Auto-labelling
Manual tagging can’t keep pace with your range.
GEMA adds physical and contextual attributes at scale, connected by relationships rather than left as loose tags.
Drop shoulderCold snap
+12 more on this product
Manage the meaning in your catalogue. Use it everywhere.
GEMA enriches every product with meaningful data, so search, recommendations and AI answers can be as relevant as your best merchandiser.
We run a sample of your own products through GEMA before the call.
Crew neck
Drop shoulder
Chunky rib knit
Relaxed fit
Cold snap
Weekend
Quiet luxury
Wide-leg trousers
Leather tote
“cosy jumper for autumn”
Chunky knit jumper
Illustrative products. Physical and contextual attributes, related products and shopper language.
Trusted by 750+ brands worldwide
One source of meaning behind every place shoppers meet your products.
Today each tool tags products its own way. With GEMA, search, product pages, feeds, AI answers and CRM all draw on the same understanding of your range.
Search and merchandising
Shoppers find “something cosy for a cold weekend” even when no product title says it.
Search and category pages understand occasion, style and fit, and merchandising rules apply to meaning, not keywords.
TODAY
“cosy jumper for a cold weekend”
No results for “cosy jumper for a cold weekend”. Showing results for jumper.
New
Keyword match only. Anything called a jumper shows up; “cosy” and “cold weekend” mean nothing.
WITH GEMA
“cosy jumper for a cold weekend”
GEMA understoodWeather: cold snapOccasion: weekendChunky knit“cosy”
24 resultsRanked by occasion, weather and warmth
Cold snap
Cosy
Weekend
Also matchedAnkle bootsWool coat
Matched on occasion and weather, not just keywords.
+2xincrease in searchable attributes, from 17 to 32 distinct attributes.
Product pages
Richer, on-brand product detail on every page.
Richer product descriptions, complete-the-look and personalised recommendations. Dynamic product detail pages drawn from the same graph, in your tone of voice.
TODAY
Home › Knitwear
Jumper
£89
Colour: Brown
SMLXL
Add to bag- Description
- Knitted jumper in a wool blend.
- Size and fit
- Regular fit. Model wears M.
- Care
- Hand wash cold. Dry flat.
The basics are there, but no context: how it wears, when to wear it, what goes with it.
WITH GEMA
Knitwear › Jumpers › Chunky knit
Chunky knit jumper
£89
Chunky rib-knit crew neck for cold snaps and weekends away.
+15 attributes from GEMA
Relaxed fitChunky rib knitWool blendCold snapWeekend away“cosy jumper”
Style this look
3xmore distinct attributes to power your product page experiences.
Feeds and answer engines
Show up in AI answers as the right product, described the right way.
Feeds and answer engines get descriptions grounded in the occasions, styles and words shoppers actually use, not just the four fields in your feed.
TODAY
“warm jumper for a cold weekend away”
AI answer
Brown knitwear jumper in a wool blend. Available in sizes S–XL.
Source: product feed
SHOPPING FEED4 fields
- title
- Jumper
- product_type
- Knitwear
- colour
- Brown
- material
- Wool blend
Four feed fields. Generic, easy to pass over.
WITH GEMA
“warm jumper for a cold weekend away”
AI answerTop pick
Chunky knit jumper · £89
A chunky rib-knit crew neck with a relaxed fit, made for cold snaps and weekends away.
yourbrand.com+ 2 more of your products cited
SHOPPING FEED4 → 19 fields
- title
- Chunky rib-knit jumper, relaxed fit
- product_type
- Knitwear › Jumpers › Chunky knit
- occasion
- Weekend away · cold snapNEW
- search_terms
- chunky knit, cosy jumperNEW
Grounded in 19 attributes and relationships in your graph.
Increase ROAS, citations and share of authority.
Landing pages and editorial
Build edits like “the cold-weekend edit” from meaning, not manual SKU lists.
Curated pages stay current as the range changes, because they are defined by attributes and rules.
TODAY
Home › Women › Knitwear › New in
Autumn editNew in: KnitwearShop all knitwear
Edited manually · 3 weeks agoSold out
Built by hand from SKU lists, out of date by next week.
WITH GEMA
Built fromWeather: cold snapOccasion: weekend
Across categories, and it updates itself as the range changes.
Shoppable landing pages and edits, generated automatically from your catalogue.
Recommendations
Recommendations that understand intent and sell across categories.
GEMA reads why a shopper is browsing, the occasion, style and need, and pairs products across categories: trousers, a tote and boots for a jumper, not three more jumpers.
TODAY
You’re viewingChunky knit jumper · £89
Customers also viewedRanked by popularity
Bestseller
More from the same category. Nothing to wear it with.
WITH GEMA
You’re viewingChunky knit jumper · £89
IntentWeekend awayCold snapRelaxed style
Complete your weekendBy occasion, style, colour
Trousers
Bags
Footwear
Complete the look £328Add all 3 to bag
Cross-category, matched on intent, not just what others bought.
+5%increase in conversion.
Outfits and Shop the Look
Complete outfits at range scale, built on your own styling rules.
Compatibility comes from your authored rules and expertise, applied across the whole catalogue.
TODAY
Chunky knit jumper£89 · Brown, wool blend
Add to bagOne product, one photo. Shoppers style it themselves, or leave.
WITH GEMA
Shop the look
Styling ruleBurgundy + ivoryWorkwearAutumn
12341
Trench coat£189 · Anchors the look
2
Ribbed turtleneck£45 · Layers underneath
3
Wide-leg trousers£69 · Balances the coat
4
Ankle boots£139 · Rain-ready
Look total · 4 items£442
Add the look to bagOutfits built from your own styling rules, for every product.
3xhigher conversion on outfits.
Email and CRM
Personalise by the styles and occasions each customer cares about.
Segments and content draw on product meaning, not just past purchases.
TODAY
Your Brandto all subscribers · 48,210
Thu 09:00New arrivals this week
New inJust landedShop now
Same email for everyone.
WITH GEMA
From: Your Brand · To: Sam
Triggered by cold-snap forecastSam, your cold-weekend edit
Relaxed fit
Weekend
Cold snap
Because Sam likesRelaxed fitWeekend“chunky knits”
Built from the styles and occasions each customer engages with.
+34%click-through rate.
Returns and size intelligence
Returns Intelligence uses a retailer’s own numbers to show what returns cost and why they happen.
Replaces assumption-led size ratios with demand forecasts from product, sales and attribute data, reducing fragmented stock, markdowns and margin leakage.
RETURNS INTELLIGENCEWhat drives returnsPutting the two togetherConfidence ceilingLast 12 months · vs prior year
Every return, by what the customer did when buying
SHARE OF RETURNS
Style only
39.1%bought two or more different garments of the same typeBoth
10.1%compared styles and bought several sizes in the same basketSize only
14.5%bought the same garment in two or more sizesOut of reach
36.4%one garment in one size, nothing to compare it withStyle is involved in 49.2% of returns and size in 24.6%.
30–40%of online fashion orders come back.
Feed it what you know. GEMA turns it into meaning. Every surface uses it.
Product data says what an item is. Your team’s knowledge and your shoppers’ language say what it means. GEMA brings them together.
1 · FEED AND TRAIN THE GRAPH
Data
Product feedWebsiteConsumer languageMarket trendsCustomer dataKnowledge
Domain Knowledge PacksMerchandising rulesMarketing calendarTone of voice2 · GEMA MANAGES MEANING
Your brand ontology
Built on Domain Knowledge Packs, controlled by you
Auto-labelling
Physical and contextual attributes, at range scale
Product and graph view
Every attribute and relationship, in one place
Brand and merchandising rules
Set once, applied on every surface
Review and approval
AI makes the first pass, your team signs off
3 · YOUR BRAND KNOWLEDGE GRAPH
Chunky knit jumper19 connections · brand graphis aKnitwear › Jumpersworn forCold snappairs withLeather toteshoppers say“chunky knit”Say “jumper”, never “sweater”Brand ontologyv7 · publishedOne connected source of meaning, owned by your brand.
4 · EVERY SURFACE
- Search and merchandising
- Product pages
- Feeds and answer engines
- Landing pages and editorial
- Recommendations
- Outfits and Shop the Look
- Email and CRM
- Returns and size intelligence
Start from expert depth, not a blank sheet.
Domain Knowledge Packs are ready-made sets of attributes, structure and shopper language for a retail category. You shape them into your own brand ontology.
Fashion
InstalledReady-made category expertise for womenswear, menswear and footwear, from dresses and tailoring to knitwear, outerwear and boots.
Womenswear6 categoriesMenswear5 categoriesFootwear4 categories
15 categories30 attribute values10 shopper phrases
PHYSICAL ATTRIBUTES15 families
NecklineSleeve lengthSilhouetteFabricFitCollarLeg shapeHeel heightToe shape+6 more
NECKLINE9 values
CrewRoll neckV-neckSquareBoatCowl+3 more
CONTEXTUAL ATTRIBUTES7 families
OccasionStyleSeasonMoodWeatherDress codeComfort
OCCASION9 values
EveningWorkWedding guestWeekend awayHoliday+4 more
ONTOLOGY SLICE
- Fashion
- Womenswear
- Knitwear
- Jumpers
- Chunky knit
Shape it into your brand
SHOPPERS SAY
“going-out top”
maps to
Occasion: eveningCategory: topsStyle: statement
TAILORED TO YOUR BRAND
sweaterjumperpantstrouserssneakerstrainers
YOUR VERSION
- Pack installedDay 1
- 24 terms tailoredWeek 2
- v7 publishedLive everywhere
Beauty
InstalledReady-made category expertise for skincare, make-up, haircare and fragrance.
Skincare5 categoriesMake-up5 categoriesHaircare3 categoriesFragrance2 categories
15 categories30 attribute values10 shopper phrases
PHYSICAL ATTRIBUTES11 families
Skin typeFinishShadeFormulationCoverageSizeHair typeScent family+3 more
FINISH7 values
MatteSatinDewyGlossyShimmer+2 more
CONTEXTUAL ATTRIBUTES6 families
Routine stepConcernSeasonOccasionLifestyleValues
CONCERN8 values
DrynessShineRednessFine linesDark spots+3 more
ONTOLOGY SLICE
- Beauty
- Skincare
- Moisturisers
- Night creams
Shape it into your brand
SHOPPERS SAY
“dewy no-make-up look”
maps to
Finish: dewyCoverage: lightOccasion: everyday
TAILORED TO YOUR BRAND
moisturizermoisturiserlipsticklip colouranti-agingskin renewal
YOUR VERSION
- Pack installedDay 1
- 24 terms tailoredWeek 2
- v7 publishedLive everywhere
Home
InstalledReady-made category expertise for textiles, furniture, lighting and decor.
Furniture5 categoriesTextiles4 categoriesLighting3 categoriesDecor4 categories
16 categories30 attribute values10 shopper phrases
PHYSICAL ATTRIBUTES11 families
MaterialDimensionsColour familyFinishPatternSeat depthWeaveCare+3 more
MATERIAL10 values
BoucléLinenVelvetWoolOakRattan+4 more
CONTEXTUAL ATTRIBUTES6 families
RoomStyleSeasonMoodLifestyleHosting
ROOM7 values
Living roomBedroomDiningHome officeHallway+2 more
ONTOLOGY SLICE
- Home
- Furniture
- Sofas
- Corner sofas
Shape it into your brand
SHOPPERS SAY
“cosy living room”
maps to
Room: living roomStyle: warm minimalMaterial: wool
TAILORED TO YOUR BRAND
couchsofacomforterduvetthrow blanketthrow
YOUR VERSION
- Pack installedDay 1
- 24 terms tailoredWeek 2
- v7 publishedLive everywhere
A product record says what it is. A graph says what it means.
A brand knowledge graph is a connected map of every product in your catalogue: what it is, what it’s for, how it relates to the rest of your range and how shoppers describe it.
- 01Flat product record
- 02Physical attributes
- 03Contextual attributes
- 04Relationships and shopper language
pairs withworn forsearched as
IN GEMAChunky knit jumperKnitwear · Brown · Wool blend
- Physical attribute
- Contextual attribute
- Related product
- Shopper language
View as a list: 19 nodes, 30 relationships
PHYSICAL
- Crew neck
- Drop shoulder
- Chunky rib knit
- Relaxed fit
- Wool blend
- Chocolate brown
CONTEXTUAL
- Cold snap
- Weekend
- Quiet luxury
- Layering piece
- Autumn
RELATED PRODUCTS
- Wide-leg trousers
- Leather tote
- Ankle boots
- Ribbed tee
SHOPPER LANGUAGE
- “chunky knit”
- “oversized jumper”
- “cosy autumn knit”
- “weekend away outfit”
Want to see your own products as a graph?
We run a sample of your catalogue through GEMA before the demo.
AI does the heavy lifting. Your team stays in control.
Labels alone aren’t enough. GEMA connects them with relationships and rules, and your team approves what goes live.
Your brand ontology
Generic taxonomies don’t speak your brand.
Built on Domain Knowledge Packs and shaped to your vocabulary. Controlled by you, for your brand.
KnitwearJumpersChunky knit
Product and graph view
You can’t manage what you can’t see.
See every attribute and relationship, the way search, recommendations and AI see your catalogue.
Wide-leg trousers
12 attributes · 6 relationshipsBrand and merchandising rules
Every tool gets set up its own way.
Set trading and brand intent once and apply it on every surface, instead of in every tool.
BRAND RULESay “wide-leg”, never “palazzo”
Review and approval
AI alone isn’t enough for your brand.
AI makes the first pass. Your team reviews and approves, with a full audit trail.
Merch review3/4 approved
Questions retailers ask
Everything you need to know about GEMA before you talk to us. Can’t find your answer?
Contact usA brand knowledge graph is a connected map of every product in your catalogue: what it is, what it’s for, how it relates to other products and how shoppers describe it, organised in your brand’s own vocabulary.
Domain Knowledge Packs are ready-made sets of attributes, structure and shopper language for a retail category, such as womenswear or beauty. You start from expert depth and shape them into your own brand ontology.
A PIM stores product records. GEMA adds meaning on top: contextual attributes, relationships and rules, managed by your team and used on every surface.
Faster than you’d think. GEMA works from the data you already have: your product feed, catalogue and attributes. We connect to it, enrich it, your team reviews, makes refinements where needed and approves. Typical time to first enriched graph is 4 weeks, with full rollout depending on what’s in scope.
No. GEMA is complementary to the tools you already use. It enhances your search, recommendations, feeds and campaigns by giving them a richer understanding of every product, so you get more from your existing stack without replacing it.
See what your website can look like with GEMA applied.
Book a demo and we’ll run a sample of your products through GEMA before the call.
WHAT YOU’LL SEE
- Your products enriched by GEMA
- Gaps in your current product data
- Where richer meaning would lift search, recommendations and answer-engine visibility
WHAT HAPPENS NEXT
- Submit the form
- Pick a time on the confirmation page
- We pull sample products and prepare your audit
- We walk through your own products in GEMA
