HOW TO TELL IF CHATGPT RECOMMENDS YOUR D2C BRAND
(AND WHAT TO DO WHEN IT DOESN'T)

If a D2C founder asks ChatGPT "what's the best clean-ingredient sunscreen under ₹1,500," one of two things happens: your brand gets named, or a competitor does. There's no third option and no partial credit. Most founders assume they're in the conversation. Our audits across 40+ Indian D2C brands in 2026 show 83% are invisible on the exact queries their highest-intent buyers are asking. This guide walks you through the 10-minute test we run first on every audit — and the six levers that move the needle once you know where you stand.

The 4-Step Test You Can Run in 10 Minutes

Step 1: Write Your Test Prompts

Don't just type your brand name into ChatGPT. Your buyers aren't searching for you by name — they're looking for solutions. Write three prompts based on how your buyers actually ask for help:

• Category query: "What are the best [your category] brands in India under [your price point]?" — this is the money query. Most brands are invisible here.
• Comparison query: "Compare [your brand] vs [top competitor]" — if ChatGPT doesn't know you well enough to compare you, that's diagnostic.
• Problem-solution query: "I have [buyer's specific problem]. What should I use?" — phrased exactly how your target buyer would speak, not how your marketing team writes.

Step 2: Test Across All Three Engines

Run each prompt on ChatGPT (GPT-4o), Perplexity, and Google Gemini. Use a fresh incognito window for each. Screenshot every response — you'll need them for comparison. Don't test in the same chat session; start fresh each time. AI responses have session-level memory that can inflate scores.

Step 3: Score Each Response

For each prompt × engine combination (9 total), mark:

Result	Definition	Points
Cited	Named positively, recommended, or featured	3
Mentioned	Named neutrally, used as example, or compared	1
Invisible	Not mentioned at all	0

Step 4: Calculate Your Score

Maximum possible: 27 points (9 combinations × 3 points). Anything below 9 is critical. Anything above 18 means you're genuinely visible. Most D2C brands we audit score between 2 and 7 — they get a mention on the comparison query if they're lucky, and nothing on the category and problem-solution queries.

What "Being Cited" Actually Means in 2026

Direct Citation

The AI names your brand in its response, positions it positively, and often includes a brief description of what makes it distinctive. This is the gold standard. It typically means you have: a clear brand description in structured web content, consistent mentions across review platforms and community sites, and content that directly addresses the query's intent.

Comparative Mention

The AI names your brand in a comparison — "X is known for Y, while [your brand] is known for Z." This is useful but not where you want to stay. It means the AI has training data about you but can't recommend you as a standalone answer. The fix is usually a content and positioning gap.

Invisible

The AI doesn't mention you at all. This doesn't mean your brand doesn't exist — it means the AI's training data doesn't have enough signal about you to include you in an answer about your category. The AI isn't choosing your competitors over you — it literally doesn't know you're an option. You're not alone: 2026 audits found 98.8% of local businesses are invisible in AI recommendations, and ChatGPT surfaces just 1.2% of locations (the 2026 AI search data).

Why D2C Brands Specifically Are Losing Here

The Generic Category Prompt Problem

Most D2C brands have optimised for Google, which rewards keyword-specific content. AI engines reward category authority — the brand that has the most consistent, structured, trustworthy signal across the web about what it is and who it's for. If your entire digital footprint is ads and product pages, you're invisible to AI.

Amazon and Aggregators Dominate Training Data

A significant portion of AI training data comes from Amazon, Flipkart, and large review aggregators. If your brand isn't well-reviewed on those platforms, or if reviews don't specifically describe your differentiation, you're not generating the signal that feeds AI recommendations.

The "Top 5" Problem

When ChatGPT answers "best sunscreens in India," it typically lists 4–6 brands. The brands in that list weren't chosen by an algorithm looking at your SEO score — they were chosen because they appear most consistently and credibly across the web as answers to that type of question. Getting into that set is a content and authority problem, not a keyword problem.

Fixing It — The 6-Lever Framework

Lever 1: Structured Content

Publish blog content that directly answers the queries you tested. Not SEO-style keyword-stuffed posts — genuine, structured Q&A content that names your category, your buyer, your specific differentiators. Add FAQ schema and HowTo schema to relevant pages. AI engines extract structured content preferentially.

Lever 2: Schema Markup

Implement Product schema on every product page (name, brand, category, description, review aggregate). Implement Organization schema on your homepage. These signals feed structured AI training data about who you are and what you sell.

Lever 3: Third-Party Citations

AI engines weight third-party mentions of your brand heavily. This means: getting reviewed on independent review sites, being mentioned in editorial content (not sponsored), appearing in "best of" lists on non-affiliate blogs, and being discussed in Reddit threads and Quora answers in your category. Each organic third-party mention is training signal.

Lever 4: Community and Forum Presence

Reddit and community forums (Quora, niche Facebook groups, Discord communities) are disproportionately present in AI training data. A brand mentioned positively in 20 genuine Reddit threads about skincare will outrank a brand with ₹5L of sponsored content in AI citations. Earn community presence, don't fake it.

Lever 5: PR and Media Mentions

A single credible media mention (YourStory, Inc42, Vogue India, The Hindu) is worth dozens of generic backlinks for AI citation purposes. The signal is: credible, non-affiliated sources have written about this brand in the context of its category. This is achievable for any brand with a genuine story.

Lever 6: Product Data Consistency

Ensure your product name, category, key ingredients/features, and target buyer are described identically across your website, Amazon listing, Flipkart listing, social bios, and any distribution partner pages. Inconsistency fragments the signal. Consistency compounds it.

How This Connects to Your SEO and Organic Traffic

Your AI visibility is not separate from your SEO — it is built on it. The pages that already rank on Google and pull organic traffic are the same sources ChatGPT and Perplexity scrape to decide who to recommend. Weak on-page SEO and thin content show up twice: once as lost rankings, once as a missing AI citation.

Fix the fundamentals — keyword research, on-page SEO, authoritative content — and you improve both numbers at once.

How We Measure It at antral.

The test above gives you a directional answer. Our full AI Answer Share-of-Voice audit runs 25+ category-specific prompts across all three major AI engines, scores how often you appear, where you rank, sentiment and framing relative to competitors, and identifies the specific content and signal gaps driving your score. It's the first deliverable in every engagement we offer.