Skip to main content

Select your location

Search

At the beginning of 2020 I wrote an article about how AI was starting to make a real difference in delivering better search experiences. For most of us the benchmark search experience on a day to day basis is still Google. Elsewhere, search experiences have been slowly improving but are still not good enough. Especially in the retail industry, search abandonment is a real issue, having a direct impact on brand loyalty and billions of dollars of lost revenue as a result.

One nightmare scenario I increasingly hear from B2B and B2C retailers is that customers abandon on-site search experiences and head to Google, where they get better results, then complete their purchase with another retailer.

This opportunity has not been lost on Google. At the height of the pandemic we were lucky to get a preview of Google's new Cloud Retail Search (CRS) product, now on general availability. Google is finally making some of its magic available to all retailers – and to me, this was a step-change in the market and a clear signal that retail search had just changed forever. CRS is the most advanced type of AI-powered search that I called out in my previous blog post. For now, the search for the next generation of retail search is over – it’s already here.

Google Cloud Search for Retail

Product discovery at the next level

CRS is part of the broader Product Discovery tools in Google Cloud and would typically be deployed alongside Autocomplete and Recommendations AI to create a fully engaged product discovery experience. This can go further to include fully integrated image search, voice search and conversational experiences.

But back to the basic point - what is different in Google’s new search product and why should all top retailers take notice?

CRS is Search-as-a-Service and runs on the same global platform as Google itself. This means that search performance, scale and Black Friday are covered as are a myriad of other technical activities that anyone running a typical search engine (including some of my experts here at Kin + Carta) is well versed in. So, with a platform that just works and scales, the focus is on how it improves the customer experience, and ultimately increases revenue for the retailer.


Better customer experience, higher revenue

There are two key areas of differentiation to understand. The first is CRS’s ability to automatically understand consumer intent and the second is how it maps that to product inventory to drive revenue.

From a customer perspective you already understand the first one, it's exactly what you expect when you use Google already. In CRS you also get great results even if you don’t use the same words as those in the product catalog. It’s the feeling that Google ‘just understands’ what you want. The simple fact is that from all of the interactions we have with Google every day, it does. And CRS can apply that understanding to the retailer’s product catalog. ‘Date night dress’ is a great example: these exact words may not be in the retailer’s product descriptions but Google still knows that most people expect this to be a dress which is more fitted, possibly black or red – definitely not one with pictures of dates on it.

Once CRS understands what the customer is looking for, the second differentiator is its ability to drive outcomes – ultimately revenue per customer – by showing products that the customer is interested in, which are also proven to drive these metrics. As a retailer you can focus on increasing clicks, conversions or order value, and CRS will automatically determine how best to rank the results to achieve this. At its highest level of capability, CRS can also personalize results for every user – a clear differentiator for this kind of experience. Using this technology, retailers are seeing up to a 20% increase in revenue per visitor against their existing search platforms.

With these two key areas of functionality in play, merchandising becomes more about directing CRS towards what you want to sell and less about trying to band-aid your search engine with hard-coded rules.

In our own tests with retailers we have observed CRS deliver these ‘wow’ moments very quickly, immediately seeing that some current pain points we focused on in their existing search now just work: “How are you doing that?”

CRS works by seamlessly combining Google’s cloud platform and query understanding with retailer data and advanced machine learning models. We believe that this highly data-driven approach requires an equally data-driven approach to prove its value. Once upon a time when a new search technology came along it was easy to be drawn in by the demos and the sizzle. As experts in both Search and Experimentation methods and technologies we have created a Solution Accelerator approach which will allow retailers to put CRS into an A/B against their existing platform and prove its value. We let the data talk.

CRS is a key component of Google’s broader Product Discovery suite which means that you will be well-placed to further enhance the shopping experience to reduce friction. Product recommendations, voice search and image search are all within easy reach.

Kin + Carta is able to rapidly demonstrate the power of Retail Search and its potential impact on search relevancy and your broader conversion metrics. In just a few weeks our self-sufficient team can ingest your catalog and events data and evaluate CRS against your existing search technology. Although it comes at a cost, it is a fraction of the benefit our customers are seeing from the solution.


We can show you what the future of Retail looks like

Chat with our experts

Share this article

Show me all