$2.4M
New annual revenue
Visual Comfort case study
29x
Return on investment
Visual Comfort case study
Many
Pages created
Automatically generated
1st
Year ROI measured
Visual Comfort case study
Your best products are hidden in plain sight
Customers who find one product rarely discover your other popular items. Your best-sellers get traffic from search, but customers don't navigate between them naturally.
Products stay siloed
Customers browsing leather sofas rarely discover your dining tables or bedroom furniture. Each product category exists in isolation, limiting cross-category discovery.
Same navigation everywhere
Static “Popular Products” blocks show the same items on every page. Customers see them once, ignore them forever. They feel like spam, not recommendations.
Manual updates don't scale
Maintaining product recommendations across thousands of pages typically requires significant engineering time. Best-sellers change, but your navigation stays stale.
Smart distribution that feels personal
The Popular Links agent adds dynamic linking blocks that connect pages to popular, high-demand destination pages based on search data, with each page showing different recommendations.
The result: your leather sofas appear more often overall because they're a high-demand page, but customers on your velvet sofa page see varied suggestions. No repetitive blocks. No “same five products everywhere” problem.
You control what determines importance: sales volume, revenue, customer engagement, or your own priority score. The Popular Links agent handles the distribution.
How priority scores work
Leather sofas (top seller)
Dining tables (high engagement)
Bedroom furniture (solid performer)
Garden furniture (seasonal)
Higher priority = more visibility across your site
Varied suggestions vs repetitive blocks
Traditional “popular products” modules often show the same items on every page. The Popular Links agent uses LLMs, SERP data, and importance signals like taxonomy depth and traffic to dynamically identify and link to the most relevant high-demand destination pages, ensuring your high-demand pages get more visibility overall.
Traditional approach
Page: Velvet Sofas
Popular products:
- 1. Leather sofas
- 2. Corner sofas
- 3. Sofa beds
Page: Dining Tables
Popular products:
- 1. Leather sofas
- 2. Corner sofas
- 3. Sofa beds
Page: Bedroom Sets
Popular products:
- 1. Leather sofas
- 2. Corner sofas
- 3. Sofa beds
Same recommendations everywhere. Customers tune them out after the first page.
With Popular Links agent
Page: Velvet Sofas
You might also like:
- 1. Leather sofas
- 2. Dining tables
- 3. Fabric armchairs
Page: Dining Tables
You might also like:
- 1. Dining chairs
- 2. Leather sofas
- 3. Bedroom sets
Page: Bedroom Sets
You might also like:
- 1. Corner sofas
- 2. Wardrobes
- 3. Leather sofas
Dynamic and data-driven “Leather sofas” appears more often overall because it's a high-demand page based on search data.
Two ways to help customers browse
Most retailers use both Related and Popular recommendations together. Customers get two clear ways to explore: similar products and popular products.
Velvet Sofas
Related products
Similar to what you're viewing
Popular across the store
What other customers are browsing
Related products
“What's similar to what I'm looking at?” Helps customers explore within a category.
Popular products
“What are other customers buying?” Helps customers discover your best-sellers.
CASE STUDY
How Visual Comfort drove $2.4M in new annual revenue
Visual Comfort, a luxury lighting retailer with over 60 premium brands, used Similar AI to automatically create new category pages and connect them with smart product recommendations.
The result: $2.4 million in new annual revenue from AI-created pages. According to Jennifer Skeen at Visual Comfort, this represented a 29x return on their investment with Similar AI.
$2.4M
New annual revenue
29x
ROI
Many
Pages created
1st
Year ROI measured
Part of a complete product discovery system
The Popular Links agent is one way Similar AI helps customers find products. Combine with other Similar AI's Linking Agent features for maximum impact.
Frequently asked questions
Does this require engineering resources?
Minimal. For all customers, implementation takes days, not months. We handle the technical complexity; your team just needs to approve the changes before they go live.
How quickly will I see results?
Most customers see measurable traffic increases within 4-8 weeks. Revenue impact follows as new pages drive organic traffic and generate revenue. The exact timeline depends on how quickly search engines re-index your pages.
How is this different from static 'Popular Products' blocks?
Traditional popular product modules often show the same items on every page. Customers see them once, then ignore them. The Popular Links agent uses LLMs, SERP data, and importance signals like taxonomy depth and traffic to dynamically identify and link to the most relevant high-demand destination pages, ensuring your best-sellers get more visibility overall. The result feels curated, not templated.
Can I control which products get priority?
Yes. You can choose what determines a product's priority for example, sales volume, revenue, customer engagement, margin, or a custom score you provide. You can also exclude specific products or categories from recommendations entirely.
How does this work with Related recommendations?
Most retailers use both. Related shows products similar to what the customer is viewing (velvet sofas → fabric sofas). Popular shows your best-sellers across categories (velvet sofas → dining tables). Together, they give customers two complementary ways to browse.
See how it works with your products
Book a demo and we'll show you how the Popular Links agent can help customers find your best products. We'll use your actual catalog. No commitment required.