Overview of Apptus search and Navigation capabilities
Live site, customer screen shotBringing back database matches....others don't......so brilliant performanceMost popular at the top
Another live site Huge catalogue , fast enough to autocomplete , for every letterFinding hits each match makes Most popular matches
Effectively bring back the user requirement – without having to set the synonym “implicit synonyms” – are a bonus!!! Don't have domain specific starter packs !Merchant specific synonyms – merchant IP
So this again underlines the importance of linking search and recommendations The recommendations engine, by knowing what was searched for, can make the most appropriate recommendations despite spelling
Dynamicnav options (iconised) for a brand landing page !Subset – size and brand sortCustomer is innovative – and they bought our technology!!
Animated illustration of how search results and navigation are dynamically optimised using eSales
This illustrates the basic “bought – bought” recommendations Look for bottom half of page next
Eg shows the films with same actors Pulls out other products with shared attributes / attribute values Can be same fabric e.g. cottonSame wood e.g. Looking at birch chairs, show other birch furniture It is a BIG DEAL !!
Illustrating recommendations and how their display can be optimised
Showing how content such as campaign banners can be targeted to a user – essentially providing automatic, personalised AB testing and optimisation
Snapshot of eSales ManagerLHS column shows available panels containing merchandising componentsMain area shows how they are placed in the page hierarchy
Showing how we report on the performance of our merchandising.
TODO – add stats for clients
Transcript of "7 michael mokhberi apptus sebc"
Intelligent Search<br />ScandinaivaneBusiness Camp, 26th of May<br />
Introducing Apptus<br />Pioneering relevance engine technology<br />Search, recommendations & content targeting for e-Commerce<br />Boosts site profits; streamlines marketing & merchandising back-office<br />Founded in 2000, profitable and VC backed<br />HQ & development in Lund - Sweden<br />Sales offices & partner network in Europe and North America <br />
The many dimensions of relevancy<br />WHAT<br />Only 10% of the users know the exact name, article id or specifics of what they are looking for<br />80% of searches address 20% of the database<br />The margin of the long tail is between 50-400% higher than the top list<br />WHERE<br /><ul><li>Nearest store where the buyer can explore the product
Nearest location where the buyer can fetch or return the goods
Location specific purchase/delivery terms</li></ul>WHEN<br /><ul><li>Seasonal influences
Shoppers will never see ‘no results found’</li></ul>More ways to browse <br /><ul><li> Encourage shoppers to linger
Opportunity for up-sell and cross-sell</li></ul>Personalize by relevance for higher conversion <br /><ul><li> Rank relevant attributes higher
Include user ratings</li></li></ul><li>Navigation on brand landing page<br />Context-sensitive filtering<br />
…optimises use of page real estate<br />eSales search finds best match to what user is looking for in each category…<br />Product <br />Product <br />Product <br />Search, navigation and layout optimised for maximum conversions based on relevance & crowd learning<br />Faceted search personalised<br />
Category-based recommendations<br />Customers who bought things in this category bought...<br />
User-driven recommendations<br />People who bought this bought that...<br />
Attribute-based recommendations<br />Other products with similar attributes....<br />
15% admit to purchasing when they see recommendations
“Retailers told us … that between 2% and 20% of their revenue could be attributed to recommendations”</li></ul>By year-end 2013 over 30% of the 100 most popular websites will use search technology or content analytics to target content at users.<br />Börge Olsen, Sales Manager<br />Personalised promotions double retention rates to 16%<br />
Referencecase: CDON<br /> CDON – the Amazon of the nordics<br />Challenges:<br />Slow search response times<br />Unstable IT environment and service outages @ peak load<br />Irrelevant products shown to buyers<br />Results with Apptus eSales:<br />Lightening fast response time regardless of load <br />99.97% uptime and excellent reliability during peak hours<br />Record sales in 2010 thanks to relevant products exposed to the buyers in different contexts<br />Mikael Olander, CEO<br /><ul><li> 10 million products
4.2 million searches/day; peak 2,000 searches/sec