BuDhaGirl
The stack was already designed in.
The recommendations just needed to understand it.
How BuDhaGirl gave shoppers a rec gallery that understood finish relationships - and turned single-set browsers into stack-builders.
→ How completion logic replaced similarity logic in BuDhaGirl's rec gallery
RecoMelon installs in under a day. First influenced orders in 60min.
No code, no engineering, no catalogue prep.
Message our team on WhatsApp to see how RecoMelon fits your store. Or, run a free PDP audit - see your rec relevance score.
- The situation
A catalogue built for stacking. A recommendation module with no concept of completion.
BuDhaGirl is not a jewellery brand that happens to sell bangles. It is a stacking brand. Every product in the catalogue is designed to be worn alongside something else - nine bangles in a set, colourways built to layer, finish families that co-ordinate across collections. The All Weather Bangle is the core, but the core is always the beginning of a stack, not the end of a purchase.
The purchasing behaviour BuDhaGirl is built to unlock is inherently multi-item. A shopper who finds the Gold Serenity Prayer AWB is not necessarily done. She is looking for the Champagne that completes her range, the Two-tone that gives her contrast, the colour that matches her intention for the day. The brand has encoded that logic into its product design for years.
What it hadn't yet encoded was the discovery experience. The recommendation module that sat on BuDhaGirl PDPs understood what was popular. It did not understand what completed a stack. It could show another gold set to a shopper who was already choosing gold. It could not show her the piece that made her gold set more versatile - the complementary finish, the co-ordinating colourway, the contrast piece that makes the stack worth building.
"The catalogue was designed for completion. The recommendation engine had no concept of it."
The BuDhaGirl stack-building journey - before RecoMelon
Shopper arrives with intention
She knows BuDhaGirl. She has a colourway, a finish, a set in mind. She is already thinking in stacks - gold with champagne, Two-tone with solid.
Finds the anchor piece
Gold AWB. Serenity Prayer. The set that anchors her stack. She is on the PDP. She is ready to build.
Recommendation module shows her more of the same
Another gold set. Popularity-ranked. Technically related. Commercially useless - she already has gold in her basket.
Completion opportunity missedNo contrast, no complement, no co-ordinate
The Champagne that would complete her range. The Two-tone that gives her contrast. The colour piece that matches her mood for the day. None surfaced.
Single-set basket
She buys what she came for. The stack she was building in her head stays there. AOV stays at native.
Stack left half-built
- The category dynamics
Colour, finish and pattern are not attributes. For BuDhaGirl shoppers, they are the purchase logic.
Most jewellery recommendation challenges are about relevance - show things that match the aesthetic, avoid jarring mismatches. BuDhaGirl's challenge is more specific. It is about complementarity. The shopper is not asking "show me something similar." She is asking "show me what completes this."
That distinction matters because the answer to "show me something similar" and "show me what completes this" can be the exact same product - or the opposite one. A Gold AWB shopper needs to see Champagne and Two-tone, not more Gold. A solid colourway shopper building a textured stack needs the contrast piece, not another solid. A single-finish purchase is a category missed, not a category matched.
The Dealbreaker™ data from BuDhaGirl's Shop Similar gallery confirms this precisely. The highest-frequency filter interactions are Two-tone, Multi-Color and Gold Leaf - not individual colours, but finish and combination types. BuDhaGirl shoppers navigate by combination logic. They are actively telling the gallery what kind of stack they are building. The rec module's job is to understand that and respond with the pieces that complete it.
Completion, not similarity. A shopper with Gold in her basket doesn't need more Gold shown to her. She needs Champagne, Two-tone, the contrast that makes the stack worth wearing. The rec module's job is to complete, not mirror.
Finish families are the purchase signal. BuDhaGirl shoppers filter by Two-tone, Multi-Color, Gold Leaf - finish combinations, not individual colours. RecoMelon reads finish relationships across the catalogue automatically.
Pattern is a distinct axis. The Pattern Pulse data - abstract, solid, stripe - shows shoppers navigating by pattern type as a second dimension. A solid-to-solid stack reads differently to a solid-with-stripe stack. Both are valid. The rec module needs to know which one the shopper is building.
Mindful Glamour® is a multi-item concept. BuDhaGirl's brand idea - turning routines into rituals - is inherently about layering intention. Nine bangles per set, each with a purpose. The multi-item basket is not upsell. It is the product working as designed.
- The intervention
Completion logic. Finish intelligence. Dealbreaker™ filters that speak the language of stacking.
RecoMelon mapped BuDhaGirl's catalogue visually - by finish family, colourway relationship, pattern type and combination logic. Not by tag, not by price band, not by popularity rank. By what the products actually look like next to each other on a wrist.
A Gold AWB on a PDP surfaces Champagne, Two-tone and complementary colourways - the pieces that complete the range, not the pieces that duplicate it. A textured finish surfaces smooth complements. A bold pattern surfaces the solid that lets it lead. The gallery becomes a stack-building tool, not a "more like this" carousel.
Dealbreaker™ filters gave shoppers the elimination layer to build with precision. Two-tone or single-tone. Abstract or solid. Gold Leaf or plain. One interaction filters the gallery to the finish family the shopper is working with - letting her build the stack she has in mind rather than scrolling through options that don't help.
The result is a rec gallery that understands BuDhaGirl's commercial model: the basket is always a stack in progress, and the rec module's job is to complete it.
How RecoMelon works for BuDhaGirl
Finish and colourway mapping
Every product mapped by finish family (Gold Leaf, Champagne, Two-tone, Multi-Color), colourway and pattern type. Relationships built from product images, not tags.
Fully automatedCompletion logic, not similarity logic
Gold surfaces Champagne and contrast. Solid surfaces textured and stripe. Abstract surfaces the plain that lets it lead. The gallery shows what completes the stack, not what duplicates it.
Dealbreaker™ filters for stack-building precision
Two-tone or single. Abstract, solid or stripe. Gold Leaf or plain. One filter interaction narrows the gallery to the finish family the shopper is working with.
The stacking vocabularyMore recs per session as the stack builds
Average recs per session doubled - from 2 to 3. Shoppers are not clicking one rec and buying. They are using the gallery to explore the stack, piece by piece.
Stack exploration in action
- The outcome
+19.3% AOV. 223 influenced orders. And 55,590 rec clicks that show exactly how shoppers build stacks.
These are the signals that matter - not just order counts, but the behavioural evidence that completion logic changes how BuDhaGirl shoppers engage with the catalogue.
- Rec clicks55,590vs 14,652 prior period - +279% growth
- Influenced orders223vs 137 prior period - +62.8% growth
- Recs per session3vs 2 prior period - +50% - stack exploration in action
- Gallery impressions39,288vs 26,622 prior period - +47.6% growth
Executive interpretation
The 55,590 rec clicks - a +279% increase on the prior period - is not a modest engagement lift. It is the sound of shoppers using the gallery as a stack-building tool. They are not clicking one rec and buying. They are navigating across finish families, exploring combinations, building the stack they had in mind when they arrived. The 3 recs per session (up from 2) is the mechanical proof of that behaviour. The +19.3% AOV is what it costs when those explorations end in a basket rather than a tab close.
- Dealbreaker™ intelligence
The filters tell you exactly how BuDhaGirl shoppers think about stacking.
141 Dealbreaker™ filter interactions in the reporting period. The top terms - Two-tone, Multi-Color, Gold Leaf, Waterproof, Stackable - are not colour filters. They are finish and combination filters. BuDhaGirl shoppers do not navigate the catalogue by colour alone. They navigate by finish relationship: what kind of stack am I building, and what type of piece do I need next?
This is precise signal. A shopper filtering for Two-tone is telling you she has a single-finish piece in her basket and is looking for the contrast. A shopper filtering for Gold Leaf is building in a specific finish family and wants to know what else belongs there. Dealbreaker™ captures those navigation decisions and returns exactly the pieces that match - not the most popular products, but the most relevant ones for the stack being built.
Dealbreaker™ top filter terms - BuDhaGirl
141 filter interactions. The dominant terms are finish combinations and stack types - not individual colours. BuDhaGirl shoppers navigate by how pieces relate to each other, not by what they look like in isolation.
Dealbreaker™ colour crush - BuDhaGirl
No dominant colour. The top 5 clicked colours are essentially tied at 2.0-2.2% each - spread evenly across the palette. BuDhaGirl shoppers are not mono-colour buyers. They are building across the full range.
The even spread across the palette is itself a finding. BuDhaGirl shoppers are not single-colour collectors. They are building ranges - which means the rec gallery's job is not to match the colour they chose, but to show them what sits alongside it across the full palette.
- Strategic takeaway
The stack was already designed in. The recommendations just needed to understand it.
BuDhaGirl's commercial opportunity was not hidden. It was visible in every product line - nine bangles to a set, colourways built to layer, the entire brand philosophy of Mindful Glamour® built around the idea that what you wear every day should be intentional, layered, personal.
What was missing was the mechanism to translate that into a purchase session. A shopper who arrives knowing what a BuDhaGirl stack is, and wanting to build one, needed a gallery that understood finish relationships. Without it, she left with one set. With it, she builds the stack she came to find.
The +279% rec click growth is not a sign that the old module was failing. It is a sign that shoppers were always willing to explore - they just had nothing worth exploring. Once the gallery showed them completions rather than duplications, they used it.
The question for any stacking brand is not whether your customers understand the product. They do. The question is: does your discovery experience understand the stack they are trying to build - and does it give them what they need to finish it?
- Implementation
No tagging. No manual curation. The finish relationships were already in the products.
The completion logic, the Dealbreaker™ finish filters, the colourway and pattern mapping - none of it required the BuDhaGirl team to tag products, write rules or curate galleries. RecoMelon's visual mapping read the finish families, the colourway relationships and the combination logic from the product imagery alone. The intelligence was already in the catalogue. RecoMelon surfaced it.
“RecoMelon understood our catalogue the way our customers do - as a set of pieces waiting to be stacked together.”
- +19.3% AOV - $186.18 rec-influenced vs $156.11 native
- 55,590 rec clicks vs 14,652 prior period - +279%
- 223 influenced orders vs 137 prior period - +62.8%
- 39,288 gallery impressions vs 26,622 prior period - +47.6%
- 3 recs per session vs 2 prior period - +50%
- 141 Dealbreaker™ filter interactions - led by Two-tone, Multi-Color, Gold Leaf
- 643 unique products with recs generated - comprehensive catalogue coverage
- Live in under one day - no tagging, no engineering, no catalogue prep
View more case studies
Shopify brands use RecoMelon to fix bounce, grow AOV, and turn PDPs into high-converting discovery tools. Here’s how.
