Reporting practices in search marketing have not caught up with how discovery actually works. Monthly reports still lead with keyword position tables, while a growing share of commercial research happens inside AI assistants where position does not exist. Citation is binary. A brand appears in an answer, or it does not. There is no third place.
Royston G King argues that reporting practices have not caught up with the mechanism. He studied business at the University of Southern California and has since been accepted into Columbia University. He founded Master Scaling in 2018 and Quantum Scaling Partners as its selective arm, and has spent close to a decade across search visibility, media placement, and reputation work. A Forbes 30 Under 30 Monaco honouree whose commentary has appeared in Entrepreneur and Inc., he now pushes clients toward a different set of numbers entirely.
Building useful measurement means replacing familiar metrics with ones that describe the new mechanism. Several are worth adopting.
Prompt-level citation rate, segmented by engine. Define a set of prompts that reflect how buyers actually ask about a category. Run them at intervals across each major assistant. Report the percentage in which the brand appears. This is the closest equivalent to a ranking report and the most directly useful number available. Segmentation by engine is essential because source pools differ dramatically between platforms, and an aggregate figure conceals which channel is working.
Mention rate alongside citation rate. Being named in the body of an answer without a linked citation still influences purchasing behaviour, and analysis of large mention datasets has found that appearing in the response text corresponds with substantially higher citation rates than not appearing. Tracking both separates two distinct outcomes that a single metric would blur.
Position and framing within the answer. Being named first as a leading option differs materially from appearing in a closing list of alternatives. So does the descriptive language attached. Capturing whether a brand is described accurately, favourably, or dismissively provides information no numeric metric does.
Which specific source URL was used. This is the most operationally valuable metric available and the most frequently omitted. Knowing that a particular video, a specific discussion thread, or a trade publication article is the asset driving citation tells a team exactly where to invest next. Without it, budget allocation is guesswork.
Branded web mention volume across the open web. Because mentions correlate with AI visibility far more strongly than backlinks do, mention volume functions as a leading
indicator. It moves before citation rate moves, which makes it useful for assessing whether current activity is working before results appear in answers.
Branded search volume. A slower second-order indicator, but a meaningful one. Rising branded search suggests the category association is strengthening.
Impressions rather than clicks in conventional search. Analysis has found that the correlation between conventional search performance and AI citation strengthens considerably when measured against impressions rather than clicks. The reason is mechanical. AI systems draw from the pool of pages considered relevant to a query, not from the pages that win the click. Impressions describe that pool. Clicks describe something else.
Two methodological points determine whether any of this produces reliable information.
The first is repetition. The same query can return different sources across repeated runs, with variation measured as high as half the time in some cross-region studies. A single check of a single prompt is noise. Meaningful measurement requires running a defined prompt set repeatedly and reporting distributions rather than snapshots.
The second is expectation setting around traffic volume. AI-referred traffic remains small in absolute terms for most organisations, and teams accustomed to conventional search volumes often read the numbers as failure. Conversion behaviour tells a different story. Ahrefs reported that AI search visitors accounted for a disproportionate share of signups relative to their share of traffic, a substantial conversion advantage over conventional organic. The visitors are fewer and considerably further along in their decision process.
Reporting that leads with position tables describes a mechanism that is steadily becoming less relevant to how buyers find suppliers. Reporting built on citation rate, mention rate, source attribution, and mention volume describes what is actually happening.
Reporting that leads with position tables describes a mechanism steadily becoming less relevant to how buyers find suppliers. The measurement framework King uses at Quantum Scaling Partners is built on citation rate, mention rate, source attribution, and mention volume instead.
Connect With Royston G. King
To learn more about Royston G. King and follow his latest work, visit his official website, connect with him on Instagram and LinkedIn, or watch his latest content on YouTube.
