s research was published independently on September 15 following work that began before Profound’s funding announcement. The timing highlights a broader issue as investment and automation accelerate across the category: marketers need to understand where underlying data comes from, how it is transformed and what resulting metrics actually measure.
“As more marketing decisions are handed to AI, the quality and provenance of the data underneath those systems matters more, not less,” said QueryBurst founder David McSweeney.
“Automation doesn’t fix uncertainty in the measurement underneath it. If your starting point is incomplete, poorly understood, or measuring something narrower than you think, every decision downstream can inherit that problem.”
The first review examines Profound’s public disclosures around its billion-scale real-user AI conversation dataset, Prompt Volumes, Answer Engine Insights and Profound Index.
The research draws exclusively from Profound’s own first-party materials, including product pages, Help Center documentation, engineering posts, research, press releases and public statements.
Profound currently describes its platform as being built on more than 2 billion real-user prompts. Earlier Index materials described the benchmark as being powered by more than 1.9 billion real-user conversations.
According to QueryBurst’s review, Profound’s methodology spans three layers.
Prompt Volumes uses observed AI conversations from double-opt-in consumer panels as source data, then applies statistical modelling to produce population-level volume estimates.
Answer Engine Insights sends prompts to consumer-facing AI systems and captures the resulting responses.
The Profound Index uses real-user conversation data to retrieve, cluster and select representative prompts, while final brand rankings are calculated from fresh AI responses to the resulting standardized prompt set.
QueryBurst concludes that the billion-scale corpus is not itself the direct sample from which individual Index rankings are calculated. Instead, it informs construction of the benchmark prompt set used for subsequent measurement.
“The point was to separate what is directly observed from what is modelled, extrapolated or constructed as a benchmark,” McSweeney said. “Profound has published considerably more methodology than many platforms in this category, but some important questions still cannot be answered from the public record.”
The review identifies unresolved areas including the counting unit behind headline corpus figures, canonical prompt counts, effective sample size behind specific rankings, whether prompt clusters are weighted by observed demand, and statistical uncertainty attached to Prompt Volume estimates or leaderboard positions.
It also examines Profound’s Black Friday Index, where real-user Prompt Volumes data identifies high-volume retail categories before 50 unbranded prompts per category are synthetically generated for the benchmark.
The research follows 31 methodology questions published in July 2026 covering prompt representativeness, weighting, uncertainty, personalization, conversation state, external validation and falsifiability. Several smaller AI visibility platforms responded. The major named platforms did not.
McSweeney said the Profound review is the first in a broader series examining how AI visibility platforms define and measure the category.
“The wider issue is that AI visibility is increasingly represented through scores, rankings and market-level claims. Those numbers may be useful, but they need to be understood in terms of what was actually sampled, modelled and measured.”
QueryBurst’s position is that AI visibility should not be treated as a single context-free score.
Rather than tracking fixed prompt panels, QueryBurst models multi-turn customer journeys through AI search to investigate where brands enter, influence and leave the conversation — and why.
“We believe improved AI visibility is an outcome, not a score,” McSweeney said. “A market is made up of people moving through different conversational states, not interchangeable prompts.”
The full research is available at:
https://queryburst.com/research/profound-methodology/
About QueryBurst
QueryBurst is a workspace for investigating and improving multi-turn customer journeys through AI search. It helps marketers understand where a brand enters, influences and leaves the conversation — and why — then act on the conditions shaping those outcomes.
QueryBurst does not offer prompt tracking or report AI visibility scores. It approaches AI visibility through people, conversations and evidence rather than prompt panels.
David McSweeney
QueryBurst
hello@queryburst.com
Legal Disclaimer:
EIN Presswire provides this news content “as is” without warranty of any kind. We do not accept any responsibility or liability
for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this
article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
Media gallery
