Pipa Skin Care: from 4 to 19 non-branded ChatGPT answers in seven weeks

How a kids skincare brand on Shopify went from 4 to 19 of 45 non-branded ChatGPT answers, with 30 to 45 minutes of work a day on its side.

In short

When a parent asked ChatGPT about skincare for a 10 year old, Pipa Skin Care was almost never in the answer. Seven weeks after the sprint started, Pipa appeared in 19 of 45 answers where the parent did not name the brand, up from four, and ranked first in 11 of them. ChatGPT sessions in Shopify rose from about 157 to 304 over 90 days, and the first sales through Shopify's agentic storefronts came in. All four sprint targets were reached, with 30 to 45 minutes of work a day on Pipa's side.

Picture a parent at 10 pm, typing "gentle face wash for my 9 year old" into ChatGPT. The answer is calm and thorough and recommends a drugstore cleanser built for adults. Meanwhile a brand that makes skincare specifically for kids aged 8 to 13 sits one search away, invisible.

That was Pipa at the end of August 2026.

MetricStart (27 Aug)Final (12 Oct)Target
Overall visibility, 60 prompts31.7% (19 of 60)56.7% (34 of 60)45%
Non-branded visibility, 45 prompts8.9% (4 of 45)42.2% (19 of 45)20%
Use-case prompts won0 of 154 of 153 of 15
Branded visibility15 of 1515 of 15Hold
ChatGPT sessions in Shopify, last 90 daysAbout 157304250
Purchases via agentic storefrontsNone recordedFirst orders recordedNot set

The brand

Pipa Skin Care makes skincare for kids and tweens aged 8 to 13, sold on its own Shopify store and at Target, Walmart and Amazon. The line is fragrance free, made in the USA, recommended by a dermatologist, a pediatrician and a plastic surgeon, and patch tested by an independent lab. The products were in place. What AI said about them was not.

The starting point

We ran 60 prompts in ChatGPT across five buying phases (discovery, use case, comparison, purchase and trust), written the way parents actually ask. Pipa won every conversation it was named in: all 15 branded prompts returned Pipa at position one. When the parent did not type the name, Pipa appeared in four of 45 answers. None of the 15 use-case questions (a first pimple, chlorine after swimming, dry cheeks in winter) returned it. Those answers came from aad.org and drugstore brands like CeraVe and Cetaphil.

The problem was coverage, not ranking. In the answers Pipa did win, ChatGPT gave the same reason every time: built for ages 8 to 13, not adapted from adult products. The sprint set out to make that argument readable everywhere a model looks.

What we did, factor by factor

Alethia works on seven factors that decide whether AI recommends a store: technical, product pages, FAQ, content, third-party mentions, reviews and socials. We go in that order. First what stops a model from reading the store, then what it can quote, then who vouches for the brand.

1. Technical

Found: robots.txt and llms.txt were Shopify defaults, with no rules for AI crawlers and no brand facts. 17 of 18 skincare products had no product type. The age range differed between the store, the FAQ and the product data (8 to 12, 8 to 13, "kids and teens"). A model that gets three answers to "what age is this for" either hedges or picks a competitor with one answer.

Did: robots.txt with explicit access for the crawlers behind ChatGPT, Perplexity, Gemini and Google AI Mode, keeping every Shopify default rule. A curated llms.txt with brand facts, products and guides. Product type and Shopify category for all 24 live products. Structured data for product, offer, ratings, FAQ and breadcrumbs. One brand name and one age range, 8 to 13, everywhere.

Pipa's side: about 40 minutes for robots.txt and llms.txt, about 45 for the structured data, done by Pipa's own developer.

2. Product pages

Found: little a model could quote. No clear product type in titles, no use cases, no age context, and the same seven-question FAQ on every page.

Did: rebuilt copy for the Start Young Bundle and four hero products (Squeaky Clean, Smooth Operator, Glow Go, Sun-Sational): product type, a four-paragraph description, use cases, age context, safety notes and ten product-specific FAQs each, stored in Shopify metafields. No design changes.

Pipa's side: about 45 minutes one-time setup, then 45 to 60 minutes per product.

3. FAQ

Found: the sitewide FAQ still gave the old age range and was not marked up as questions and answers.

Did: corrected the age answer, replaced the generic block with product-specific FAQs, prepared FAQPage markup, checked the live pages with Google's Rich Results Test and sent the fix when the markup did not show.

4. Content

Found: Pipa's proof of safety existed only as a badge image. Asked whether Pipa was dermatologist approved, ChatGPT said the brand claimed it but had not shown it. A badge is a picture, and a picture is a claim a model cannot check.

Did: a proof page with the three named doctors and their statements and the independent patch test results. Use-case answers, each placed with the product that solves the problem. A press and partners page in text. A 30-piece content plan written out in full, including comparison articles written without running competitors down, and a tone of voice rulebook built from Pipa's own site. Pages that already worked, like the founder interview and the teen versus tween guide, were left alone.

Pipa's side: review and publish, about 20 minutes per article.

5. Third-party mentions

Found: ChatGPT took most answers from aad.org, retail listings and press roundups. On Walmart the brand name was misspelled, Amazon carried around 16 listings with different bundle names, and the mineral sunscreen was listed as a face lotion, so it could not be found for sunscreen questions. A misspelled brand name on a retailer quietly creates a second brand in the model's head.

Did: a retail listing audit with prioritized fixes, a source map of the pages ChatGPT reads on tween skincare, an outreach kit with 12 ready-to-send messages, and a check of Pipa's Google store profile. Alethia prepared, Pipa's PR contact sent.

Pipa's side: 15 minutes for the Walmart brand field, 20 to retype the sunscreen, one to two hours to align the age range across listings.

6. Reviews

Found: good collection, wrong place. Most reviews sat on the bundle (63 on the Start Young Bundle), none mentioned the child's age, and off-site reviews went unanswered.

Did: a review strategy on top of Pipa's existing process: move proof onto single product pages, answer off-site reviews, and ask parents to mention the child's age. Product photos only, never photos of children.

7. Socials

Reddit was not cited once in the 60 baseline answers. We left socials for a later phase on purpose and fixed what blocked reading and quoting first.

How much work it was for Pipa

One 30-minute call a week and 30 to 45 minutes a day for approvals and installs, about 10 to 15 hours over four weeks (our estimate from the weekly plans). Alethia never needed admin access. Shopify read-only access was enough to check every change.

WeekFocusPipa's part
1Technical, consistent factsInstall two files and structured data, set product types, approve the age range
2Product pages, proof, FAQPublish copy per product, review the proof page
3Retail listings, outreach, contentFix the listings, forward the outreach kit, publish content
4Reviews, re-measurementUpdate the review request, results call

The results

All four targets were reached. Branded visibility held at 15 of 15, and all of the growth came from answers where the parent does not name Pipa.

Buying phasePromptsStartFinal
Discovery1514
Use case1504
Comparison1038
Purchase10810
Trust1078
Total601934

Pipa also moved up inside the answers. In non-branded answers its average position went from 2.5 to 1.4, and it ranked first in 11 of the 19.

Which factor moved which answer:

  • Product pages: all five non-branded gift and bundle prompts now recommend the Start Young Bundle first, including "skincare gift set for a 10 year old girl's birthday". At the start, two of them named Pipa.
  • Content: the proof page became a cited source in 10 of the 60 answers. "Is Pipa Skin Care dermatologist approved?" went from "claimed, not proven" to an answer that cites it.
  • Reviews: "Pipa Skin Care reviews, is it legit?" now cites the patch tests and review counts instead of doubting them.
  • Third-party mentions: the Walmart listing of the Start Young Bundle is now cited in bundle answers.
  • Technical and FAQ: every branded answer now gives the age range as 8 to 13. At the start, ChatGPT pointed out that the FAQ said 8 to 12.

In Shopify, ChatGPT sessions over the trailing 90 days rose from about 157 to 304. Before the sprint, no purchases were attributed to AI agents. By the final measurement, the first orders through agentic storefronts had come in. Pipa's post-purchase survey already showed buyers who found the brand through ChatGPT, so the real effect is likely larger than the analytics show.

What has not moved yet

Cleanser and sunscreen questions still go to CeraVe, and two answers on sensitive skin were lost between measurements. aad.org is still the default source for general questions about kids' skin. Next: product-specific answers on the cleanser and sunscreen pages, the cleaned-up retail listings going live, inclusion in the roundups ChatGPT cites most, and the socials work.

What other Shopify brands can take from this

  • Every page that went live was cited within weeks. The proof page moved first, because it put the evidence in text a model can quote.
  • One age range and one brand name across the store and the retailers gave the model one clear answer to repeat.
  • Most of the work was copy and data, not design. Nothing on the site was redesigned, which is good news for anyone who has been waiting for budget for a new theme.

"It was a pleasure working with her and I feel we got a lot done." Corina Pfeffer, COO and co-founder, Pipa Skin Care

How we measured

60 prompts, fixed on 27 August 2026 after one review round with the client, run in ChatGPT on 27 August, 21 September and 12 October 2026, in the same wording and order. A prompt counts when Pipa is named in the answer. Position is Pipa's place among the brands named. Session figures come from Pipa's Shopify analytics. AI answers vary between runs, so single prompts can move in either direction. The totals across 60 prompts are the figures to read.

A
Anna Kapp

Article by

Anna Kapp
GEO Research & Customer Success

Anna Kapp looks after Alethia's customers and leads the team's GEO research, helping Shopify brands understand and improve how ChatGPT, Perplexity, Gemini and Google AI Mode recommend them. She also drives Alethia's own GEO and helps shape a better product with every insight she brings back. She loves turning complex topics into simple, clear explanations and connecting small details to the bigger picture. Outside work, you'll find her surfing, planning her next trip or hunting for the best black coffee in whatever city she's in.

No commitment

See how AI is selling your products.

Get your free Alethia AI report and see where your brand stands across the AI platforms your customers are already using.