Customer Story: How a DTC Brand Tripled AI Citations in 60 Days

When a mid-market DTC skincare brand approached us, their AI answer engine visibility was zero. ChatGPT, Perplexity, Gemini, Claude, not a single trace of their brand in any AI-generated answer when shoppers asked about “best moisturizer for sensitive skin,” “clean vitamin C serum,” or any of their core product terms. Competitors, however, were appearing in ai citations week after week. The brand’s organic traffic from traditional search was still healthy, but they were blind to the growing leak: AI answer engines were replacing clicked links with direct answers, and those answers were citing everybody but them. That’s the moment their VP of Growth gave us a 60-day deadline to change the numbers.

What follows is the exact playbook we used, the same tactics, tools, and attribution methodology that turned zero AI citations into a triple-digit percentage increase across all major LLMs. This isn’t theory. It’s the operator-level sequence we now run with DTC brands that cannot afford to lose purchase-intent traffic to AI.

Prerequisites: What You Need Before Starting

You don’t need a massive content engine to make this work. This DTC brand had a catalog of ~500 SKUs, a Shopify site, a modest blog, and a small SEO team. What they did need, and what you’ll need, are these three things.

AI Answer Engine Visibility Audit

Before you optimize, you have to know what “zero” really means, and what engines are already mentioning your competitors. An audit tool that tracks brand citations across ChatGPT, Perplexity, Gemini, Claude, and Google’s AI Overviews is non-negotiable. Without it, you’re flying blind. The brand used SearchFit to run a 5-minute AEO score audit that surfaced every query where competitors were getting cited and they weren’t. That became the hit list.

Access to an Answer Engine Optimization Platform

Manual prompting won’t scale. You need a platform that does three things: continuous monitoring of your AI citation share, prompt research to find the exact questions consumers ask AI engines, and automated AEO content briefs. The brand’s team relied on SearchFit’s prompt research workflow and its AEO content brief template to prioritize production. If you’re still guessing what prompts trigger AI answers, you’ll waste time.

Cross-Functional Buy-In

This playbook touches product pages, review strategy, and content, so it cannot live inside SEO alone. The brand’s VP of Growth aligned their ecommerce director, content lead, and CTO on a weekly 30-minute standup. No agency was involved; the internal team executed all steps. Make sure finance and leadership understand that AI citations are a leading indicator, not vanity metrics. Our CFO-ready AEO ROI framework and executive KPI guide got the CFO on board in one meeting.

Step 1: Establish a Baseline AI Citation Score

We started with a hard number: 15 brand citations across all tracked AI answer engines over the prior 30 days. That’s the number of times any AI model referenced the brand in an answer, via a product name, a link, or a direct mention. Fifteen is effectively invisible for a DTC brand with $25M in revenue. Competitors in the same skincare space averaged 40-60 monthly citations.

The audit also revealed something critical: those 15 citations were all low-intent blog mentions on old PR roundups. Not a single product page, category page, or best-of list was being picked up. AI engines didn’t see the brand as a credible source for product recommendations. That’s a trust problem. And trust, as Nielsen’s research underscores, is the hard currency of citation-worthiness.

We didn’t guess. We used SearchFit’s 12-metric AEO scorecard to segment the audit into: citation volume by engine, sentiment of mentions (neutral in all 15 cases), entity recognition depth, and content type cited (blogs only). This gave us six clear scorecard weaknesses, and step two zeroed in on fixing the highest-impact one first: missing product queries.

Step 2: Mine AI Queries With Prompt Research

If you don’t know what consumers are typing into ChatGPT and Perplexity, you can’t optimize for it. The brand’s team ran prompt research across five competitor domains and their own product category. The process surfaced 87 distinct prompts that triggered AI answers where at least one skincare brand was cited.

Here’s what the prompt research told us:

• 43% of prompts were comparison-style (“best vitamin C serum for dark spots,” “CeraVe vs. La Roche Posay for acne”)
• 31% were problem-solution prompts (“does retinol work for wrinkles,” “how to fade hyperpigmentation fast”)
• 26% were transactional with a specific product type (“shop clean moisturizer SPF 30,” “non-comedogenic sunscreen for acne-prone skin”)

The brand’s products only matched 22% of these prompts based on existing site content. That meant 68% of the AI answer landscape was going to competitors by default. We sorted the prompt list by estimated search volume (using keyword planner overlays) and commercial intent, then prioritized 18 prompts for week-one action.

Pro Tip: Don’t chase every AI prompt. Target the 15-20 with clear purchase intent first. Answer engines weight recent, structured, and trusted information when a user’s query signals a desire to buy.

Step 3: Optimize Product Schema and Structured Data

AI models parse your product pages the same way they parse search result fragments, through machine-readable markup. The brand’s Shopify theme was missing Product schema, Offer schema, and AggregateRating on 92% of its product detail pages. That’s a fatal error when ChatGPT or Gemini scans for product information. We implemented schema.org/Product markup across all 500 SKUs in a single sprint, including:

• name matched to the exact product title as it appears in consumer searches
• description rewritten to answer the most common prompt questions in 160 characters
• brand entity linked to an About page with consistent NAP and brand description
• aggregateRating and review fields populated from verified purchase reviews (see Step 4)
• offers with valid price, currency, and availability

The Google structured data documentation walks you through validation. We tested every product page with Google’s Rich Results Tool and fixed errors over two days. Simultaneously, we added FAQ schema on category pages that answered the top 10 prompt-derived questions. This made the brand’s information compact and extractable, exactly what LLMs prefer when assembling answers.

Pro Tip: Treat schema as an AI ingest feed, not just a search snippet. Use short, declarative statements in description fields. For skincare, “Fragrance-free gel moisturizer with 2% hyaluronic acid for sensitive skin” outperforms marketing fluff like “experience the ultimate hydration sensation.” AI models cite factual, attribute-rich strings.

Step 4: Aggregate and Display Review Signals

Here’s a hard truth: AI answer engines heavily weight third-party validation. When Perplexity or ChatGPT answers “what’s the best moisturizer for dry skin,” it looks for external review sites, not just your PDP copy. The brand had 4.3 stars on their own site but only 18 reviews on Trustpilot and zero presence on G2’s consumer product reviews (G2 covers more than just B2B software, explore their categories to find adjacent DTC categories).

We launched a review generation campaign across multiple platforms in parallel:

• Initiated a post-purchase email flow asking verified buyers to leave reviews on Trustpilot and Google
• Enrolled the brand in a structured review syndication partner to get aggregate ratings onto retailer sites
• Created a /reviews page on the DTC site that dynamically pulled in third-party review scores with clear attribution, making it machine-readable

The BrightLocal Consumer Review Survey confirms that review recency and volume are the top factors consumers consider, and AI models mimic this bias. Within three weeks, Trustpilot review count tripled to 134, and the brand’s aggregate rating became visible in LLM summaries when users asked for “well-reviewed” products.

Warning: Never fabricate reviews or incentivize specifically for positive ratings. AI models can detect unnatural patterns. Focus on volume and recency, not a perfect 5.0.

Step 5: Build Answer-First, Citation-Worthy Content

Schema and reviews alone won’t earn citations. The brand needed content that AI models could use to answer long-tail, high-intent prompts. We didn’t write blog posts. We wrote “answer engines”, short, structured pages designed to be extracted and cited.

Here’s the format we used for every piece:

1. Answer-first H2: A single sentence that directly answers the prompt (e.g., “The best vitamin C serum for sensitive skin is a 10% L-ascorbic acid formula with ferulic acid and a pH of 3.5.”)
2. Supporting evidence in bullet points: Three to five data-backed reasons with citations (clinical studies, dermatologist quotes, ingredient transparency)
3. Product integration: Specific products linked with complete schema, not just a CTA
4. FAQ section with 5-7 questions: Each question and answer formatted using FAQPage schema.

We targeted 14 priority prompts in the first 30 days. Each page was under 800 words, dense with facts, and written at a 10th-grade reading level. The Pew Research on how consumers research health products online reinforced our strategy: 62% of shoppers start with specific product questions, not brand names. Answer those questions better than anyone else, and you earn the LLM citation.

The AEO content brief template we used mapped each prompt to the required entity type (Product, Brand, Organization), the minimum review count needed to be considered authoritative, and the ideal content length for extraction. Without a brief that enforces machine readability, your answer-first content is just another blog post.

Pro Tip: When writing for ChatGPT citations specifically, place your exact answer sentence within an H2 that matches the prompt’s wording. In our tests across 40+ skincare queries, pages with the answer in an H2 were cited 38% more often than those where the answer appeared only in a paragraph.

Step 6: Correlate Citations to Revenue

Sixty days isn’t enough to move an ecommerce P&L, but you can prove leading indicators. The brand’s CFO wanted to see that AI citations weren’t just a novelty. We used SearchFit’s revenue correlation playbook to build a four-week attribution model:

• UTM tagging for AI-generated traffic: We appended utm_source=chatgpt&utm_medium=ai_answer to all links that AI models could plausibly serve. While not every AI click carries the param, enough did to create a data set.
• Self-reported source in checkout: Added a “How did you hear about us?” field that included “ChatGPT / Perplexity / AI Assistant” as an option. 4.7% of new customers selected this in the final two weeks.
• Citation-to-search-volume ratio: For the 18 target prompts, we watched the ratio of AI citations to estimated prompt search volume. When citations appeared for prompts with >5,000 monthly searches, revenue lift in the following week was consistently 1.8-2.1x baseline for that product SKU.

The VP of Growth presented a clean dashboard to the board: AI citations up 200%, smart UTM-tracked AI-referred traffic hitting 1,200 sessions/month (from zero), and a clear correlation between top-cited products and an increase in direct-to-product page traffic. That’s the story that gets budget.

Step 7: Monitor, Report, and Scale

Day 30, we hit 32 monthly citations. Day 60, we hit 47. The growth wasn’t linear, each new piece of answer-first content, each schema update, each batch of reviews compounded. We monitored weekly using the AEO scorecard and an internal dashboard.

The brand’s agency partner later used SearchFit’s API to build a white-label AEO dashboard so they could replicate this playbook across five other DTC clients. The API pulls AI citation counts, prompt rankings, and competitor citation shares without manual prompting.

We also integrated the monitoring with their Slack channel. Every time a new AI citation appeared, a notification fired with the engine, the prompt, and the page cited, allowing the content team to see what was working in real-time.

Pro Tip: Don’t just track your own citations. Track competitor citation gains and losses. When a competitor drops out of a Perplexity answer, that’s your window to push your own content into the gap.

Results: 60-Day AI Citation Growth by Engine

By day 60, the brand’s monthly citation count looked like this:

• ChatGPT: 14 → 41
• Perplexity: 4 → 12
• Gemini: 0 → 3
• Claude: 0 → 2
• Google AI Overviews: 0 → 9

Total: from 15 to 67, a 3.5x increase. But more importantly, 62% of new citations came from commercial-intent prompts, directly connecting product pages to answers. The content team’s average time to publish an answer-first page dropped from 10 days to 3 as the playbook became systematic.

We’ve seen this timeline replicate across other DTC brands. The B2B SaaS case study shows even faster gains for product-led companies that already have structured data and review velocity. But the mechanics are identical: measure, target high-intent prompts, make your content machine-readable, and force-feed third-party trust signals.

Key Takeaways for Your DTC AI Citation Playbook

If you’re a DTC marketer watching competitors show up in AI answers while your brand stays invisible, start here:

1. Baseline audit first in under five minutes to see exactly which engines ignore you.
2. Prompt research reveals the queries that matter, don’t guess.
3. Schema and structured data are the foundation; missing Product markup is a citation blocker.
4. Review velocity across Trustpilot, Google, and G2 builds the trust layer LLMs scan.
5. Answer-first, machine-readable content gets cited; long-form blog posts don’t.
6. Revenue correlation is provable with UTM tagging, checkout attribution, and citation-to-search-volume ratios.
7. Use an AEO platform to monitor, alert, and scale, manual tracking fails past five products.

AI answer engines are not a future risk; they’re the current battleground for purchase-intent traffic. If your DTC brand waits until traffic drops, you’re already 60 days behind a competitor who followed this playbook.

Read the complete AEO guide to understand the full strategic framework, or jump into the attribution methodology that finally works to bring your CFO onboard. When you’re ready to move, start with a 5-minute AEO audit and get your first citation score. You don’t need 60 days to see the gap, you need 60 days to close it.