For many UK eBay sellers, cracking the code of search visibility can feel like chasing shadows. Why does one listing consistently appear on the first page while another, with an identical product, sinks without a trace? The answer lies in the Cassini eBay in-house search engine that decides what buyers see, in what order, and when.
“Ostersund, Sweden – July 24, 2011: Close-up of eBay’s website on a computer screen. eBay is one of the largest online auction and shopping websites in the world.”
Cassini powers eBay marketplace across its 130 million active buyers and nearly two billion live listings worldwide. First rolled out in 2013 to replace the older Voyager system, Cassini has grown into one of the most advanced commercial search engines in existence. Its job is simple but critical: understand what shoppers mean when they search and surface the listings most likely to result in a sale.
Cassini: Built for Buying, Not Just Searching
Unlike Google, which scans the open web, Cassini operates inside a commercial ecosystem where success is measured in completed transactions, not clicks. It analyses mountains of behavioural data search terms, click-through rates, item specifics, seller performance, and buyer histories — to learn what drives conversions.
When a shopper in Manchester searches “wireless headphones”, Cassini doesn’t just look for text matches. It factors in the shopper’s previous searches, brand preferences, average spend, and even time of day. Two people could type the same phrase and see entirely different results.
eBay’s engineers describe Cassini as a behavioural search engine, built to personalise results in real time and reward sellers whose listings attract engagement and trust.
The Three Pillars: Relevance, Performance and Trust
Cassini’s ranking logic is built on three intertwined principles: relevance, performance, and trust.
Relevance is about content — the accuracy and clarity of titles, item specifics, and descriptions. Cassini’s natural-language algorithms can detect keyword stuffing and downgrade spam-like titles. A concise, human-readable headline such as “Apple iPhone 14 Pro Max Silicone Case – Midnight” performs better than a cluttered one crammed with duplicates and variations.
Performance covers behavioural data: click-through rate, conversion rate, and sales velocity. Listings that attract clicks and close sales climb the rankings, creating a virtuous cycle. According to a 2024 Marketplace Pulse analysis, listings with strong engagement metrics receive up to 40% more impressions than those with similar prices but weaker performance.
Finally, trust reflects a seller’s reputation — feedback scores, delivery reliability, and customer service. In the UK, Top Rated Sellers routinely appear higher in search because Cassini associates their accounts with lower risk and higher satisfaction.
Working With Cassini, Not Against It
Cassini isn’t something to trick; it’s something to collaborate with. Successful UK sellers increasingly use data-driven optimisation — drawing on tools like eBay’s Seller Hub performance reports and Terapeak Research — to make evidence-based adjustments.
Keyword strategy should mirror how real buyers search. Use British spellings (“colour” not “color”), local phrasing (“trainers” not “sneakers”), and relevant categories to match UK user intent. Including complete item specifics — brand, size, condition, material — gives Cassini structured data to index your product correctly.
Images and pricing remain powerful signals. eBay’s own testing found that listings with at least six clear photos had a 12% higher conversion rate. Competitive pricing and transparent postage policies (including Free Postage where possible) boost click-through and purchase likelihood — two of Cassini’s key engagement metrics.
The AI Shift in eBay Search
Cassini’s evolution mirrors broader changes in AI-driven retail. Its models are increasingly context-aware, analysing user behaviour across entire sessions rather than single searches. The goal is to interpret intent, not just keywords.
In practice, that means Cassini can tell whether “football boots” refers to adult or junior sizes, or whether “iPad case” implies a specific model generation. The system continuously retrains itself on millions of data points from the UK marketplace — price trends, shipping expectations, and seasonality — to deliver more relevant results with every query.
The Bottom Line
Cassini isn’t an enigma; it’s an algorithm built around human behaviour. It ranks listings based on how well they satisfy real shoppers — rewarding those that are accurate, engaging, and trustworthy.
For UK sellers, the key to success lies not in gaming the system but in understanding it. Craft clear titles, use complete item data, maintain strong feedback, and deliver consistently. Cassini measures the same things your customers care about: credibility, quality, and value.
In a crowded marketplace, that’s what turns visibility into sales — and browsers into buyers.
Feature Image from FreePiK
