Fashion Brand Reduces Ad Dependency by 30% Via Visual Search.
How we turned an invisible fashion brand image catalog into the brand's highest-performing acquisition channel through precision image optimization and visual search engineering.
Key Results
visual search conversions
return on organic search spend
paid ad dependency maintained
load time reduced from 4.2s to 0.9s
01 — Challenge
The Challenge.
Despite spending £12,000/month on Meta and Instagram ads, ROAS had declined 40% over 18 months as CPMs rose. The brand had 3,200 high-resolution product images with zero alt-text, filenames like IMG_3847.jpg, and no metadata. Google Lens and image search users found competitors but not this brand. Product listing pages took 4.2 seconds to load on mobile, generating an 80% bounce rate.
Tagged all 3,200 images with structured semantic alt-text. Migrated product assets to AVIF format, cutting page weight 65%. Structured Google Merchant Center with live product schema including inventory, pricing, and aggregated review scores. Integrated direct-to-purchase visual search pathways.
02 — Solution
The Solution.
03 — Strategy
Strategic Thinking.
The underlying reasoning and competitive insight that shaped our approach.
Fashion is one of the highest-volume visual search categories — Google Lens processes over 12 billion visual searches monthly, with apparel representing approximately 17% of that volume. Our strategy identified that the brand was invisible across this high-intent discovery surface because its images lacked the machine-readable metadata required for image indexation. The solution combined image optimization for Google Images and Lens with Google Merchant Center product schema to capture buyers across both the discovery and purchase intent phases.
04 — Execution
How We Did It.
Audited all 3,200 product images for alt-text coverage, file naming conventions, and format — finding 0% had meaningful metadata.
Built a structured alt-text generation system following a precise template: [garment type] + [primary colour] + [material] + [occasion] + [brand name] — applied systematically across the catalogue.
Converted all product imagery from JPEG to AVIF format via a batch processing pipeline, reducing average product page weight from 3.2MB to 680KB.
Rebuilt the product page load sequence to inline critical CSS and lazy-load below-fold images, reducing Largest Contentful Paint from 4.2s to 0.9s.
Set up a Google Merchant Center feed with live ProductListing schema: SKU, availability, price, currency, condition, brand, and aggregateRating from verified purchaser reviews.
Implemented ImageObject schema with product links, enabling direct-to-purchase pathways from Google Lens taps.
05 — Outcomes
What Happened.
Visual search conversions increased 190% within the first quarter post-launch. Mobile page load time dropped from 4.2s to 0.9s, reducing bounce rate from 80% to 31%. Organic search became the second-highest revenue channel within 90 days. Google Lens became the #1 traffic source by session volume in month 4. Meta ad spend was reduced by 30% while total revenue increased.
visual search conversions
return on organic search spend
paid ad dependency maintained
load time reduced from 4.2s to 0.9s
06 — Conclusion
The Takeaway.
Fashion brands are leaving significant organic revenue on the table by treating their product photography as a marketing asset rather than a search asset. The same images that drive paid ad clicks can, with proper metadata, drive free visual search traffic at scale. The compound effect of image optimization is particularly powerful in fashion: each new product added to a properly structured catalog extends the organic reach without additional media spend.
More Case Studies
Further Reading.
Build Your Story
Want results like these?
We audit your current organic position in Fashion & Apparel, map competitive gaps, and build a tailored acquisition system from day one.