Search for E-commerce and Retail

Salient Features

Facets

Makes it easier for your customers to filter or narrow down their product search using categories and sub-categories. Offer site admin to easily configure the product facets and customize them as per the shopping domain. SearchAssist automatically creates facet categories and subcategories based on the indexed data.

Contextual Search

SearchAssist uses the context of the page, past product searches, and visitor demographic, to bring the most relevant results. Whether they are looking for a specific product of a brand or a price range, or if they are filtering a subcategory within a chosen category, SearchAssist makes it super easy for you to provide contextual results.

Recommendations

Provide contextual personalized search results that are tailor-made for your shoppers. Analyze the user profile, search pattern, and search history to predict and provide personalized or contextual recommendations, proactively.

Fine Tune Relevance

Improve the relevancy of the results by tuning the search results. SearchAssist lets you easily adjust the relevancy score of products based on various parameters like click-through rates and popularity or boost the ranking.

Experiments

SearchAssist provides you with the option to run experiments between your different search experiences among your customers to test the effectiveness of the search experience and make incremental changes as required to provide the best performing option

Challenges and Limitations of Conventional Search

The following are the key challenges and limitations faced by businesses in implementing traditional search in their eCommerce channels namely websites or mobile apps:

• Maintaining dedicated resources in the non-core areas of Search and Conversational technologies
• Reactive and passive user engagement limited to information link pointing and snippets
• Customizing  search experiences
• Analyzing performance and Experimentation

SearchAssist Solutions

SearchAssist addresses each of the aforementioned challenges with the following key features:

• a Platform as a Service (PaaS) approach to subscribe, configure and use to streamline search costs effectively
• proactive user engagement transforming traditional eCommerce into a Conversational commerce experience presenting both information and call-to-actions.
• designing custom search experiences and personalizing results
• presenting insights on user engagement, queries, results for feedback, and defining experiments.

Best Practice Recommendations

In the context of eCommerce, most of the features of SearchAssist can be applied to achieve the best outcomes. The following  pre-requisites and best practices are recommended to help you maximize the benefits of a SearchAssistant deployment:

• Maintaining updated product catalogs in structured data files like JSON, or CSV
• Maintaining product FAQs in SearchAssist mandated format
• Having clear localization and marketing promotional offer requirements
• Having product image gallery and URLs ready for each product

Practical Guidelines in the eCommerce context

Consider an example where an eCommerce business that sells smartphones decides to deploy Kore.ai’s SearchAssistant onto its channels namely website or mobile app. Assume they have added the following sources of the content:

• the product catalog in structured data either in CSV or JSON formats
• the associated FAQs are all reviewed and approved,
• user manuals if applicable etc.
• SearchAssistant is linked to a live retail Support virtual assistant called ShopAssistant.

SearchAssist by default indexes the ingested content with the default index configuration. Additionally, you can create custom index configurations, personalize results, design UIs, and customize the search experience with SearchAssist.

The following guides give practical pointers on how you can utilize each feature to deliver an advanced search experience to your users and meet your business objectives:

• Ingesting Data
• Managing Idexing Sequence
• Managing Relevance
• Personalizing Results
• Designing Search experiences
• Experimenting with Variants
• Analyzing Search Performance