DTC Brand AEO Case Study: From 5% to 18% AI Mention Rate in 120 Days

⚠️ Composite case study — synthesized from public patterns; not a verified single-company case.

TL;DR

A mid-market DTC skincare brand (anonymized) ran a 120-day AEO program targeting AI Overviews and ChatGPT mention rate. Three workstreams — PDP rewrite, schema upgrades, and routine guide content — lifted brand mention rate from 5% to 18% in priority prompts and grew AI-referred sessions 22%. The playbook is reproducible across DTC verticals.

Background

The brand sells across skincare, body care, and wellness with ~60 active SKUs. Going into the program, baseline data showed:

• 5% AI mention rate in 200 priority prompts
• 11% AI mention rate among top 5 incumbent competitors (worst-of-pack)
• AI Overview presence: rare
• AI-referred sessions: ~1.4% of organic

Workstream 1 — PDP rewrite (weeks 1-6)

Problem: PDPs read like marketing copy, not extractable answers. Generative engines could not surface ingredient-level claims because no FAQ or comparison structure existed.

Action:

1. Rewrote 24 hero SKUs with an opening 40-60 word "What is" answer.
2. Added Q&A blocks for the top 5 buyer questions per SKU (~120 Q&As total).
3. Inserted a comparison block ("vs alternative formulations") on each PDP.

Outcome: AI mention rate on the rewritten 24 SKUs lifted from 6% to 14% within 30 days of crawl.

Workstream 2 — Schema upgrades (weeks 4-8)

Problem: No FAQ schema, generic Product schema, no Author or Organization markup.

Action:

1. Added FAQPage schema to every PDP with the new Q&A blocks.
2. Upgraded Product to include gtin, brand, aggregateRating, and review arrays from verified buyer reviews.
3. Added Person schema with sameAs to the brand's chief formulator.
4. Added Organization schema sitewide with sameAs to Wikidata, LinkedIn, and Crunchbase.

Outcome: Perplexity and ChatGPT citation rate lifted from 4% to 10% in priority prompts.

Workstream 3 — Routine guides (weeks 6-16)

Problem: No long-form expert content. AI engines defaulted to citing dermatologist publications, not the brand.

Action:

1. Shipped 12 routine guides (1,500-2,500 words each) authored by the chief formulator.
2. Each guide answered a high-volume "how do I" prompt and linked to relevant PDPs.
3. Each guide carried Article + Person schema, an AI summary block, TL;DR, FAQ, and a hub link.

Outcome: Routine guides earned 38% citation rate in their target prompts within 60 days. Cross-link traffic to PDPs grew 17%.

Results

What worked

• Answer-first PDP rewrites were the highest-leverage move — measurable lift inside 30 days.
• Author schema (chief formulator with sameAs) materially affected Perplexity citations.
• Routine guides authored under a credible Person built durable citation share.

What did not move the needle

• Adding ClaimReview schema without third-party fact-check sources — engines did not honor it.
• Submitting to llms.txt registries — too early in 2026 for ecommerce-side adoption signal.
• Aggressive internal linking from blog to PDPs — marginal lift only.

How to apply this playbook

1. Inventory PDPs and rank by current AI mention rate using a visibility tool.
2. Rewrite the top 20-40 PDPs with answer-first openings and FAQ blocks.
3. Add FAQPage, full Product, Organization, and Person schema in parallel.
4. Ship 8-12 expert-authored routine guides per quarter that link to relevant PDPs.
5. Measure citation share weekly; recalibrate at day 60 and day 120.

FAQ

Q: Why did mention rate move so fast?

DTC verticals have lower AEO competition than B2B SaaS. Most competing brands had not yet shipped FAQ schema or expert-authored long form, so the lift came largely from filling an empty signal surface.

Q: Did paid ads contribute?

No. The program ran without changes to paid budget. Lift is attributable to organic AEO work.

Q: How many SKUs do I need to rewrite?

Start with the top 20-40 (revenue-weighted). Diminishing returns set in beyond ~50 unless your catalog is highly diverse.

Q: What schema fields mattered most?

FAQPage, Product.brand, Person with sameAs, and Organization.sameAs. Most other fields are nice-to-have.

Q: Can a brand under $5M ARR run this playbook?

Yes — the playbook is content- and schema-led, not budget-led. The bottleneck is editorial throughput, not spend.