Transforming Telecom search and CX with Google Discovery AI and commercetools

Recent years are marked by dramatic changes in the Telecommunications business environment, driven by the pandemic, economic turbulence, and political instability. As Forrester put it, “Constant consumer change is a defining element of the “new normal.” Consumers changed their shopping habits, their brand loyalty is shaken and they are ready and willing to try new ways to shop online.

Leading Telecom companies are stepping up to embrace this change rather than resist it, and are modernizing their platforms with a new architectural approach - composable commerce.

Composable Commerce promotes using best-in-breed business capabilities from different vendors, available as cloud APIs, focusing on business flexibility, independence, and promptness, rather than relying on a single vendor platform. Composable commerce is powered by MACH architecture, which focuses on seamless integration of API-first services by design.

commercetools is the most popular composable MACH commerce platform, and Grid Dynamics has successfully implemented the platform for a number of well-known brands. commercetools also has integrations with a number of leading cloud-native SaaS providers covering key digital commerce capabilities.

One of the most crucial capabilities in digital commerce is search, enabling customers to find products they love, browse product catalogs with ease and receive relevant product recommendations that are personalized to their needs and wants.

Recently, Google expanded its Discovery solutions for Retail, announcing general availability of its next generation product search engine: Google Cloud Retail Search, naming Grid Dynamics the official launch implementation partner.
Naturally, we integrated Google Cloud Retail Search into a composable commerce blueprint and developed a Google Cloud Discovery AI Starter Kit for commercetools that provides essential integration capabilities.

What is Google Cloud Retail Search?

Over the years, product search has experienced several paradigm shifts. It has evolved from simple word-matching in the product title to an AI-based approach that leverages deep learning models to find the most relevant products based on query intent and customer behavior data.

Google Cloud Retail Search is the next-generation search service from the Google Cloud Discovery AI for Retail suite of solutions that represents this new AI-based approach to search. It combines state-of-the-art natural language processing and relevance models with a vast Google knowledge graph to truly understand both customer queries and product data. It empowers digital retailers to provide best-in-class, personalized search and browse experiences by seamlessly and conveniently connecting customers with products they want, need and love.

Google Cloud Retail Search enables retailers to enhance the search experience within their e-commerce ecosystems with a variety of advanced features, including:

• Advanced query understanding to provide better search results from broad and ambiguous queries, including queries unrelated to the product.
• Semantic search allows customers to find more personalized, meaningful, and relevant search results by analyzing the context of queries and customer preferences.
• Optimized results ranking that leverages user interaction and advanced ranking models which can be optimized for specific business goals, such as engagement or convergence.
• State-of-the-art security and privacy practices ensure retailer data is protected with strong access controls, and is only used to deliver relevant search results on their properties.

What is commercetools?

commercetools is the leading platform for next-generation B2C and B2B composable commerce. It acts as the backbone API and event bus, and integrates all the concerns of modern digital commerce platforms, implemented as “packaged business capabilities” or PBCs.

Some of the PBCs are provided with the commercetools platform, but others can be integrated from 3rd party providers or implemented in-house as groups of microservices.

The key aspect of a composable commerce architecture is that PBCs are well isolated from each other and interact only through APIs and messages. This makes it straightforward to substitute them with another implementation if needed. This enables very granular build-vs-buy decisions, and allows businesses to select, integrate and experiment with best-in-class SaaS platforms.

The headless, API-first nature of the platform promotes and enforces decoupling the front-end from the business services backend, allowing ultimate flexibility and differentiation in customer experience.

Businesses that embrace the composable commerce approach and MACH architecture are optimized for business flexibility. They are empowered to react quickly to changes in the business environment and consumer demand, and outpace the competition considerably.

In a nutshell: Google Cloud Discovery AI Starter Kit for commercetools

The starter kit for commercetools is a data bridge that enables communication between the Google Discovery AI APIs and the commercetools API. Catalog data is extracted from commercetools, staged in Google Cloud Storage, transformed and mapped into a Google Discovery AI product data format, and loaded into the Google Discovery AI services. Catalog, price and inventory changes are streamed to Google Discovery AI to ensure up-to-date search results.

The starter kit is built following a cloud-native approach, leveraging powerful Google Cloud data processing services such as PubSub, DataFlow, CloudFunction and BigQuery, to facilitate data synchronization between commercetools and Google Retail Search.

The current implementation of the starter kit allows customers to develop a data pipeline for integration of catalog/price/inventory data. We are planning more useful features as the starter kit evolves, such as WebUI integration, promotional data ingestion, and much more.

The Google Cloud Discovery AI Starter Kit for commercetools is already available for Grid Dynamics customers and will be published on the marketplace soon. Stay tuned for more details or get in touch with us to get started!