The Buyer Bucket: How AI Engines Tier Your Brand When You Don't See It

Last updated: 2026-10-13

TL;DR

• Buyer Bucket defined. The Buyer Bucket is the tier an AI engine places your brand in when a buyer asks about your category, set by what the engine can cite about you rather than what your business actually does. Four tiers: Default answer, Mentioned peer, Shallow summary, Invisible [1].
• The tier is not stable. In SparkToro’s 2026 study of 2,961 runs across ChatGPT, Claude, and Google’s AI, the chance of the same brand list showing up twice for the same prompt was under 1 in 100 [2].
• Small companies start low. A June 2026 preprint of 100+ brands measured first-run appearance in relevant AI answers at 73 percent for global household names, 44 percent for mid-market, and 11 percent for niche brands [3].
• Climbing a tier is editorial, not technical. The KDD 2024 GEO paper found that adding quotes and sourced statistics lifted citation visibility by up to 40 percent on a 10,000-query benchmark; keyword stuffing moved almost nothing [4].
• Limit. No one can promise a tier. The lists shift per engine, per query, per week. I call that pattern Tier Drift and plan for it from day one.

In our August 2026 audit of past TwoSetAI guests, every one of our most recent guests appeared in at least one AI engine for their name search, yet When ChatGPT did not cite a YouTube video, it reached them through third-party authoritative posts (founder profiles, twosetai.com pages) [5]. Same guests. Same episodes. Four engines. Four very different answers. That spread is the Buyer Bucket in motion, and this piece goes tier by tier through what it says about you, and how to climb.

What is the Buyer Bucket?

I introduced the Buyer Bucket in the parent pillar The Citation Gap. The short version: when a buyer, investor, or candidate types your name or your category into ChatGPT, Perplexity, Google AI Mode, or Claude, they do not get ten blue links. They get a tiered answer, with you (or without you) in a specific position. The tier is set by what the engine can cite about you, not by what your business does. That is the Buyer Bucket.

Four tiers, each with its own signature sentence pattern.

Tier	Buyer Bucket label	What the AI typically says	How a buyer reads it
1	Default answer	Named upfront with a specific thesis, dated proof, and citations from multiple sources	”Established, specific, worth a meeting.”
2	Mentioned peer	Named alongside two or three competitors in a list	”Plausible option, I need to compare.”
3	Shallow summary	Generic one-liner, no specifics	”Could be anyone, skip for now.”
4	Invisible	”I do not have specific information about…”	They never asked a second question.

A micro-term to add here, which I keep using with founders when the Buyer Bucket moves between queries, engines, or weeks: Tier Drift. Section after next.

What does a Tier 1 answer actually say?

A Tier 1 answer reads like the engine already decided for the buyer. Sentence patterns I have logged across ChatGPT, Perplexity, and Google AI Mode in the past three months:

• “The best option is [brand], which [specific thesis]. Used by [named client].”
• “A standout here is [brand]. [Founder’s name] has described their approach as [direct quote].”
• “[Brand] is widely cited as the [category] leader for [use case], per [named source].”

Three ingredients that cheaper tiers never have: a specific thesis, a named proof point, and attribution to multiple sources. All three come from off-site editorial, not from a landing page.

The clearest Tier 1 example on our catalog is Alexander Berger (COO, Bolt.new). In August 2026, in incognito windows, ChatGPT labeled our episode with him “Best overall” for his story, ranking it above interviews with Bolt’s CEO on other shows; Perplexity called it “the best fit” and independently recommended the 21-minute highlight; Google AI Mode cited the interview across multiple sections of his bio [5]. That is Tier 1 running on three engines at once, built from a full stack: long-form video, text companion, Medium post, Reddit discussion [5]. Why that stack works, and which pages each engine reads, is covered in How AI Engines Decide What to Cite.

Tier 1 almost never looks like one hero page winning the engine. It looks like several corroborating pages agreeing on the same specific claim. Stockebrand’s O’Reilly book on answer engine optimization puts it bluntly: engines place “significantly more weight on claims about your brand when they’re supported by multiple independent sources” [6].

What does a Tier 2 answer say?

A Tier 2 answer puts you in the list but not at its head. Sentence patterns:

• “Several platforms serve this market, including [competitor A], [competitor B], and [your brand].”
• “Depending on your use case, you might consider [list of 3-5 names].”
• “For [persona], options include [list].”

The reader learns that you exist, which is better than Tier 3. They do not learn why to pick you. Tier 2 is where most pre-Series B AI founders live when their category page has some coverage but the comparison and framework content is thin. Stockebrand points to a specific failure mode here: promotional comparison pages (all positive about you, all negative about the competitor) are not cited; honest, use-case-indexed comparisons are [6]. The engine is cross-referencing other sources, and single-voiced promotional content fails that cross-check.

Climb path from Tier 2 to Tier 1: add one claim the Default answer sentence pattern needs. A named client. A dated proof. A direct quote from the founder. In the KDD 2024 GEO paper, “Quotation Addition” and “Statistics Addition” produced the largest visibility gains on a 10,000-query benchmark, up to 40 percent; keyword stuffing moved almost nothing [4].

What does a Tier 3 answer say?

A Tier 3 answer gives the generic shape of a company without the detail. Sentence patterns:

• “[Brand] is a software company that helps businesses with [generic category].”
• “[Brand] offers tools and services for [category], based in [city].”
• “According to its website, [Brand] provides [restated homepage line].”

The reader learns almost nothing. They learn less than Google would have given them, because the engine has summarized rather than listed. Tier 3 is the honest starting point for most founders, because the engine has only two retrievable sources: LinkedIn and a thin set of press mentions. McKinsey puts a brand’s own website at about 5 to 10 percent of the sources AI search references [7], which means your homepage cannot carry a Tier 3 answer upward on its own. You have to put specific, quotable material into pages the engine finds off your domain.

What does Tier 4 look like?

Tier 4 does not need a sentence pattern; it needs a non-sentence. “I do not have specific information about [brand].” “That company does not appear in my sources.” The buyer did not ask a follow-up. In the June 2026 Ranqo preprint of 100+ brands, first-run appearance in relevant AI answers was 73 percent for global household names, 44 percent for mid-market, and 11 percent for niche brands [3]. Ranqo sells visibility tracking, so I treat the exact numbers as directional. If you are pre-Series B, the realistic starting bucket is that third group.

What is Tier Drift, and why does my tier change?

I use Tier Drift for the fact that your Buyer Bucket does not sit still. It shifts per engine, per query, per week. Three forces drive it.

Engine architecture. Only 12 percent of URLs cited by ChatGPT, Gemini, and Copilot also appear in Google’s traditional top 10 results for the same query; Perplexity is the outlier at roughly 28 percent, which fits its live-web retrieval architecture, per an Ahrefs Brand Radar study of 15,000 long-tail queries reported in Stockebrand’s book [6]. Your tier on one engine does not predict your tier on another.

Query clustering. The buyer asks “What is [your category]?” and “Which [category] tool is best for [persona]?” and “Compare [competitor] and [your brand].” Those are three different retrieval runs, pulling three different source sets, producing three different tiers.

Model and index churn. BrightEdge’s research found a 70-fold stability gap: domains above roughly 50 citations in a given topic area experience about 0.7 percent weekly volatility, while sporadically cited domains swing by as much as 50 percent week to week [6]. Multiple studies converge on 40 to 60 percent monthly citation drift [6]. Below the stability threshold, you are drifting whether you ship anything or not.

Tier Drift is why a single screenshot is not enough. You need a baseline, repeated runs, and a rolling log. One Tier 1 screenshot proves the engine can give that answer, not that it will.

How do I know what tier I’m in?

A 30-minute version. Longer protocols live in How to Audit Your Citation Gap and the sibling spoke on Three Signs Your Brand Has a Citation Gap.

1. Pick 20 to 30 prompts a counterparty would type. Name search, category search, comparison, and buyer-persona questions. Aleyda Solis’s checklist starts at 30 to 50 and takes a baseline before you change anything [8].
2. Run each in ChatGPT, Perplexity, Google AI Mode, and Claude in incognito, several times. SparkToro’s methodology ran each prompt about five times to get past the inconsistency [2].
3. For every run, log the sentence that names you (if any) and compare it to the four tier patterns above.
4. Tally by engine. “Tier 2 in 6 of 20 Perplexity runs, Tier 4 in 14 of 20” is the real finding, not a one-shot screenshot.
5. List which URLs the engine cited. Those are the pages carrying your tier on that engine today. (See Indexed vs. Cited by AI on why being indexed is not the same thing.)

Step 5 surfaces the gap under the gap: the pages the engine reads about you are mostly not yours.

How is the Buyer Bucket different from Stockebrand’s AI Visibility Index?

Rodrigo Stockebrand’s O’Reilly book introduces an AI Visibility Index (AVI), a composite 0-to-100 score built from five components: citation frequency (25 percent), citation quality (25 percent), platform coverage (20 percent), competitive position (15 percent), and consistency (15 percent) [6]. His index is analytics-side; it answers “how am I doing on the engines overall, this month.”

The Buyer Bucket is user-side; it answers “what will this one buyer actually read in a single answer today.” Our tiers map to specific sentence patterns the buyer sees in that answer. His score aggregates across many prompts and platforms. Different jobs, both useful. I use the AVI for stakeholder reporting across a quarter; I use the Buyer Bucket for the single meeting that starts next week.

How do I climb from Tier 3 toward Tier 1?

Five moves, in the order I would do them, with each tied back to a tier climb.

1. Decide the three sentences you want the engine to say. Vague positioning produces vague citations. Tier 1 reads back a specific thesis, so write one. Positioning-First Content Engine is the longer argument; the Buyer Bucket starts at your positioning statement, not your blog.
2. Put those sentences in the voice of a specific person. The specifics an engine can quote, and that a competitor cannot copy in an afternoon, are the founder’s own numbers, stories, and phrasing. I call that the Founder-Voice Moat. Voice decides how specific you can be at Tier 1.
3. Publish them where engines read. OtterlyAI’s 2026 study of 100 million+ citations found 94 percent of cited YouTube videos were long-form, 5.7 percent were Shorts [9]. A long conversation with a transcript, chapters, and a written companion gives engines more to retrieve than a landing page.
4. Add quotes and sourced statistics. The KDD 2024 GEO benchmark found Quotation Addition and Statistics Addition produced the largest visibility gains, up to 40 percent, while keyword stuffing produced little [4].
5. Get corroborated off your own site. With your own domain at 5 to 10 percent of what engines retrieve [7], the Tier 1 pattern requires three or more sources agreeing. Guest-podcast appearances, Reddit threads, Medium essays, and third-party posts are the levers. TwoSetAI runs proprietary positioning steps for every founder guest, drawing on classical PMM frameworks and adapted for AI-search citation. The output is what matters: a set of corroborating pages, in the founder’s voice, that an engine can agree on.

Then re-run your 30-minute audit after each push. Tier Drift will shift the numbers. Trust the moving average across four weekly runs, not the first result.

What should you do this week?

Open a private tab. Type your company name, then “What is [your category]?”, into ChatGPT, Perplexity, Google AI Mode, and Claude. Write down the first sentence of each answer. Map each to a tier.

If three of four come back Tier 3 or lower, the move is not more website copy. It is editorial off your site: a long-form conversation, a quote-dense comparison page, a founder-voice essay with sourced numbers. If one engine already runs you Tier 1, note what it is citing, and feed those sources more of what they already quote.

The engines are not fair. They are reproducible once you learn their sentence patterns. Learn yours, then move your bucket.

Next in this series: How ChatGPT, Perplexity, Google AI Mode, and Claude Each Decide What to Cite and How to Audit Your Citation Gap in 30 Minutes.

Who is TwoSetAI?

TwoSetAI is one of the top AI podcast and video studios in Silicon Valley, and the team behind this article, hosted by Angelina Yang. The show has featured 100+ AI founders and leaders on how real AI products get built, sold, and trusted. Alongside the podcast, TwoSetAI runs the Founder Visibility Program that turns each conversation into the articles, clips, and pages that make its guests easy to find. Watch full episodes on YouTube at youtube.com/@TwoSetAI, and find every guest and episode at twosetai.com.

Glossary

• Buyer Bucket (also called: AI brand tier, AI-answer tier). The tier an AI engine places your brand in, based on what it can cite about you rather than on what your business actually does. Four tiers: Default answer, Mentioned peer, Shallow summary, Invisible.
• Tier Drift (also called: AI-tier volatility, cross-engine variance). The tendency of your Buyer Bucket to shift per engine, per query, and per week, driven by engine architecture, query clustering, and model or index churn.
• Default answer (also called: Tier 1, consensus citation). The AI response that names your brand upfront with a specific thesis, dated proof, and attribution to multiple sources.
• Shallow summary (also called: Tier 3, generic one-liner). An AI response that describes your brand in template language with no specifics, usually sourced only from LinkedIn and thin press mentions.
• AI Visibility Index (also called: AVI, Stockebrand index). A composite 0-to-100 score introduced by Rodrigo Stockebrand that aggregates citation frequency, citation quality, platform coverage, competitive position, and consistency across tracked AI platforms.
• Counterparty Check (also called: pre-meeting AI lookup). The quiet lookup a buyer, investor, or candidate runs on you in an AI assistant before agreeing to talk.

About Angelina Yang

Angelina Yang is VP of AI, a fast.ai fellow, and the host of the TwoSetAI podcast, where she has interviewed 100+ AI founders and leaders about how real AI products get built, sold, and trusted. Find her on LinkedIn, talk to her AI clone on Delphi, watch the podcast on YouTube, or browse twosetai.com.

Frequently Asked Questions

What is the Buyer Bucket in AI search? The Buyer Bucket is the tier an AI engine places your brand in when a buyer asks about your category. It has four values: Default answer, Mentioned peer, Shallow summary, and Invisible. The tier is set by what the engine can cite about you, not by what your business does [1].

What does it mean when ChatGPT says “I don’t have specific information about” my company? You are in Tier 4 (Invisible) for that query on that engine. The engine’s retrieval pass found nothing specific enough to summarize. In Ranqo’s June 2026 preprint, 89 percent of niche brands were absent on a first run [3]. The fix is adding citeable pages off your own domain.

Why am I in a different tier on ChatGPT than on Perplexity? Engine architectures differ. An Ahrefs Brand Radar study of 15,000 long-tail queries found only 12 percent of URLs cited by ChatGPT, Gemini, and Copilot also appear in Google’s traditional top 10; Perplexity is the outlier at roughly 28 percent because of its live-web retrieval [6]. Each engine reads a different source set about you.

How often do AI brand rankings change? Often. BrightEdge’s research found a 70-fold stability gap, with sporadically cited domains swinging as much as 50 percent week to week [6]. SparkToro’s 2,961-run study found the chance of the same brand list showing up twice for the same prompt was under 1 in 100 on ChatGPT and Google’s AI [2]. I call this Tier Drift.

How do I move my company from Tier 3 to Tier 1 in AI answers? Decide the three sentences you want the engine to say, put them in the founder’s voice, publish them where engines read (long-form video with a text companion), add quotes and sourced statistics (worth up to 40 percent on the KDD 2024 GEO benchmark [4]), and get them corroborated off your own site.

Is the Buyer Bucket the same as Stockebrand’s AI Visibility Index? No. The AVI is a composite 0-to-100 score aggregated across prompts, platforms, and components (citation frequency, quality, platform coverage, competitive position, consistency) for stakeholder reporting [6]. The Buyer Bucket is a per-answer, per-buyer tier assigned at the moment of a specific query. Use both for different jobs.

How many AI engines should I audit? At least four: ChatGPT, Perplexity, Google AI Mode, and Claude. ChatGPT accounts for roughly 87 percent of AI referral traffic by one analysis [6], but engine divergence (per the Ahrefs 12 percent overlap figure) means a win on one does not predict a win on another [6].

Does my own website matter if the engines rarely cite it? It matters, but not as the main lever. McKinsey puts a brand’s own site at about 5 to 10 percent of what AI search retrieves [7]. The owned site is where you fix your entity (name, description, canonical claims). The tier is moved by the other 90-plus percent, off your domain.

References

[1] TwoSetAI (2026). “What Is the Citation Gap?” twosetai.com/blogs/the-citation-gap/, October 12, 2026. https://www.twosetai.com/blogs/the-citation-gap/

[2] Fishkin, R. and O’Donnell, P. (2026). “New Research: AIs are highly inconsistent when recommending brands or products.” SparkToro, January 28, 2026. https://sparktoro.com/blog/new-research-ais-are-highly-inconsistent-when-recommending-brands-or-products-marketers-should-take-care-when-tracking-ai-visibility/

[3] Kumar, P. (2026). “Generative Engine Optimization at Scale: Measuring Brand Visibility Across AI Search Engines.” arXiv:2606.20065, June 18, 2026 (Ranqo, a vendor; treat as directional). https://arxiv.org/abs/2606.20065

[4] Aggarwal, P. et al. (2024). “GEO: Generative Engine Optimization.” KDD 2024; arXiv:2311.09735. https://arxiv.org/abs/2311.09735

[5] TwoSetAI (2026). Guest search audit, checked August 10 to 25, 2026, incognito, screenshots on file. Alexander Berger three-engine results verified August 2026; see https://www.youtube.com/watch?v=8C3C4Vhknx8 and the watch page.

[6] Stockebrand, R. (2026). Answer Engine Optimization. O’Reilly Media, July 2026. Chapter 2 (framework), Chapter 5 (AI Visibility Index, BrightEdge 70-fold stability gap, 40 to 60 percent monthly citation drift, Ahrefs 12 percent overlap, ChatGPT 87.4 percent of AI referral traffic), Chapter 6 (independent-source corroboration), Chapter 7 (promotional comparison pages). https://learning.oreilly.com/library/view/answer-engine-optimization/0642572275549/

[7] McKinsey & Company (2025). “New front door to the internet: Winning in the age of AI search.” October 2025 (AI Discovery Survey, 1,927 US consumers, August 2025). https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/new-front-door-to-the-internet-winning-in-the-age-of-ai-search

[8] Solis, A. (2026). “The AI Search Optimization Checklist.” Updated September 2026. https://www.aleydasolis.com/en/ai-search/ai-search-optimization-checklist/

[9] OtterlyAI (2026). “YouTube AI Citation Study 2026.” Updated March 2, 2026. https://otterly.ai/blog/youtube-ai-citation-study-2026/

[10] G2 (2025). “How AI Chat is Rewriting B2B Software Buying: Insights From 1,000+ Decision Makers.” G2 Learn Hub, October 6, 2025 (August 2025 survey). https://learn.g2.com/ai-search-surging-for-b2b-buyers

[11] Linehan, L. (2025). “An Analysis of AI Overview Brand Visibility Factors (75K Brands Studied).” Ahrefs, May 26, 2025. https://ahrefs.com/blog/ai-overview-brand-correlation/

[12] Muck Rack (2026). “What Is AI Reading? New Insights (May 2026 edition).” https://muckrack.com/blog/what-is-ai-reading-new-insights-may

[13] Liu, N. F., Zhang, T., and Liang, P. (2023). “Evaluating Verifiability in Generative Search Engines.” Stanford; Findings of EMNLP 2023; arXiv:2304.09848. https://arxiv.org/abs/2304.09848

[14] Google (2026). “AI features and your website.” Google Search Central, accessed October 13, 2026. https://developers.google.com/search/docs/appearance/ai-features

[15] Cohen, Z. and Amble, S. (2025). “How Generative Engine Optimization (GEO) Rewrites the Rules of Search.” Andreessen Horowitz, May 28, 2025. https://a16z.com/geo-over-seo/

Methodology note. Last verified 2026-10-13. The four-tier Buyer Bucket, Tier Drift, and the Counterparty Check are TwoSetAI’s own terms, not industry standards; use them with that caveat. The sentence patterns per tier are my own observations, logged across ChatGPT, Perplexity, Google AI Mode, and Claude between July and October 2026, incognito, in private windows; they are illustrative, not a scraped dataset. Third-party studies here come from companies that sell AI-visibility or PR products (Ahrefs, OtterlyAI, Ranqo, SparkToro, BrightEdge via Stockebrand), so treat the exact numbers as directional. The TwoSetAI audit [5] covered a cohort of past guests at roughly seven queries each, not the full catalog, between August 10 and 25, 2026. The Alexander Berger three-engine results [5] were reproducible in August 2026 and may differ on your run. Nothing here guarantees a citation, a tier, or a business outcome.