Google Cloud Retail Search, now available to all retailers, provides the ability to integrate Google-like search features on digital properties without the need to preprocess data, train or tune machine learning models. It is part of Product Discovery Solutions for Retail.

“While we’ve come a long way from the days when search was largely based on keywords and boolean rules, shoppers still struggle to find what they’re looking for,” Srikanth Belwadi, group product manager at Google Cloud, wrote in a post. “”They often have to come up with a perfectly-worded query that a retailer’s site search engine will understand, sometimes rephrasing several times before getting the results they want — if they find anything relevant at all.”

The technology to support retail websites builds on years of knowledge based on Google Web search. The cloud service can load balance or manually provision the infrastructure to support unpredictable traffic spikes.

Retailers can customize the fully managed service based on innovations in search indexing, retrieval, and ranking.

Google estimates that in the U.S. each year, retailers lose more than $300 billion from bad online search experiences on their websites — likely because consumers expect search engines, especially on retail sites, to understand their intent, return relevant results faster, and help them discover new products with personalized recommendations. 

The technology behind Google Cloud Retail Search supports advanced query understanding that produces better results from even broad queries, like non-product searches. It offers semantic search to match product attributes with website content, optimizes results that leverage user interaction and ranking models to meet specific business goals, and provides access controls for security and privacy.

Search is the most common and frequently used feature on a retail website. Despite frequent use, 94% of U.S. consumers report they have received irrelevant query results.

Some 76% of consumer report after an unsuccessful search they will abandon the site, and 48% will purchase the item elsewhere. About 52% said they typically abandon their entire cart and go elsewhere if there’s at least one item they can’t find.

Lowe’s, an early adopter of the technology, partnered with Google Cloud to give its customers relevant results for long-tail searches. The company has seen an increase in click-through and search conversion, and a drop in the “No Results Found” rate since launched, according to Neelima Sharma, senior vice president, technology, e-commerce, marketing and merchandising at Lowe’s.

Sharma believes that with limited customer signals and no historical data, descriptive long-tail searches are some of the most challenging queries to understand, especially on retail sites.

“We have been partnering with Google Cloud to give our customers relevant results for long-tail searches and have seen an increase in click-through and search conversion and a drop in our ‘No Results Found’ rate since we launched,” she wrote in a blog post.

Key partners also include GroupBy, Lucidworks, GridDynamics, SpringML, and others.

Google Cloud Retail Search, now available to all retailers, provides the ability to integrate Google-like search features on digital properties without the need to preprocess data, train or tune machine-learning models. It’s part of Product Discovery Solutions for Retail.