Brand Services/AI References, Brand Visibility, Category Positioning, ToB Branding, GEO
This set of figures comes from "Crackle 100: 100 Most-Cited B2B Tech Brands in AI Search 2026" released by Crackle PR, updated on November 13, 2026. The sample is 4 AI search engines × 200 purchase intention questions × 3 runs. The statistics show the citation of B2B technology brands in supplier selection questions, and the public caliber is CC BY 4.0. It measures the reference position in the AI answer, does not represent the overall brand awareness, nor does it correspond to revenue or market share.
Top 10: OpenAI(100.0), Anthropic(96.4,+4), NVIDIA(94.1,+1), Microsoft(92.7,− 2), Google Cloud(90.8), AWS(89.5,− 2), Databricks(84.6,+3), and Snowflake(82.4,− 1). The top 10 together took 88.5 percent of all top-3 citations.
The two most concentrated are AI infrastructure (OpenAI, Anthropic, NVIDIA) and cloud (Microsoft, Google Cloud, AWS). The rest of the category has a longer tail, and the first place does not form the same degree of monopoly.
The meaning of the long tail brand is: to compete for the top three positions in the mature category, the difficulty is concentrated in the head; In the subdivision category that is taking shape, it is easier for the brand that makes the category word clear first to get the position.88.5 percent is the result of concentration, not a restriction on new entrants.
In plain English,narrative momentumIt is the strength of a brand being written into a category story: when others describe this category, will they write it as a representative case, standard scheme or typical player. Falling into the AI answer, the reference position corresponds to this accumulation-the position repeatedly confirmed in the category story.
This layer of position is saved: the third party repeatedly writes and the reviewable facts are repeatedly quoted before the category position settles down; One activity or one round of delivery cannot exchange a stable reference position.
Do another layer of contrast: brand awareness answer "do not know", brand value answer "value", narrative momentum answer "in which category of story 」. The three can exist at the same time, or they can be out of sync-well-known brands may not be in the new category narrative.
Illustration of the sequential relationship between narrative accumulation and reference position
In the 2026 edition of the list, the brands with the highest increases each have one thing in common: they got category-defining reports in the scanning window. Cursor up 38, Perplexity 22,Wiz 18,ElevenLabs 16,Supabase 15,Chainguard 12,Vercel 11. They correspond to the emerging category stories of AI programming IDE, AI search, cloud security integration, generated voice, edge developer platform, container security, and edge deployment.
Falling side: Loom− 4, Okta− 4, HashiCorp − 4, Zoom− 5. They are all leaders in mature categories, and no new category narratives enter in the scanning window.
Let's look at one thing in common: rising brands get the definition of new categories, not the packaging of old categories. The problem with falling brands is not the quantity of content, but that there is no new category relationship to enter and accumulation does not continue to stack up.
The conclusions of the report's authors are straightforward:AI citation share is a lagging indicator of earned media momentum, not an indicator of ARR.In other words, the reference position reflects where the external narrative accumulation has gone, not the current income scale.
Reference position rise and fall brand comparison
There is a common misjudgment here: attributing the reference back to insufficient placement. The reference position is a lagging indicator, first look back at whether the external narrative accumulation has changed, and then talk about the rhythm of delivery. Before the category position is not moved, the launch will bring more exposure and will not automatically turn into category roles in other populations.
High narrative momentum does not mean high income, it only describes the lag relationship of "narrative first, citation later", and does not promise cause and effect. Changes in third-party reporting and indexing are not within the control of the service provider.
The actual measurement of 5W Research on November 20, 2026 also gives circumstantial evidence: being mentioned and cited are two independent results, and brands can be named without being cited. Narrative momentum acts in the latter step-first written into the category story, and then into the material that can be cited.
The reference is compared with the two calibers cited.
you can ask yourself four sentences: have the category words been decided? are there any factual materials that can be rechecked? which category does the outside put us in? when was the last time you saw the narrative position?
斯托亚德洞察
First of all, the "what do you do" this sentence clear, so that others can say a word to understand; then the product parameters, customer cases, these can withstand the check of the content, organized into AI can read the page; then every once in a while, together to see what kind of articles outside you into, AI answer to mention you.