Natural Touch: Turning a Large Beauty Catalogue Into AI-Readable Product Knowledge
A shopper asking which hair oil fights frizz, or whether argan or jojoba oil is better for damaged hair, isn't going to read twenty product descriptions to find out. Natural Touch's catalogue, spanning hair, skin, body, and fragrance, already had the right products. It didn't yet have the structure to tell an AI system which one to recommend, and why.
Why beauty shopping is increasingly a conversation, not a browse.
Natural Touch sells across hair, skin, body, fragrance, and men's care, a catalogue built from real, genuinely differentiated products: Argan, Amla, and Jojoba hair oils, a Super Repair treatment line, curl-specific care, dozens of body butters and shower gels across named scent collections, and a full range of eau de parfums, perfume oils, and solid perfumes. That's a large volume of exactly the kind of question beauty shoppers increasingly ask AI tools directly: which oil is best for dry hair, argan versus jojoba for damaged hair, what to use for very dry skin. The catalogue had real answers to all of it. The site wasn't yet built to hand those answers to an AI system in a form it could use.
Beauty shoppers rarely start with a product name. They start with a concern or a comparison: what's the best oil for dry hair, argan oil vs. jojoba oil for damaged hair, what body care actually works for very dry skin, or more ambitiously, build me a full haircare routine for dry, damaged hair. Each of those questions requires reasoning across concern, ingredient, hair or skin type, and intended use, exactly the kind of structured comparison an AI assistant is built to do, provided the underlying facts are stated explicitly somewhere it can find them.
Before: a big, good catalogue that was hard to tell apart.
Natural Touch's range is large by design: dozens of body butters and shower gels across named scent collections, a full hair oil lineup, multiple curl- and repair-focused hair care lines, and an extensive fragrance range split across EDPs, perfume oils, and solid formats. That's a genuine strength for a shopper who already knows what they want. It's a real obstacle for one who doesn't, and increasingly, for the AI tool trying to narrow the field on their behalf.
Product copy across the catalogue was descriptive, evocative scent and texture language, premium positioning, real ingredients named, but not consistently structured around the specific dimensions a comparison question actually needs: which skin or hair concern a product targets, which hair or skin type it's suited for, how it's meant to be used, and who it isn't the right fit for. Two genuinely different hair oils, say Argan and Jojoba, could both read as "good for hair" without a shopper, or a model, being able to tell which one actually fits a dry, damaged-hair concern versus a shine-and-frizz concern.
The GEO work: structure, routines, and comparisons AI can use.
We started by standardizing a consistent set of attributes across the catalogue, active ingredients, the specific concern each product targets, hair or skin type suitability, and intended usage, so the same structured facts exist for every product rather than varying by however each listing happened to be written. On top of that structure, we built explicit "best for" and "not ideal for" language directly into product content, so a product doesn't just describe itself, it states plainly who it's actually right for.
We built routine-based content, a dry-hair routine built from specific products in sequence rather than a single listing in isolation, and genuine comparison pages organized around real decisions shoppers are making, including an Argan vs. Jojoba oil comparison built around damaged hair specifically. We implemented richer product structured data across the catalogue, built FAQ content directly from the real questions customers ask, and made sure the same core product facts stayed consistent everywhere they appeared on the site. It's the same AI-led discovery foundation behind the AI-powered product discovery platform we built for Natural Touch, extended from the shopping experience itself into the content that feeds it.
“We've always known which oil was right for which hair type. The gap was that a customer, or an AI tool shopping on their behalf, had no reliable way to find that out without asking us directly.”
- Natural Touch
What This Project Proved
A large catalogue is a GEO advantage, not a liability, once it's structured
More genuinely differentiated products means more specific comparisons an AI system can make, as long as the differentiation is explicit rather than implied.
"Good for hair" isn't a usable attribute. "Best for dry, damaged hair" is.
Specific, concern-based language is what lets a model distinguish between two similar products confidently.
Routine content does work comparison pages can't
A single product answers one question. A routine, what to use, in what order, for a specific concern, answers the fuller question a shopper, or an AI building a recommendation, is actually asking.
Consistency across the site matters as much as the content itself
The same product facts need to match wherever they appear, category pages, product pages, blog content, or they read as unreliable rather than authoritative.
Is Your Catalogue Big Enough to Compete, But Structured in a Way AI Can Actually Compare It?
SnapTec helps beauty, personal care, and other comparison-heavy catalogues restructure product content so AI systems can confidently recommend the right product for the right need.
Talk to Our GEO Team