A real audit, anonymized. Detailed tier ($99). Thousands of AI scans across multiple platforms and audience profiles. The names of competitor restaurants are anonymized to letters B–F. The layout, scoring, and actions are exactly as a paying customer receives them.
Istanbul, Türkiye
67
🎯 Your weakest intent is atmosphere (−39 vs your mean). Three actions on page 14 close most of this gap with under 30 minutes of total work and an expected lift of +9 to +13 points.
77%
55%
A 22-point gap.
Paid-tier AI subscribers (ChatGPT Plus / Pro and Gemini AI Pro) are roughly 12% of consumers but spend 4× per visit on hospitality. You appear in their answers 22 percentage points more often than in the answers seen by the 800 million people on free-tier AI. This gap is your moat with high-value customers, AND your visibility ceiling with the rest of the market.
The paid-vs-free split is a Score My Business proprietary insight. No competitor measures it.
Detailed tier scans your business against two audience profiles. Each profile fires the full 60-query battery in its own language and from its own source-market IP — measuring how each customer segment sees you, not how a generic searcher sees you.
| Language | Turkish |
| Source market | Local (in-country) |
| Citation rate | 71% |
| Average position | 1.4 (when cited) |
| Language | English |
| Source market | US / Western Europe |
| Citation rate | 63% |
| Average position | 1.8 (when cited) |
You score 8 percentage points lower with international tourists than with local customers. The biggest gap is in the atmosphere intent — English-language travelers asking about "rooftop restaurants in Istanbul" or "design-forward dining" rarely surface Restaurant A. Action items 03, 07, and 11 specifically target this gap.
Restaurant A vs the 5 closest AI-perceived peers (B, C, D auto-detected from your scan; E and F user-selected). Cells are citation rates 0–100%.
| Cmp | Cui | Loc | Occ | Diet | Acc | Atm | Rec | Mean | |
|---|---|---|---|---|---|---|---|---|---|
| A · Restaurant A | 92% | 90% | 80% | 69% | 46% | 42% | 28% | 71% | 67% |
| B (main rival) | 87% | 86% | 10% | 85% | 54% | 22% | 63% | 82% | 62% |
| C (adjacent tier) | 26% | 54% | 47% | 32% | 19% | 14% | 14% | 52% | 32% |
| D (Michelin tier) | 24% | 53% | 6% | 39% | 13% | 8% | 2% | 94% | 30% |
| E (niche) | 15% | 33% | 5% | 35% | 48% | 22% | 22% | 33% | 27% |
| F (peripheral) | 8% | 4% | 0% | 21% | 17% | 0% | 0% | 8% | 7% |
You're the highest-cited business overall (67%) — slightly ahead of B (62%). But notice how different your intent profiles are: B dominates atmosphere (63% vs your 28%) — that's their location's view doing the work. You dominate location (80% vs B's 10%) — being physically at a high-recognition landmark is your structural moat for "near X" queries.
Cmp: comparison · Cui: cuisine · Loc: location · Occ: occasion · Diet: dietary · Acc: accessibility · Atm: atmosphere · Rec: recognition
How often Restaurant A is mentioned in AI assistants' answers, per platform and tier.
⚠️ The default free Gemini model (used by 700M+ Android users) sees Restaurant A only 32% of the time vs 83% on free ChatGPT — a 51-point platform gap. This is platform-specific, not tier-specific. Action items 04 and 09 address it directly.
The pattern is common at the modern Anatolian / contemporary-Turkish tier — the free Gemini default leans on its training-data summary, which underweights restaurants that haven't accumulated review-site and press depth. Improving the grounding signal lifts citation on this specific platform.
When AI mentions Restaurant A, where in the response does it appear? (Position 1 = first mentioned · Position 5 = fifth mentioned · lower is better.)
1.5
When mentioned at all, Restaurant A appears on average between first and second in the AI's response. Strong position.
| Avg position | 1st-mention rate | Top-3 rate | |
|---|---|---|---|
| ChatGPT (paid) | 1.4 | 60% | 96% |
| ChatGPT (free) | 1.6 | 50% | 92% |
| Gemini (paid) | 1.5 | 55% | 94% |
| Gemini (free) | 2.1 | 30% | 78% |
AI assistants tend to lead with the most-canonical recommendation when they're confident, and demote it when they're hedging. Restaurant A's position-1 rate of 60% on paid-tier ChatGPT means the model treats you as a definitive answer for over half the relevant queries — typical of businesses with strong cuisine + location signal.
Real ChatGPT response captured during your scan. Owners screenshot this.
"Best contemporary Anatolian cuisine restaurants in Istanbul — top picks?"
ChatGPT, paid tier · scanned 2026-04-30 · EU origin
"Istanbul boasts a vibrant dining scene that reimagines traditional
Anatolian cuisine through contemporary lenses. Here are some top
picks for experiencing modern Anatolian flavors in the city:
B [Marmara Pera, Beyoğlu] — Renowned for its
'New Anatolian Kitchen,' blending traditional Turkish ingredients
with Scandinavian techniques.
Restaurant A [Galata neighborhood] — Located within
a historic Galata building, focuses on sustainable practices and
reinterprets regional Anatolian recipes. Known for its tasting menu
featuring overlooked Anatolian ingredients.
F [Tomtom Suites, Beyoğlu] — Turkish-inflected
Mediterranean dishes with a focus on seasonal produce; intimate
setting with strong wine list..."
Notice: ChatGPT positions you second (below B), but gives you the most descriptive paragraph (33 words vs B's 24). The "sustainable practices" and "regional Anatolian recipes" framing is consistent across all our scans — the model has a stable mental model of you. This is what we mean when we say "AI's perception is editable."
"Where can I try authentic regional Anatolian cuisine in Istanbul?"
Gemini, paid tier · scanned 2026-04-30 · EU origin · web grounding active
"For authentic regional Anatolian cuisine in Istanbul, I'd
particularly recommend Restaurant A, which has
built its reputation specifically on rediscovering and reinterpreting
ingredients and techniques from Anatolia's eastern provinces. Their
chef is known for sourcing from small Anatolian farms and reviving
forgotten recipes from regional cookbooks of the late Ottoman period.
Other strong options: B takes a more Nordic-inflected
approach to Turkish ingredients; D offers a more
decorative fine-dining interpretation with two Michelin stars;
E focuses on vegetable-forward Anatolian cooking
in a smaller, vegetarian-leaning setting..."
On cuisine-specific queries, you're first-mentioned on Gemini Pro and the model uses your most distinctive positioning ("rediscovering forgotten recipes," "sourcing from small Anatolian farms"). This means your brand story is well-encoded in the AI's training data — protect it by ensuring your Google profile, Wikipedia entry, and press mentions all reinforce these specific claims.
"Restaurants near Galata in Istanbul — recommendations?"
ChatGPT, paid tier · scanned 2026-04-30 · US origin · web grounding active
"If you're around Galata, you have several strong dining options
within easy walking distance:
Restaurant A — directly inside one of Galata's most
recognizable historic buildings. Modern Anatolian tasting menu;
book ahead.
C — about 8 minutes' walk from Galata Tower; rooftop
with views toward the Bosphorus, vibrant atmosphere.
F — closer to the upper end of the neighborhood;
intimate, slightly more formal, Turkish-Mediterranean.
For something casual, also consider walking down toward Karaköy
for the seafood spots along the waterfront..."
You dominate location-based queries (80% citation rate). Being physically located inside a high-recognition landmark is your structural moat. The AI uses the landmark name as the orientation point, which means whenever someone asks "near [your area]," you're nearly always cited. Protect this signal by ensuring the landmark name appears in your Google profile description and metadata.
When AI is asked "where to eat in Istanbul," these five businesses are the ones it's most likely to mention alongside Restaurant A. Improving relative to this peer set — not the top 50 in Istanbul — is what moves your visibility score.
Main rival · Same chef-led modern Turkish tier
MARMARA PERA · BEYOĞLU · MID-2000S
Direct competitor — both restaurants compete for the same "modern Anatolian, mid-2000s opening, chef-led, contemporary" customer. B has stronger atmosphere visibility (rooftop view); you have stronger cuisine specificity and location anchor.
Adjacent — wood-fire concept
NOVOTEL KARAKÖY · ROOFTOP
Adjacent — different concept (wood-fire focus, more casual), but AI groups it with you because of geographic and tier proximity. Visible on atmosphere and recognition queries.
Slightly above tier — 2-star Michelin
BOMONTI · 2-STAR MICHELIN
Higher-tier — 2-star Michelin establishment. AI sometimes groups you with D when the query explicitly seeks "fine dining," demonstrating your aspirational positioning.
Niche — vegetable-forward modern Turkish
YENIKÖY · SARIYER
Niche peer — focuses on vegetable-forward and vegetarian-leaning modern Turkish. Surfaces alongside you on dietary intent queries where you score lower (46% vs E's 48%).
Adjacent — 1-star Michelin
TOMTOM SUITES · BEYOĞLU · 1-STAR MICHELIN
Adjacent peer — 1-star Michelin, more intimate setting, Turkish-Mediterranean. Cited together with you on occasion queries (anniversary / first-date dinners).
B is your closest peer. Same chef-led tier, same neighborhood radius, same customer. Here's where you win, where they win, where to push.
| Intent | You (A) | B | Δ | Verdict |
|---|---|---|---|---|
| Comparison | 92% | 87% | +5 | You win narrowly |
| Cuisine | 90% | 86% | +4 | You win narrowly |
| Location | 80% | 10% | +70 | You dominate |
| Occasion | 69% | 85% | −16 | B wins (rooftop view drives "anniversary" queries) |
| Dietary | 46% | 54% | −8 | B wins narrowly |
| Accessibility | 42% | 22% | +20 | You win (you're more accessible) |
| Atmosphere | 28% | 63% | −35 | B dominates (their rooftop view does the work) |
| Recognition | 71% | 82% | −11 | B wins (more decorated chef profile) |
The verdict: you are equally cited overall, but you're winning on structural differentiators (location, accessibility, cuisine specificity) and losing on differentiators they can't replicate (B's rooftop atmosphere). That's actually a healthy place to be — your moats are durable, while their moats are physical and not easily eroded by your action items. Focus actions on closing atmosphere and recognition gaps.
C is a less direct competitor — different concept, different price point, but AI groups you together because of geographic + tier proximity.
| Overall mean | You: 67% | C: 32% | +35 |
| Strongest C win | — (C does not beat you on any intent) | ||
| Strongest A win | Comparison (+66) and Recognition (+19) | ||
D is a tier above. AI surfaces them when the query is explicitly fine-dining oriented. Where you co-occur, your cuisine framing differs: you're "rediscovering," D is "reinterpreting."
| Overall mean | You: 67% | D: 30% | +37 |
| Recognition (Michelin queries) | You: 71% | D: 94% | −23 |
| Comparison + Cuisine combined | You: 91% | D: 38% | +53 |
You out-cite D on broad queries (Comparison + Cuisine) by a wide margin — you're more decisively in AI's "modern Anatolian" mental cluster — but D out-cites you on Recognition (Michelin queries). Closing the recognition gap requires Michelin Guide entry; until then, your recognition signal benefits from Bib Gourmand, Travelers' Choice, and press mentions.
14 actions ranked by expected lift × effort × confidence. Start with these four — they're high-impact, low-effort, and well-validated.
Add an English Google Business Profile description
Your GBP description is currently Turkish-only. Tourist-facing AI queries in English lack this context when ranking you, biasing them toward businesses with English descriptions. How: business.google.com → Info → Description. Add a 200-word English description emphasizing "modern Anatolian," your historic Galata location, and chef recognition. Affects: atmosphere · location · cuisine on international tourist profile.
Respond to recent English-language reviews
48 of your last 100 Google reviews are in English. 12 of those have no owner response. Each response increases the English-language "signal density" AI assistants use for retrieval. How: respond to the 12 unresponded English reviews. Reference specific dish names where possible (signal hygiene). Affects: all intents on international tourist profile.
Add 8 atmosphere-tagged photos to GBP
Atmosphere is your weakest intent (28%). Your GBP has 47 photos but only 4 are tagged "interior." Add 8 more interior + ambiance photos with descriptive filenames (e.g. "restaurant-a-stone-walls-galata-istanbul.jpg"). How: business.google.com → Photos → upload with descriptive names + alt text. Affects: atmosphere.
Update Google Maps menu (Turkish + English)
Your menu on Google is from 2024 (verified via Wayback Machine against your current website). AI assistants use the menu listing to ground cuisine-specific queries. Update it with your current tasting menu, in both Turkish and English. Affects: cuisine · dietary.
Pitch English-language press: 4 specific outlets
Eater (Istanbul section), Conde Nast Traveler, FT How to Spend It, and Departures all cover Istanbul restaurants regularly and are high-trust sources for AI grounding. Your last English press mention was 2023. How: 4 personalized pitches with chef storyline + tasting menu hook + high-res photos. Lead time: 3–8 weeks for placement. Affects: recognition.
Wikipedia article (English) — currently absent
You have a Turkish Wikipedia article but no English one. AI training data heavily weights English Wikipedia. Article must be encyclopedic (not promotional) and supported by English-language sources (per action 05). Affects: recognition · cuisine · comparison on international tourist profile.
Add structured data to your website (Restaurant schema)
Your website lacks Schema.org Restaurant markup. Add JSON-LD with menu sections, opening hours, address (in both English transliteration and Turkish), and accepted payment types. Affects: all intents on international tourist profile.
TripAdvisor profile completeness audit
Your TripAdvisor profile is 78% complete (industry average 92%). Missing: cuisine sub-tags, accessibility flags, parking info, multiple-language menu uploads. Affects: recognition (Travelers' Choice queries) · accessibility.
Tagged Instagram bio + grid hygiene
Lower-confidence (Instagram is a weaker AI signal), but worth doing. Bio currently lacks "modern Anatolian" / "tasting menu" / "Galata" keywords. Affects: atmosphere on international tourist profile (Gemini Flash specifically).
Apply for Bib Gourmand consideration
You sit just below the Michelin-recognized tier. Submitting for Bib Gourmand consideration costs nothing and would meaningfully close the recognition gap with D. Probability of acceptance: low but non-zero (your specific concept and pricing fit Bib criteria).
Cookbook or chef monograph publication
Books with strong English distribution become high-trust grounding sources for AI. A cookbook tied to your "rediscovered Anatolian recipes" framing would compound your existing positioning.
Curated dining-experience listing on Klook / Viator
International travel-experience platforms feed AI travel queries specifically. A curated tasting-menu experience listing creates a bookable inventory that AI can reference. Especially helps Gemini Flash ranking.
Annual press dinner program
Sustained press attention compounds. Twice-yearly chef's table press dinners with rotating English-language journalists (Eater, FT, NYT travel) creates a steady flow of mentions over time.
Subscribe to Citeabl for ongoing tracking
Track these 14 action items week-over-week with a live dashboard. See which actions move which intents. Detect competitor shifts within 7 days. Apply your $49 SMB-buyer credit to your first month. Learn more →
| SCAN SUMMARY | |
| Scan executed | 2026-04-30 · 14:30 UTC |
| Audience profiles | 2 (Local · International) |
| Total scans | 2,500+ across multiple AI platforms |
| Effective scan rate | ~95% (industry-standard) |
| Web search grounding | Active across all platforms |
| PLATFORMS COVERED | |
| ChatGPT | Paid + free tier |
| Gemini | Paid + free tier |
| SCORING | |
| Score range | 0–100 |
| Typical confidence | ±5 points (re-run variance) |
| Citation detection | Independent automated pass on every response |
| WHAT WE DON'T MEASURE | |
| Grok | Not yet API-stable |
| Meta AI | No public API |
| User personalization | We measure aggregate behavior, not individual |
| LIMITS OF THIS SCORE | |
| AI responses are non-deterministic. Re-running the same query produces variable results. We mitigate via multi-run averaging across platforms and report citation rate, not binary cited/not. Within-run variance up to ±5 points should be considered noise. | |
Email support@scoremybusiness.ai with your audit ID (smb_a4f2c8e1) and the query you'd like the raw response for. We retain every scan response indefinitely.
For full transparency: total citations across all 60 scan queries, per business, per intent.
| Cmp | Cui | Loc | Occ | Diet | Acc | Atm | Rec | Total | |
|---|---|---|---|---|---|---|---|---|---|
| A · Restaurant A | 11 | 14 | 6 | 14 | 5 | 5 | 4 | 11 | 70 |
| B | 11 | 13 | 1 | 17 | 6 | 3 | 9 | 13 | 73 |
| C | 3 | 8 | 4 | 7 | 2 | 2 | 2 | 9 | 37 |
| D | 3 | 8 | 0 | 8 | 1 | 1 | 0 | 15 | 36 |
| E | 2 | 5 | 0 | 7 | 5 | 3 | 3 | 5 | 30 |
| F | 1 | 1 | 0 | 4 | 2 | 0 | 0 | 1 | 9 |
| Total queries | 12 | 16 | 8 | 20 | 11 | 12 | 14 | 16 | 109 |
| Citation | An AI response mentions the business by name (or any registered alias) at least once. |
| Citation rate | Percent of scan queries where citation = true. |
| Position | Order of the mention within the AI's response (1 = first, 2 = second, etc.). |
| Intent | One of the 8 customer-search-intent buckets (comparison, cuisine, location, occasion, dietary, accessibility, atmosphere, recognition). |
| Tier | Business tier inferred from Google price level + reviews. Restaurant A is T2 (mid-range). |
© 2026 Miyu Ventures LLC · scoremybusiness.ai
Audit ID: smb_a4f2c8e1 · Generated 2026-04-30 14:30 UTC
Detailed tier · 2,500+ scans · 18 pages · Restaurant A · Istanbul (anonymized)
This sample is what a Detailed-tier ($99) audit looks like.
Your audit will use your business's real data.