Venue Stories: an image led way to browse wedding venues
A scroll first gallery view, A/B tested across four markets
Senior Product Designer
Product strategy, UX, Interaction design, A/B test design, Mobile design
I led the design of Gallery View: a full screen, image led way to browse wedding venues, built as an alternative to the list view for couples still looking for inspiration.
Senior Product Designer, partnering with product and engineering.
I owned:
The gallery interaction model: scroll up or down to change venue, swipe for more photos
Mobile UI and content card states
User flows from both the home and search entry points
Aligning the design to engagement and retention metrics
A/B test setup across two variants and four markets
Why this mattered
On mobile, couples could only browse venues as a list. For lower intent couples still looking for inspiration, that felt transactional: comparing venues took effort and offered little sense of discovery. We believed a passive, immersive way to browse would match their mindset, deepen engagement and create the emotional hook that brings couples back and, ultimately, drives stronger enquiries.
What we set out to achieve
Business Goals
Sign up to non concierge enquiry: 20% to 23%
Sign up to favourites: 33% to 37%
Scroll depth past the 5th result: 55% to 65%
Week 1 retention: 19.43% to 20.43%
User Needs
Research and session recordings showed couples:
Want inspiration, not just information, early on
Engage more when content feels effortless and entertaining
Respond to visual formats they know from social media
Want to shortlist venues that feel right, not just ones that fit
Design Hypothesis
Couples who use the gallery view will engage more deeply because they:
Can browse without committing to a full profile
Connect emotionally through rich visuals
Have a more enjoyable reason to keep scrolling and return
Could a TikTok style, image led browse lift passive discovery, retention and enquiries, without disrupting high intent flows?
What I learned before designing
Research & Insights
I reviewed TikTok and Reels research alongside FullStory session recordings. Key insights:
Short, scroll driven formats increase the sense of discovery
Entertainment drives continued use more than information alone
Low intent couples respond better to content led formats than to search queries
Information Architecture
Gallery View became an opt in third view on search results, behind a view toggle, with one test variant landing on it by default. List view stayed intact for high intent couples.
What we built, and the limits we built within
A third way to view search results, reachable from home and search. In Gallery View, couples: Scroll up or down to move between venues Swipe left or right for photos, map, reviews and video Tap persistent buttons to view the profile, favourite, share or request a brochure Filter results as usual
Technical
Mobile only (90% of users)
Four markets: UK, IE, DE, FR
Fewer images in DE and FR (FR median 9, UK median 14)
Design
A third view that couldn’t disrupt list and map
Card design resilient to low image counts
Time on card, scroll depth and discards captured for future ML
How it works for the people using it
Vertical scroll: moving between venues
Full screen cards for one handed use
Up for next, down for previous, like Reels
One swipe, one venue
Horizontal swipe: exploring a venue
Photos, map, reviews and video in one card
Familiar carousel behaviour
Content types kept visually distinct
Action buttons
Persistent on the right: View profile, Favourite, Share, Brochure
Secondary to the content, by design
Profile tap opens the full venue profile
Visual design
Full bleed imagery for emotional impact
High contrast buttons over varied photos
Minimal chrome
Fits the design system while feeling made for discovery
The calls I made, and why
1. Vertical scroll between venues. Matches the Reels model our audience already uses, so there’s almost nothing to learn.
2. Full profile on tap. Keeps the high intent path intact. Gallery View is an on ramp, not a replacement.
3. Time on card as a signal. Attention becomes an implicit shortlist, ready for future ML recommendations without asking couples for input.
4. Separate opt in and default variants. Shows whether it works when chosen versus when surfaced; the default tests the strongest version of the hypothesis.
How we proved it
Control: list and map views only
Variant 1: list view with a toggle to Gallery View
Variant 2: Gallery View by default
Markets: UK, IE, DE, FR, mobile only
Succeeds at: enquiries +3pp, favourites +4pp, scroll depth +10pp, Week 1 retention +1pp
Guardrails: enquiry conversion and retention must not drop
Looking back
What worked
Passive browsing matched how low intent couples behave: they wanted to browse, not hunt
Time on card opened a promising personalisation path
The interaction was simple enough to test the core hypothesis cleanly
What I’d do next
Build a “For You” feed combining image search, Gallery View and Venue Stories
Test native video in cards
Close the DE and FR image gap with enrichment or graceful fallbacks
Gallery View tested whether passive, image led discovery could turn low intent browsing into real engagement, and whether that would compound into retention and enquiries.
It meant balancing a strong vision with real limits: thin image libraries outside the UK, a mobile only scope and existing flows that already performed well.
My approach: a clear hypothesis, the simplest testable version, careful instrumentation, and a design ready for what comes next.
