Skip to content
Industry

The Coffee Wars in AI: How Dunkin', Blue Bottle, and 8 Chains Compare Across 4 AI Engines

geobuddy.co
G
Industry Deep-DiveCoffee Chain9 Brands × 4 Engines

The Coffee Wars in AI: How Dunkin', Blue Bottle, and 8 Chains Compare Across 4 AI Engines

We asked ChatGPT, Claude, Gemini & Perplexity to recommend coffee chains 108 times. Dunkin' scores 67% visibility but only as an 'alternative.' Blue Bottle scores 33% but earns the only Primary Recommendation. Here's every data point.

G
GeoBuddy Research
March 22, 202612 min read

Here's a question no coffee CMO has asked yet — but should: When a consumer asks ChatGPT “what coffee should I try,” does it recommend your brand?

We ran 9 coffee chains through all 4 major AI engines — ChatGPT, Claude, Gemini, and Perplexity — with 12 different prompts each. The results are not what you'd expect. The chain with the most locations isn't the most recommended. The chain with the highest visibility score isn't positioned as the best choice. And the one that AI actually recommends first? It has half the visibility of the leader.

☕ The headline finding:

Dunkin' scores 67% AI visibility — but every single mention positions it as an “alternative.” Blue Bottle scores just 33% — but it's the only coffee chain in our dataset to earn a Primary Recommendation from any AI engine. In the AI coffee wars, quality of mention beats quantity of mention.

9

Coffee Brands Tested

across all 4 AI engines

12

AI Prompts Per Brand

3 prompts × 4 engines

67%

Top Scorer

Dunkin' — but only as 'alternative'

Blue Bottle

Primary Rec Winner

33% vis — but recommended first

The AI Coffee Leaderboard: 9 Chains Ranked

We tracked AI visibility — the percentage of prompts where each brand was mentioned by at least one engine. Here's the full ranking:

AI Visibility Score: 9 Coffee Chains Ranked

Percentage of AI prompts where each brand was mentioned (higher = more visible)

🟡 Gold = Primary Recommendation | 🟣 Purple = Alternative/Comparison | ⚪ Gray = Invisible

At first glance, this looks like a market-share chart — Dunkin' on top, smaller chains at the bottom. But look closer. Blank Street, one of the fastest-growing coffee startups in the US, scores a flat 0%. Not a single AI engine mentions it. And Starbucks — the 800-pound gorilla of coffee — isn't even tracked as a standalone brand because AI uses it as the default benchmark to compare everyone else against.

The AI Engine Heatmap: Who Recommends What

Each AI engine was tested with 3 distinct prompts. Here's the granular breakdown of how many times each engine mentioned each brand:

Which AI Engine Mentions Which Coffee Brand?

Number of mentions per engine (out of 3 prompts each). Green = mentioned, blank = invisible.

BrandChatGPTClaudeGeminiPerplexityTotal
Dunkin'22228/12
Peet's Coffee22116/12
Dutch Bros1225/12
Blue Bottle2114/12
Caribou Coffee123/12
Costa Coffee112/12
Tim Hortons11/12
McCafé11/12
Blank Street0/12

🔍 Key insight: Dunkin' is the only brand mentioned by ALL four engines. Dutch Bros completely disappears on Claude (0 mentions) but shows up on every other engine. Blue Bottle is invisible on Gemini despite being a premium brand. These aren't random — each engine has a systematic bias.

Claude ignores most mass-market chains — it didn't mention Dutch Bros, Caribou, Costa, Tim Hortons, or Blank Street at all. It reserves its mentions for specialty brands (Blue Bottle, Peet's) or the undeniably ubiquitous (Dunkin').

Gemini has a blind spot for specialty coffee — it recommends Dunkin' and Dutch Bros (both drive-thru/convenience brands) but completely ignores Blue Bottle, the only brand that earns a Primary Recommendation on other engines.

The Blue Bottle Paradox: Why 33% Beats 67%

This is the most counterintuitive finding in our entire dataset. Blue Bottle Coffee has half the visibility of Dunkin' — yet it's positioned dramatically better.

Here's what the data shows:

  • Dunkin' (67% visibility): Mentioned 8 times, but always as “alternative,” always alongside 5+ other brands, average rank #2-4. AI describes it as “affordable,” “quick,” “fast” — functional language, not aspirational.
  • Blue Bottle (33% visibility): Mentioned only 4 times, but one of those is a Primary Recommendation from Claude with the context: “Focus: Freshness, precision roasting. Best for: All brewing methods.” Sentiment score: 0.70 vs Dunkin's 0.47.

The Blue Bottle Paradox: Less Visible, but Better Positioned

Visibility (%) vs Sentiment Score — Blue Bottle is the only brand with 'Primary Recommendation' status despite lower visibility

Primary Recommendation Alternative / Comparison(Bubble size = recommendation quality)

For CMOs: This is the AI equivalent of shelf placement. Dunkin' is in every aisle but always on the bottom shelf. Blue Bottle is in fewer stores but at eye level with a “Staff Pick” sticker. When AI users are choosing, position matters more than presence.

4 AI Engines, 4 Coffee Personalities

Each AI engine approaches coffee recommendations with a distinct personality:

  • ChatGPT — The People-Pleaser: Mentions the most brands (6/9) but hedges every recommendation. Lists 7-10 options per response. Everyone gets a mention, nobody gets the crown. Dunkin' at rank #2, Blue Bottle at rank #2-3.
  • Claude — The Specialty Snob: Only mentions 4 brands but gives the only Primary Recommendation in the entire dataset (Blue Bottle). Ignores drive-thru chains. Uses precise language: “precision roasting,” “freshness.”
  • Gemini — The Practical Guide: Focuses on accessibility and value. Loves Dunkin' and Dutch Bros. Completely ignores Blue Bottle. Mentions “affordable” and “drive-thru” repeatedly.
  • Perplexity — The Data Nerd: Highest average sentiment (0.63). Cites real sources. Mentions Dutch Bros' “#5 in America's favorite chains” ranking. Most likely to include market data in responses.

How Each AI Engine Treats Coffee Brands

Each engine has a distinct 'personality' when recommending coffee chains

Primary Recommendation vs “Also Mentioned”: The Role Gap

Not all AI mentions are created equal. Being mentioned as an “alternative” is fundamentally different from being the Primary Recommendation. Here's how AI positions each brand across all 12 responses:

How AI Positions Each Coffee Brand (12 Responses)

Primary Recommendation vs Alternative vs Neutral vs Invisible — out of 12 total AI responses per brand

The data reveals a brutal truth: out of 108 total AI responses across all 9 brands, only ONE resulted in a Primary Recommendation — and it went to Blue Bottle, from Claude. Every other mention across every brand was either an “alternative,” “neutral comparison,” or outright invisible.

This means that for the coffee industry, AI is essentially a comparison engine, not a recommendation engine. It tells you what exists. It rarely tells you what's best. The exception? Brands with strong editorial authority and clear differentiation — like Blue Bottle.

Being Seen ≠ Being Loved: The Sentiment Gap

Visibility tells you how often AI mentions your brand. Sentiment tells you how it talks about you. These two metrics tell very different stories:

AI Sentiment vs Visibility: Being Seen ≠ Being Loved

Dunkin' is mentioned most but with lukewarm sentiment. Blue Bottle is mentioned less but praised more warmly.

Peet's Coffee has the highest sentiment (0.71) despite only 50% visibility. AI describes it with language like “rich, bold flavors,” “more finesse” than Starbucks, and “consistent quality.” Meanwhile, Dunkin' (67% visibility) gets described as “affordable,” “fast,” and “simpler” — functional but uninspiring language.

Caribou Coffee has a sentiment crisis: 25% visibility with only 0.17 sentiment. When AI does mention Caribou, the language is clinical: “bakery pairings in 19 states” — a factual mention with zero enthusiasm.

The Starbucks Shadow: The Brand AI Uses but Never Recommends

Starbucks wasn't one of the 9 brands we tracked — but it appeared organically in our data more than any tracked brand. AI engines used Starbucks as the default reference point in 18 out of 108 responses, making it the most-referenced coffee brand in our entire dataset.

But here's the twist: AI almost never recommends Starbucks — it recommends alternatives to Starbucks. When users ask “what coffee should I try” or “best coffee chains,” AI engines position Starbucks as the known quantity and then suggest other options.

The Starbucks Shadow: AI Uses Starbucks as the Default Benchmark

How many times each brand appears in the same AI response as Starbucks, and in what role

Starbucks wasn't tracked as a separate brand, but appeared as a competitor reference in 18 of 108 total AI responses.

What this means for Starbucks: Being the default is a double-edged sword. AI treats Starbucks like Google treats Wikipedia — always referenced, rarely the final answer. Every competitor benefits from the “Starbucks alternative” framing. For Starbucks' CMO, the strategic question is: how do you shift from being the benchmark to being the recommendation?

What Coffee Brands Should Do Right Now

1. Track AI visibility across all 4 engines — not just one

Dutch Bros is invisible on Claude but visible on Gemini and Perplexity. Blue Bottle is invisible on Gemini but gets the only Primary Recommendation on Claude. A single-engine check gives you 25% of the picture. Check your brand across all 4 engines free →

2. Invest in editorial authority, not just advertising

Blue Bottle's AI advantage comes from specialty coffee press, third-party reviews, and a clear brand narrative (“freshness within 48 hours”). AI engines cite editorial sources, not ad spend. The brands that win in AI are the ones that win in earned media.

3. Differentiate your brand story — AI penalizes the generic

Notice how AI describes Dunkin': “affordable, quick, fast.” Now Blue Bottle: “precision roasting, freshness, best for all brewing methods.” AI reflects the distinctiveness of your brand narrative. If your brand story is “we have coffee,” AI will treat you accordingly.

4. Optimize for the Starbucks comparison

Like it or not, AI frames most coffee recommendations relative to Starbucks. Brands that lean into this comparison (Peet's: “more finesse than Starbucks,” Dunkin': “cheaper than Starbucks”) get clearer positioning. Don't fight the frame — use it.

5. Don't confuse visibility with quality

Dunkin' at 67% proves you can be highly visible and still be positioned as second choice. Blue Bottle at 33% proves you can be less visible and still be positioned as the best. The goal isn't more mentions — it's better mentions.

FAQ

Which coffee chain does AI recommend most?

Dunkin' has the highest overall AI visibility at 67%, but every mention positions it as an “alternative.” Blue Bottle (33%) is the only chain to earn a Primary Recommendation from any engine.

Does ChatGPT recommend the same coffee brands as Claude or Gemini?

No. ChatGPT mentions the most brands but hedges. Claude is selective but gives the only Primary Recommendation (Blue Bottle). Gemini focuses on drive-thru/value chains. Perplexity has the highest sentiment scores.

Why does Blue Bottle get recommended by AI despite being smaller than Dunkin'?

Blue Bottle's AI advantage comes from brand positioning, not market size. AI engines evaluate editorial coverage, specialty authority, and clear brand narratives. Blue Bottle's “freshness within 48 hours” story gets cited verbatim.

Is Starbucks invisible to AI?

No — Starbucks appears in 18 of 108 responses as a competitor reference. But AI uses Starbucks as a benchmark, not a recommendation. When users ask for suggestions, AI engines recommend alternatives to Starbucks.

How can coffee brands improve their AI visibility?

Invest in editorial coverage, build a distinct brand narrative, optimize for all 4 AI engines separately, and track your visibility regularly. Start with a free check across all 4 engines →

Share:
Coffee ChainAI VisibilityBlue BottleDunkin'StarbucksChatGPTClaudeGeminiPerplexityIndustry ResearchGEO

How Does Your Brand Compare Across 4 AI Engines?

Find out in 60 seconds. Free check across ChatGPT, Claude, Gemini & Perplexity.

Check Your Brand Free