Navigation Built for Discovery: IA Research & Funnel Strategy at The Sage Lifestyle
The Challenge
A boutique fragrance brand was investing in paid acquisition while users abandoned the site before viewing a single product. The navigation structure was the silent drain on marketing ROI.
The Outcome
Phase 1 IA restructuring sustained funnel performance through the off-season. Directional results included a 69.6% increase in sample discovery and a lift in add-to-cart rates, compared against the prior peak season benchmark.
Context
The Sage Lifestyle was founded in 1998 by jewelry designer Sage Machado. The brand sells vegan, cruelty-free fragrance products inspired by the personalities of gemstones, sold online and through select retailers. The engagement started as an SEO project. The data revealed a different problem entirely.
I worked as the embedded UX researcher and strategist: a three to four week research window running alongside concurrent Shopify SEO work, with budget constraints shaping every methodological decision toward highest-impact methods. Research outputs doubled as strategic handoff documentation. Implementation was scoped to a separate development partner, so findings needed to be immediately actionable without me in the room.
Who was arriving
72% of holiday-season visitors were new to the site, so the navigation had to work for people with no prior knowledge of the brand.
Most holiday visitors were seeing the site for the first time.
Problem
North Star
The core objective was straightforward: protect the marketing investment by removing the friction between paid acquisition and purchase. The north star was a “Guided Discovery” engine, navigation that matched the site's high-end brand narrative and created a clear path from first visit to loyal customer.
The Sage Lifestyle was running paid acquisition across Google Ads, TikTok, and Klaviyo, spending to bring users to the site. GA4 purchase journey data (Dec 16, 2024–Jan 12, 2025) revealed what happened after they arrived:
- 53.6% of sessions ended before a user viewed a single product
- 84% of users who reached a product page abandoned before adding to cart
The SEO audit I was originally engaged for confirmed the misdiagnosis: 862 indexed keywords, 984 estimated monthly organic traffic, $1.4K in equivalent traffic cost. Organic search presence wasn't the problem. The leak happened after users arrived.
Peak season baseline (Dec 16 to Jan 12). 84% of product viewers left before adding to cart.
Research-to-Strategy
Competitive Audit
I audited six fragrance brands to map navigation patterns and identify gaps. Three patterns emerged across the competitive set: filter navigation organized by collection, function, and scent family; scent family categorization models; and icon-based scent notes at the product level. None existed in Sage's current IA.
Sage sits near the center, leaning slightly niche and character-driven.
Usability Testing
Hotjar session replays showed users looping through the menu repeatedly before leaving: a behavioral signature of navigation frustration, not low intent. The search bar was being used as a substitute for navigation, with users searching for scent notes rather than browsing the category structure.
Seven participants (Dec 18 to Jan 2): 2 hybrid card sorting, 2 desktop tree testing, 3 mobile tree testing.
Most visitors never scrolled to the discovery links.
What the research found
Across card sorting, tree testing, and session replays, three patterns repeated.
01
Users preferred intent-based filtering over technical taxonomy.
Card sorting participants grouped products by fragrance type and format rather than the gemstone-scent model. Hotjar confirmed users were searching for scent notes rather than using existing filters, which were limited to availability and price.
02
Unclear labels and overwhelming choices caused abandonment.
Tree testing revealed users struggled to locate basic product information through the navigation. Session replays showed most clicks concentrated on the back button and menu. Informational pages (Shop by Gemstone Scent, Shop Scent Categories, Shop Scent Styles) showed reduced time on page.
03
The navigation mixed purchase paths with informational content, no hierarchy.
14+ items in a flat list. “Discover All Gemstone Scents” and “Shop by Gemstone Scent” pointed to the same destination under different labels. Sampling options were buried equally with product links. No clear entry point for a first-time visitor.
Key Decision
Grouping by User Intent, Not Site Architecture
The existing navigation reflected the site's back-end logic: product types, gemstone collections, informational pages, and brand content at the same level with no hierarchy. Card sorting and tree testing consistently showed users approached the site with one of three intents: to shop, to discover which scent suited them, or to sample before committing.
The proposed navigation restructured around those three intents: Shop, Find Your Scent, Try Before You Buy, collapsing redundant gemstone links and surfacing sample kits as a primary entry point rather than burying them in the product catalog.
A flat 14-item menu became three groups: shop, find your scent, and try before you buy.
Filters added
Three filters followed the new navigation: fragrance type (Woody, Floral, Fresh), format (sample, full size, travel), and scent notes (Vanilla, Sandalwood, Amber). Each lets shoppers narrow by how they already think about scent.
This decision was shaped by project constraints: with a short research window and implementation scoped to a Shopify development partner, the solution needed to be achievable through label and structure changes rather than platform-level development. Intent-based grouping could be implemented without custom Shopify development, the highest-ROI intervention within scope.
I synthesized the full research findings into a strategy and recommendations deck presented to the client, prioritizing Phase 1 quick-wins and scoping the Phase 2 development investment against validated user needs.
Outcomes
Phase 1 launched as the brand moved from its holiday peak into the slow season. Fewer visitors were expected. The real test was whether the visitors who did arrive could find products and commit to them more easily.
Sample discovery views
+69.6%
The clearest sign that intent-based labels were routing shoppers toward low-risk trial products
Add-to-cart rate from product views
17.6% → 20.6%
Checkout starts rose from 50.1% to 53.0% of carts
Purchase completion held
57.9%
Mobile drove 65.6% of purchases
What the data showed
Results from January to April were compared against the prior peak season, October to December. Because the windows differ in length and season, step rates are the fairest measure.
Discovery shifted toward trial. Sample discovery views rose 69.6%, a sign that intent-based labels were routing first-time shoppers to low-risk entry points.
Shoppers who reached a product converted better. Add-to-cart rose from 17.6% to 20.6%, and checkout starts from 50.1% to 53.0% of carts.
Entry is the next gap. Fewer sessions reached product pages (40.2% vs 55.5% at peak), which points Phase 2 toward the homepage and landing experience.
A note on the data
These are directional results from a period with seasonal shifts and concurrent SEO changes, not a controlled test.
Share of visitors continuing to each step, peak vs. off-season. Fewer reached a product page, but more of those who did added to cart and began checkout.
Roadmap
Phase 1 findings pointed toward a deeper filtering layer as the next priority. Three filter types were scoped for the Shopify development handoff. Phase 2 is gated on Phase 1 conversion data. Stabilized CVR and reduced exit rates serve as the green light for development investment.
01
Phase 1: Quick wins
In progress at handoff
Validate IA changes and set a performance baseline.
- Rename page titles and categories to intent-based labels
- Add scent-note badges to collection pages
- Track GA4 and Hotjar data on exit and bounce rates
02
Phase 2: Strategic handoff
Partner-led
Align Shopify developers with research-backed mental models.
- Unlock dev budget only after Phase 1 stabilizes the funnel
- Package the Mood and Occasion taxonomy as user stories
- Hand off sitemap documentation
03
Phase 3: Scalable discovery
Future
Build a full guided-discovery experience.
- Dynamic Shopify filters
- Mobile-first tree testing
- Sample-to-full-size purchase paths
Reflection
The fastest wins came from changing words, not code.
This project reframed what the client thought was a traffic problem into a navigation problem. That pivot happened in the first week of data analysis, before any research was conducted. The SEO audit didn't uncover a new problem; it ruled out the assumed one, which created the space to look at what the on-site data was actually saying.
The short research window forced sharper method choices, and findings were consistent enough across methods to recommend with confidence.
The finding I'd carry into Phase 2: users searching for scent notes rather than browsing filters signals that the mental model for fragrance discovery doesn't match any navigation structure the site currently offers, neither the old one nor the new one. Whether Phase 2's filtering work changes that search bar behavior is the metric worth tracking.
Get in touchFrom the client
“Jillian was a pleasure to work with. She delivered incredible research and strategic presentations identifying multiple high-value areas for improvement on our website. … Highly recommend working with her.”