Typeahead lowers bounce rates and raises conversions. By combining an ultra-fast auto-complete solution that starts at the first letter, with a ranking algorithm that accounts for both popularity and newness, customers are more likely to quickly find what they are looking for.
Misspellings are very common in search queries, and proper spelling corrections are a must. Corrections offered by dv-search take context into account, and allow you to replicate a Google-like experience, with "Did you mean?" type suggestions.
Is "champagne" a color or a type of wine? It depends on the context. By finding and classifying entities from a user search query, dv-search can disambiguate the intent, and then map the results to aggregations (facets) to further improve the search results quality.
In the physical world, an experienced salesperson can quickly recommend products that partially match the original intent of the customer. Query relaxation is a feature offered by dv-search that does this automatically: provide a less specific query that retains the core original intent.
Our dv-search solution supports all the typical operations found in advanced search engines: boolean operators, stemming, faceted search, exclusions, sorting, pagination and more.
Query understanding requires the reliance on a wide range of Natural Language concepts, so as to overcome the limitations of keyword searches. Our domain specific Knowledge Graph, combined with query understanding provides the building blocks to gradually handle more complex queries, expressed as the customer thinks.
Ranking is continuously updated to account for customer selections from search results, as well as new product introductions and inventory levels.
Track performance indicators such as search conversion, new trending searches or unmet demand, through convenient dashboards that can be further broken down by geography or other dimension such as product attributes.