Your AI Visibility Rank Might Be Misleading

A static AI visibility score captures a moment. Similarweb, a digital intelligence and web analytics platform, published momentum data alongside its 2026 AI Brand Visibility Index that captures a direction. The platform tracked brand-level AI visibility from April last year through January of this year, with the starting index set to 100.

Some brands tripled. Others lost nearly half their visibility. And several brands that hold strong positions in the static rankings are declining at a rate that puts their current standing on borrowed time.

Ulta, a major beauty retailer, reached an index of 319 by January, more than tripling its AI visibility over nine months through sustained multi-month acceleration rather than a single spike. B&H Photo, a photography and electronics retailer, hit 296.9, nearly tripling over the same period. Washington Post reached 271.5, growing 2.7x. Best Buy climbed to 239.7. Southwest Airlines landed at 139.9. NerdWallet, a personal finance comparison platform, grew to 134.4.

Nike dropped to 86.5. Airbnb fell to 82.8. Coach declined to 71.5. AP News fell to 66.1. The Wall Street Journal collapsed to 52.3, losing nearly half its AI visibility in under a year.

A brand looking at the static index and seeing a respectable rank may be looking at a number that has been shrinking every month. A brand with a modest rank that has been climbing steadily is in a structurally stronger position than the static score suggests.

The fastest risers are not the sector leaders

The momentum table reveals something the static rankings hide: the brands gaining AI visibility fastest are often mid-table names, not the ones leading their category.

B&H Photo ranks seventh in Electronics in the static index. Apple dominates the category with a 54.38% mention share. But B&H Photo’s momentum index nearly tripled, which means its rate of growth far outpaces Apple’s. The gap between them in the static rankings is enormous. The gap in trajectory tells a different story. B&H Photo has been expanding its content footprint in exactly the kinds of specialist areas (product comparisons, setup guides, compatibility documentation) that AI retrieval systems favor, and the momentum data reflects that expansion compounding month over month.

Washington Post ranks sixth in News in the static index. Reuters leads with a perfect score. But Washington Post’s momentum index grew 2.7x, the third-fastest rise in the entire dataset. The static rank shows a mid-table publisher. The momentum shows a brand rapidly gaining ground on the leaders.

Ulta in Beauty tells the same story from a different starting position. Ulta already ranks second in the static index behind CeraVe, but its momentum index of 319 suggests the gap between them is narrowing fast. If the trajectory holds, the static rankings at the next measurement point will look different from the ones published in January.

The pattern holds across sectors: the brands building momentum are the ones actively expanding their content into the specific, structured, question-answering formats that AI systems retrieve and cite most frequently. The brands losing momentum tend to be ones relying on existing brand recognition without feeding the retrieval pool with new, extractable content.

Declining brands are losing for identifiable reasons

The brands losing momentum share recognizable patterns, and the News sector makes the mechanism most visible because the variable is easiest to isolate.

The Wall Street Journal’s collapse to 52.3 and AP News’s decline to 66.1 both correlate with access restrictions. The Journal operates a paywall and restricts AI crawler access. AP News has more open access but has been declining in AI visibility momentum nonetheless, suggesting that access alone is not the full explanation but is a significant contributing factor.

Reuters, The Guardian, and The Independent, all with open or partially open access policies, lead the News sector in both static rank and momentum. The Times and the Journal, both paywalled with AI crawler restrictions, trail in both. The pattern is consistent enough to treat access policy as a direct input into AI visibility momentum, not just a theoretical factor.

In Fashion, Nike’s decline to 86.5 comes despite holding a perfect 100.0 score in the static index. Nike still leads the category, but the direction of travel is negative. The brands gaining Fashion momentum (New Balance, Uniqlo, Gap, H&M) are all positioned around utility, value, and structured product data rather than heritage and brand storytelling. AI retrieval systems respond to content they can extract answers from, and utility-focused product content generates more extractable answers than brand narrative does.

Airbnb’s decline to 82.8 in Travel follows a similar logic. The platform grew on community-generated content and SEO-driven discovery. AI retrieval systems favor structured comparison data and planning tools over community reviews, which is why Skyscanner and Travelmath have been gaining AI visibility despite having far less branded search volume.

Coach’s decline to 71.5 in Fashion parallels the broader pattern of heritage and aspirational brands losing ground to accessible, utility-positioned competitors. The AI retrieval system does not browse or aspire. It answers questions, and the brands whose content is built around answering questions gain momentum while the ones built around creating desire lose it.

Rank is lagging. Momentum is leading.

A brand holding a strong AI visibility rank today may be in a structurally weakening position if its momentum index has been declining. The static rank reflects accumulated past performance. The momentum index reflects current trajectory. When the two diverge, the momentum index is the better predictor of where the brand will stand at the next measurement point.

Nike holds a perfect static score and a declining momentum index. That combination means the static rank overstates the brand’s current competitive position. Nike is still the most-mentioned Fashion brand in AI responses, but the rate at which it earns new mentions is falling while competitors are accelerating.

Ulta holds the second position in Beauty with a momentum index of 319. That combination means the static rank understates the brand’s trajectory. Ulta is gaining ground on CeraVe faster than the current static gap would suggest.

The divergence between rank and momentum creates a window. Brands with declining momentum and strong static ranks have time to course-correct before the static rank catches up to the trajectory. Brands with rising momentum and weak static ranks have confirmation that their content and citation-building strategy is working, even if the visible results have not fully materialized in the static index.

The referral traffic plateau and why it changes the measurement question

Alongside the momentum data, Similarweb reported that AI platform visits grew 28.6% between January last year and January of this year (US, desktop and mobile combined). AI referrals to external websites over the same period stayed flat.

More people are using AI platforms more often. Those platforms are not sending proportionally more traffic to external sites. The platforms retain attention, synthesize answers, and influence decisions without routing users to the sources they drew from.

The referral plateau means that measuring AI performance by referral traffic volume produces a misleading picture. A brand could be gaining AI visibility momentum, appearing in more responses, influencing more purchase decisions at the discovery stage, and still see flat or negligible referral traffic from AI platforms. The traffic metric would suggest nothing is happening. The momentum metric would show a brand on the rise.

The useful measurement tracks brand mention share: what percentage of relevant AI responses include the brand, and how that percentage changes over time. Similarweb’s AI Search Intelligence product tracks this at the brand and category level. Other tools are emerging to cover similar ground. The specific platform matters less than the recognition that referral traffic is no longer the right proxy for AI visibility performance.

Building momentum rather than protecting rank

The momentum data from six sectors points at a consistent set of inputs that separate brands gaining momentum from brands losing it.

Accessible content performs better than gated content. Brands with open or partially open access policies gain momentum faster than brands blocking AI crawlers or gating content behind paywalls. The content behind the gate might be superior, but the AI model cannot retrieve content it cannot access.

Structured, specific, extractable content performs better than broad brand narrative. The brands gaining momentum are the ones producing comparison pages, specification tables, detailed how-to guides, and question-answering content that AI retrieval systems can parse and cite. Brand storytelling, aspirational imagery, and marketing copy do not generate the kind of structured answers AI models are looking for.

Third-party citation presence accelerates momentum. The brands appearing most frequently in AI responses are also the ones cited most often across independent publishers, review platforms, and editorial content on other domains. Link building and digital PR produce those third-party citations as a direct output. Every placement on a credible publisher, every editorial mention earned through outreach, every guest post on a domain with editorial standards adds to the citation pool that AI retrieval systems draw from when assembling answers.

Depth beats breadth. The momentum data aligns with a separate finding from the same report: a third of ChatGPT citations come from pages three folders deep in a site’s URL structure. Detailed product pages outperform generic category pages. Long-form comparison guides outperform short overviews. AI retrieval systems reward content that answers a specific question completely rather than content that covers a broad topic superficially.

Link insertions into existing authoritative content offer a way to build momentum on a compressed timeline. If AI models favor pages with established trust and accumulated citations, inserting a brand reference into a page that already carries that authority puts the brand inside the citation pool without waiting for a new page to build momentum from zero.

The momentum window is open but not permanent

The AI Brand Visibility Index captures a market where rankings are still forming. The brands that invested early in structured content and third-party citation presence are compounding now, and the momentum data shows how quickly that compounding accelerates once it starts. Ulta did not triple its AI visibility in a single spike. The growth was sustained, month over month, through consistent expansion of the content and citation footprint that AI retrieval systems reward.

Brands starting from behind are not locked out. The same levers, accessible content, structured pages, third-party editorial presence, depth over breadth, are available to everyone. But the compounding nature of AI visibility means the gap between brands gaining momentum and brands losing it widens with each passing month. The momentum window favors brands that start now over brands that wait for the measurement tools to mature before committing to a strategy.

The static rank tells a brand where it stands. The momentum index tells it where it is heading. When the two numbers disagree, the momentum index is the one to trust.