Webinar: Increase Conversion With Better Search
2 likes777 views
This document discusses a partnership between IBM Commerce and Lucidworks to improve e-commerce search experiences. Key points: 1. The partnership will integrate Lucidworks' Apache Solr-based search platform Fusion with IBM Commerce to power search and recommendations on IBM Commerce sites. 2. Fusion will enrich product content, queries, and results to improve findability. It will also use signals from user interactions for more relevant results and personalized recommendations. 3. The integration aims to improve customer experiences and conversions by ensuring customers can find products through various query types and discover related items to buy.
Webinar: Increase Conversion With Better Search
- 1. Sarath Jarugula, VP Lucidworks @sarath August 26, 2015 IF THEY CAN’T FIND IT, THEY CAN’T BUY IT
- 2. IBM Commerce Technology Ecosystem + Lucidworks Iris Yuan - iyuan@us.ibm.com August 2015
- 3. Partnership with Lucidworks Certified, open ISV ecosystem Validated and integrated with Websphere Commerce Websphere Commerce (WC): eCommerce platform to deliver complete omnichannel shopping experience - mobile, social, in-store • Out-of-the-box storefronts for B2B and B2C commerce • Customer experience management (Commerce Composer) Open, extensible ecosystem to plug into and build on top of WC
- 4. Extending the Commerce Platform with Search Leverage a complete search solution on top of core WSC capabilities Drive conversions and personalized shopping with scalable, responsive, relevant search experience to every customer Powerful analytics and marketing to tie app and network performance to business results/ campaigns Admin interface to maintain and scale search configuration, capture user activity Apply advanced pipeline, signal processing, and recommendation features
- 5. Sarath Jarugula - sarath@lucidworks.com August 2015
- 6. Lucidworks and Solr Commercial steward of Apache Solr project Employs 1/3 of active Solr committers Contributing 70% of committed code Sponsors Lucene/Solr Revolution, the largest open source user conference dedicated to Apache Solr
- 7. ENVIRONMENT FEATURES SUPPORT LEVEL ADDITIONAL SUPPORT Availability Response Time Number of Incidents Pricing Model DEVELOPER SUPPORT SOLR ENTERPRISE FUSION DEVELOPMENT PRODUCTION • How-To Support • Knowledge Base • Fusion Support • Security • Log Analysis / SiLK Support • Dashboards & Reporting • Enhanced Admin UI • Security • Crawlers & Connectors • Log Analysis / SiLK Support • Enhanced Admin UI • Data Enrichment • Machine Learning • Recommendation • Advanced Relevancy Tuning • WebSphere Integration • 9x5 • SLA-Backed • Unlimited Incidents • Per Named Developer • 24x7 • SLA-Backed • Unlimited Incidents • Per Node • Dev Support (4 Contacts) • Operational Support • Regular Health Checks • 24x7 • SLA-Backed • Unlimited Incidents • Per Node • Dev Support (4 Contacts) • Operational Support • Regular Health Checks Product Offering
- 8. Delightful Commerce Experience Delivered 1. Content and Query 2. Enrich Content, Query, and Results 3. Signals 4. Recommendations
- 9. Optimize Content & Query
- 10. The 12 Queries 1. Exact Search 2. Product Type Search 3. Feature Search 4. Thematic Search 5. Relational Search 6. Compatibility Search 7. Slang, Abbreviation, and Symbol Search 8. Subjective Search 9. Symptom Search 10. Implicit Search 11. Non-Product Search 12. Natural Language Search Access white paper: http://bitly.com/12_queries
- 11. If They Can’t Find It, It Doesn’t Exist Is this what your customers are experiencing? A recent large-scale ecommerce survey observing users’ search functionality shopping experience
- 12. Users Perception Assumed Relevancy Expect Powerful, Helpful Search Google Experience
- 13. Customer Assumes Store Doesn’t Carry the Item • Include multiple title spellings • Variations with other query types • Intelligent handling of misspellings. Examples • keurig k45 • stuhrling 879.03 mens watch • nikoncoolpixs2800
- 14. Customers Have Difficulty Finding Products on Your Site • Product Type Synonyms as Categories • Include categories that are and aren’t part of the site’s hierarchy • Suggest Categories as search scopes • Landing Pages Examples • sandals • sofas • barstools
- 15. Customers Expect to Find Products by Their Features • Store all Product Attributes • Add Tags from Description and External Sources • Users Combine “Feature Search” with Other Query Types Examples • red knit sweaters • ceramic coffee grinders • manual espresso machine • 10gb ssd • waterproof bluetoothspeaker
- 16. Customers Expect to Find Products by Their Interests and Love • Combine “Relational” Queries with Other Search Query Types • Highlights and Contextual Snippets • Suggestions and Recommendations Examples • new tom hanks movie • new anne rice novel • second matrix dvd
- 17. Find Products by Customer’s Interpretation, Attributes, and Opinion • Enrich Data for Subjective Approximations and Proxies • Look Beyond Catalog Data • Analyze Interpretive and Taste-based Search Examples • high quality tea kettle • cheap wine • light weight tent
- 18. Deliver User’s Personalized Search Experience • Use All Available Environmental Data • Learn from Past History • Refine Query • Suggest Relevant Similar Searches Examples • pants (from a Women’s Apparel category page) • charger cable(from an iOS Devices landing page)
- 19. Deliver Results based on Meaning of User’s Spoken Language • Go Beyond Keyword Matching • Mainstream with Mobile Usage (Voice) • Closest to In-Store Experience • Integrate NLP into Your eCommerce Platform Examples • men’s sneakers that are red and available in size 7.5
- 20. Search Experience Delivered by most eCommerce Businesses Impacts • Shopping Experience • Conversions • Units per Transaction
- 21. Enrich: For Enhanced Experience
- 22. Enrich Across Content, Query, and Results QUERY MODIFICATION Increase the findability of documents and records with automatic creation of tags, fields and meta-data Curate the user experience in your application with artificial result ranking, document injections and obfuscation RESULT MANIPULATIONINDEX TIME ENRICHMENT Perform real time decision making and routing in order to map a users intention or enterprise policy
- 23. Lucidworks Fusion Pipelines Leverage pre-defined OOB processes to add a stage to enhance the catalogue data Instantly review how the data is processed at every stage before it’s updated in the index Create custom stages to bring metadata from different repositories to enrich the product catalogue Simple admin to add query stages and user profiles to enhance simple user’s query phrase Instantly review how the query is processed at every stage and the final search results presented to the user Create user specific personalized search experience • Landing Pages • Security Trimming • Javascript (for custom scripting) • User Profiles • Tika Parser • Exclusion Filter • Field Mapper • XML Transform Stage • OpenNLP Entity Extraction • Gazetter Extractor • Regular Expression Extractor • Javascript (for custom scripting) • Search Fields/Parameters • Facets • Boost Documents • Block Documents Sample OOB Index Pipeline stage Sample OOB Query Pipeline stage Stage-1 Stage-2 Stage-3 Stage-n Solr Index (Collection) Stage-1 Stage-2 Stage-3 Stage-n User ExperienceQuery PipelinesIndex Pipelines Solr Index (Collection) Solr Index (Collection) Solr Index (Collection) Solr Index (Collection) Solr Index (Collection) Index Cluster
- 24. Realtime Analytics - Respond to Interest Spikes and Events
- 25. Real time interactive analytics • Dashboards display real time users interaction • Integration will deliver pre-defined dashboards with most common analytics • Drill down into the analytics data all the way to a single event or user interaction • Create time-series to understand patterns and anomalies over time Configure role based personalized dashboards • Administration interface to build new dashboards with minimal effort • Create personalized dashboard views based on business unit or job role • Admin can setup dashboards per their business requirements to enable realtime analysis of their products and user activity Proactive alerts • Configure alerts to notify new events • Realtime proactive alerts help businesses react in realtime Search Driven Analytics
- 26. Signals - Differentiate from Competition
- 27. Signals power relevance. Clicks, tweets, ratings, locations and much more can all be leveraged to provide high quality recommendations to users and deeper insight for data scientists. Connector Framework Index Pipelines (ETL) ( )Scale Fault Tolerance Real-Time Fusion APIs Recommendations Personalization Contextual Search Relevancy Tool Machine Learning / Signal Processing Analytics Security Ecommerce Site Customer Analytics Product Catalog User History Conversion Data
- 28. Signals power relevance eCommerce Platform and IBM Analytics captures powerful signals User’s activity of an eCommerce site including browsing and navigating through the landing and search result pages Search and search activity • Select (click) on a product • Rate / recommend a product • Add products to a shopping cart or save to shopping list Algorithms to aggregate signals data to drive improved user experience and business performance Signals framework is built to integrate events data from any application and data source. Schemaless architecture makes it easy to load both structured and unstructured data
- 29. Play nice with elephants Combine the power of Lucidworks Fusion + Hadoop. Immediate access to customer, social, and promotional data—all in one place. Search backed analytics makes every user a data scientist. Lucidworks Fusion has unmatched scalability in search.
- 30. • User interaction signals • Clicks • Add to Wishlist / watch item • Rating • Reviews • Select navigation choices • Add to Shopping cart • Checkout, etc. • Social Signals • Twitter mentions of products • Positive and negative mentions • Learn user behavior via simple dashboards • Configure or build new dashboards through admin screens. • Search within your log, social, and clickstream data to discover insights and patterns RealTime Analysis of Signals Track users events User’s device, os, browser data Specific user Users Geo- Location
- 31. Recommendations - Increase Satisfaction & Transaction Size
- 32. Organic pre built machine learning algorithms offer OOB recommendations Integrated Apache Spark helps to add custom machine learning components to leverage the data captured in IBM analytics and Commerce platforms Predictive Recommendations
- 33. Up sell better products Cross sell related products Popular Recommendations
- 34. IBM Commerce + Lucidworks Fusion: Improve Shopping Experience & Conversion
- 35. Connector Framework Index Pipelines (ETL) ( )Scale Fault Tolerance Real-Time Fusion APIs Recommendations Personalization Contextual Search Relevancy Tool Machine Learning / Signal Processing Analytics Security Apps Mobile Silk Database Web File Logs Hadoop IBM Commerce Integration with Fusion
- 36. https://github.com/LucidWorks/fusion-solr-plugins Simple Integration Modify Solrconfig.xml to load the jars and enable search components Import the WebSphere Solr collection in to Fusion. Configure snowplow in eCommerce application Download IBM Commerce plugin jar 21 3 4
- 37. Demo
- 38. Co-Occurrence Graph Analysis • Incoming users and sessions • Co-occurring Products (products that were clicked on in the same search session). • product id • query • user • session
- 39. Questions? Download Fusion http://www.lucidworks.com/products/fusion Contact Lucidworks http://lucidworks.com/company/contact/ Contact me sarath@lucidworks.com @sarath