Reimagining customer experience in the telecom industry with AI-powered e-commerce search

The telecom industry has faced numerous challenges in recent years, including waning revenues and return on invested capital (ROIC), as well as difficulty keeping up with post-pandemic customer demands. It would be expected that, by now, telecom operators would have adopted AI-powered solutions to overcome the turbulence of the past few years, and to remain resilient in the face of the uncertainty that lies ahead. Yet, many surveys show that the industry has been slow to fully embrace AI or an AI-focused mindset. And where telcos have invested in AI, the investments are often not aligned with top-level management priorities or customer demands. Is there an AI entry point for telcos that promises tangible results and immediate customer satisfaction? Yes!

Dominating the customer experience trend in 2023 is hyper-personalization, with AI-powered e-commerce search solutions as a leading enabler. And what better AI-powered search could there be than Google-quality search? In this blog, we explore the benefits of vector search, product recommendations and image search from the Google Discovery AI for Retail suite. As Google Cloud Partners, and Discover AI for Retail Integration Partners, Grid Dynamics is uniquely positioned to make the migration to Google Retail Search fast, efficient and seamless. Get started quickly with data ingestion and change streaming pipelines, integration with clickstream data, data quality framework, search results comparison tools, and much more.

The COVID-19 pandemic disrupted the lives and livelihoods of consumers and businesses across the globe, forcing them to embrace digital as the new normal. Bound to the confines of their homes, some 92% of consumers abandoned erstwhile shopping habits to become online shopping converts, and e-commerce accelerated years of digital modernization into months to meet the needs of this new explosion of digital consumerism.

This new breed of digital consumers expects services on demand, within minutes, and a personalized experience to boot.

How has this impacted CTOs and business leaders in the telecom industry? Meeting these lofty expectations requires a shift not only in technology, but a better understanding of how customer behavior is changing, and how best to respond to it. And remember, with a plethora of online shopping options for customers in today’s digital marketplace, it only takes one negative experience for a customer to ditch your brand in favor of a competitor’s.

If you’re in the Telecommunications industry, you should already understand how important your e-commerce search solution is for creating exceptional digital customer experiences, boosting conversions, and reducing customer churn. To reinforce just how critical it is, consider the following stats from a recent Google Search Abandonment Survey:

• 94% of global consumers report receiving search results that were not relevant to what they were searching for on a retail website.
• 85% of these consumers say they view the brand differently after an unsuccessful search, and 74% say they avoid websites where they’ve experienced search difficulties in the past.
• Retailers lose over $300 billion each year from bad online search experiences in the US alone.

It should, therefore, be a priority for telcos to personalize the product search journey and provide Google-quality search and recommendations to attract, retain and delight customers in 2023, and beyond.

Thankfully, we are witnessing the emergence of a revolutionary new approach to producing personalized and context-driven search in retail based on semantic vectors  – an approach that Google Discovery AI for Retail leverages for ultimate search and recommendations performance.

Google Discovery AI for Retail is the next-generation answer to context-driven search that enables telcos to implement Google-quality product search. Google Discovery’s global-scale AI quickly and easily enables retailers to curate unique and competitive path-to-product digital user experiences. It extends decades of Google’s advancements in query understanding and semantic search capabilities, specifically designed to minimize shopping friction and maximize revenue.

It is bundled with the following sophisticated tools to deliver highly personalized search results, and product recommendations at scale across channels, and help shoppers easily search for products with an image on their mobile device:

• Retail search and browse leverages deep query understanding and semantic vector capabilities to deliver relevant, personalized search results.
• Recommendations AI works to understand nuances behind customer behavior, context, and SKUs in order to drive engagement across channels through relevant recommendations.
• Vision Product Search uses ML-powered object recognition and lookup to provide real-time results of similar or complementary items from your product catalog.

Increasingly, customers prefer to interact with search in a natural, conversational way and expect the solution to logically understand their intent, remember their preferences, and predict their shopping behavior. Unlike traditional search engines that relied primarily on text-pattern-matching techniques and complex manual rules without intelligently comprehending the properties or attributes of the underlying product, Discovery AI runs through all available search data to determine what the customer’s buying intent is, and what they are actually looking for. It calculates additional parameters like viability and personalization, to understand how the search quality matches the user’s preferences, and then arranges the results to maximize revenue.

Google Discovery AI for Retail is not just a search tool for users, but an irreplaceable interactive media solution that helps customers discover and immediately connect with the products they actually want. When customers search, they search to buy. That’s why Discovery AI ranks products not only by relevance but also buyability and revenue potential.

So what powers Discovery AI besides the next-generation search engine technologies and approach?

• Security, privacy and compliance with GDPR is standard to ensure that your data is yours alone. Your data is never shared with the rest of Google.
• Discovery treats a search query as a successful call to the SearchService.SearchMethod that returns a list of search results for a set of query strings, filters, or facets.
• Machine learning models enable omnichannel search and recommendations implementations. These machine learning models use features like color, brands, price, size, and seasonality, and have been proven to be very effective in increasing the conversion rate and per-visit revenue.
• There is no need to preprocess data, train or hypertune machine learning models, load balance or manually provision your infrastructure to handle unpredictable traffic spikes. Google does it all for you automatically.
• Discovery AI allows uploading historical clickstream data in bulk, with serving controls and request modifiers that enable retailers to easily implement contextual merchandising rules to curate more bespoke and optimized search and recommendation experiences.
• You can integrate data, manage models, and monitor performance, all from an intuitive graphical user interface. Quickly connect data with your existing tools like Google Analytics 360, Tag Manager, Merchant Center, Cloud Storage, or BigQuery.
• The autosuggest feature adds predefined entries or rules to autocomplete user queries using synonyms and spell correction.
• Using decades of Google Search and Shopping insights, Discovery’s advanced query understanding automatically adapts to ensure the customer’s intent is accurately evaluated.
• Discovery AI surfaces the most relevant and profitable products, even when a customer performs the search using misspelled or complex query terms.
• Indexing is powered by rich linguistic support to include additional language and postprocessing in text and digital fields.
• Discovery AI delivers advanced document comprehension capabilities to truly understand attributes to surface the most relevant products for a given customer intent.
• With Discovery AI, there are no throughput limits for queries per request and no extra charges for storage and indexing.
• Using the impressive capacity of Google core features  – unavailable to other search engines  – Discovery AI scales automatically and seamlessly, provides automated search and recommendation model tuning, data quality insights, and personalization.
• With the pay-as-you-go contract model from Google, you get a discount based on your volume commits.

Now let’s look at the Discovery AI for Retail’s suite of products in more detail:

• Retail Search and Browse
• Recommendations AI
• Vision Product Search

It wasn’t long ago that all search was keyword-based and relied on exact term matches to return relevant results. However, today’s customers expect a reply from a search engine that understands the context of the thing or idea they are searching for.

Google Retail Search and Browse goes beyond traditional keyword search and enables retailers to provide highly personalized product results using a semantic-based approach that works to understand the context and intent behind customer queries.

Semantic vector search is a deep learning approach in which a model learns from shopper behavior to capture the latent features of queries and products and represent them as semantic vectors in a multidimensional vector space – similar to the way in which retail products are organized and categorized in a physical store. Semantically similar products and queries are clustered together in the vector space, where nearest neighbors are ranked as the most relevant matches.

Semantic vector search is capable of processing language ambiguities like synonymy and polysemy, out-of-vocabulary natural language, as well as thematic and subjective queries.

These innovative new approaches to understanding the context behind search queries makes semantic vector search the new frontier in product discovery.

In short: Retail Search and Browse help shoppers find exactly what they want faster with personalized results, and it helps retailers better understand the needs of their customers to optimize the experience for higher conversion rates and happier, more brand-loyal customers.

Available across your website and mobile applications, the retail search and browse solution is fully customizable, uses machine learning to improve results with every search, and also includes autocomplete suggestions, and the ability to boost relevant promotions or prioritize ranking based on business goals.

Coupled with recommendations and visual search, described below, customers are now empowered to find the products they need, want and love easier and faster than ever before!

Recommendations AI uses Google’s latest machine learning architectures, which dynamically adapt to real-time customer behavior like views, clicks, in-store purchases and online activity, as well as changes in product variables like assortment, pricing, and special offers. This means that your customers are served with the most relevant and personalized product recommendations, while your product strategies and business goals are bolstered at the same time.

Furthermore, you can track conversion rate, click-through rate, and revenue optimization to help you more precisely optimize for your business goals. You can also take it a step further with the API model and personalize your shopper journey across all channels – from mobile apps and personalized email recommendations, to store kiosks or call center applications.

Vision Product Search is Google’s answer to object detection in the retail industry. Using powerful pre-trained machine learning models through REST and RPC APIs, the search tool can detect objects and faces, and read printed and handwritten text.

Retailers can encode the product images in their catalog with visual identifiers that describe the products from different viewpoints, and categorize them according to the following supported product sets: homegoods, apparel, toys, packaged goods, and general.

Now, when a customer uploads an image of the product they are searching for in your e-commerce search solution, Vision API Product Search applies machine learning to compare the product in the customer’s query image with the images in your product set, and then returns a ranked list of visually and semantically similar results.

Wondering how to integrate Discovery AI for Retail into your e-commerce platform in the most controlled and efficient way?

Migrating to Discovery AI is no easy feat. Every retail environment is unique, with specific product discovery requirements and customer experience needs. In order to get maximum ROI, faster speed-to-market and optimal value out of your telecom search solution, you need a trusted partner with retail search expertise, optimized integration methodologies and Google Cloud specific experience.

You’ve come to the right place!

As a Google Cloud Premier Partner (reserved for the top 3% of participating companies) and an Implementation Partner for Google Cloud Discovery AI for Retail, Grid Dynamics is uniquely equipped to provide all the necessary analysis, customizations, performance and relevance tuning to help retailers connect customers with the products they love in the fastest, most effective way.

Our Google Cloud Discovery AI for Retail services include:

• Trusted implementation approach: With a number of successful migrations under our belt, we have developed a proven Google Discovery AI integration methodology and blueprint that covers all the project needs from initial analysis, through design & implementation, all the way to the customer traffic migration.
• Starter kit and value-add tools: Make the most out of your Google Discovery AI services implementation with our Google Discovery AI for Retail Starter Kit. Get started quickly with data ingestion and change streaming pipelines, integration with clickstream data, data quality framework, results in comparison tools, and much more.
• Merchant’s console: We empower merchants to curate digital customer experience and optimize for business goals with a powerful and flexible business rule system. Provide hints and instructions to Google’s AI algorithms to align results with goals.
• Smart autocomplete: We add intelligence to your search box to deliver personalized experiences that lead to faster conversions with smart autocomplete and instant search. We generate a comprehensive list of autocomplete suggestions, taking into account your catalog data and customer search history, in order to provide highly relevant, typo-tolerant search suggestions.
• Near real-time update streaming pipelines: Track inventory, prices, offers, promotions and customer behavior in real time, and react immediately to ensure that accurate and up-to-date information is always visible to your customers.

Our battle-proven blueprints and starter kits help achieve the smoothest and fastest upgrade of search experience across all channels, resulting in the following benefits for your business:

• 20-30% conversion rate uplift
• 25% less abandoned search sessions
• 30% CTR improvements

Interested in achieving the same results? Get in touch with us to learn more about how we can optimize your digital customer experience with Google Cloud Discovery AI for Retail.