Upon first glance, the ease of a search engine seems obvious. You arrive at a website, type a word or phrase into the search bar, and get exactly what you were looking for. Right?
Maybe not. That tried-and-true process has begun to feel like a long-winded chore — consumers today are generally less interested in searching and more interested in discovery. Importantly, this subtle shift has enormous implications for businesses.
A few years ago, you might have needed to type an exact brand name, product, color and size to end up with your desired result. Now, most of us don’t feel the need to follow that formula. Instead, we might input a broader category — “sweaters,” “coffee table,” “gardening tools” — and expect that the search engine will do the work for us, ultimately providing us with relevant items that we might not have even known existed.
Implicit in this shift is the rise of personalization in the customer experience (CX). We live in a world where 91% of consumers are more likely to do business with brands that remember, recognize and provide them with relevant recommendations and offers — and where 80% of consumers want personalization in retail to the extent that they would depart for a competitor if a vendor gets it wrong.
When it comes to tailored results, expectations are rising: consumers don’t want to be the ones to do the work of sorting through an infinitely long list of search results anymore. What’s more, they don’t only want to find the one product they had initially wanted to look for, buy it, and move on. Instead, they’re open to discovering new items that they’ll be just as excited about.
In fact, 64% of consumers will immediately try new products or services from companies that offer a high-quality customer experience. Thus, while fast, convenient online commerce experiences are growing in popularity, it seems like we still crave the shopping journey, not just the end result.
Think of a music streaming platform, for example. There are situations (perhaps during a moment of deep focus) when you’re not interested in hearing anything new. But there are other times when you’re more inclined to discover. Algorithms can now discern certain aspects of songs that you’ve listened to and incrementally introduce you to new songs that are similar in certain ways, effectively bridging the gap for you and enabling you to find more of what you love.
The same concept applies when it comes to tailored search experiences. The most effective tools aren’t the ones that simply display what most people have historically been interested in (or the products that have paid for a top spot), but rather the ones leveraging individual user activity and machine learning to deliver relevant results, selected just for that user. Effectively, that’s recommendation — the best of which skips over user-based search to instead provide content that suits an individual’s own taste.
The problem, however, is that most businesses don’t have the resources to build truly personalized, accurate recommendation systems. Bigger, more established companies have been able to progress when it comes to personalizing search and discovery, but still tend to have a long way to go — think of your frustration after buying a refrigerator (the ultimate one-time purchase) only to be met with Amazon recommending you refrigerators for weeks after the fact.
And regardless of sophistication, most of these systems often compromise consumer privacy in order to provide targeted advertising, which simply isn’t sustainable. In fact, privacy laws such as the General Data Protection Regulation (GDPR) and the California Consumer Privacy Act (CCPA) exist because consumers are rapidly becoming aware of, and unwilling to tolerate, these unsustainable strategies.
What we typically see with search engines today is a two-tiered system, in which a user initially experiences pure filtering (think: the results after searching a broad keyword, like “jazz”), and then sees recommendations for the best jazz music for them specifically. Going forward, we’ll start to see the pure search tier slowly fade out, with recommendation becoming more prominent.
This shift will be especially impactful in subjective and highly personal industries, not only with music but also with things like art and fashion. On a clothing website, for example, you can input every search term possible, but you still might never find the perfect item. Recommendation enables us to find that item, faster.
Within the last decade, there’s been an unprecedented amount of investment in AI, including natural language processing (NLP) and computer vision. Much of that has been relatively easy for businesses to implement, but AI-powered personalization is a different story.
Due to the difficulties of real-time deployment of truly individualized recommender systems, it’s significantly tougher for businesses to take advantage of. Yet the stakes are high given the new, pandemic-induced opportunity for recruiting and retaining customers — U.S. ecommerce revenues grew by 32.4% in 2020, more than double 2019’s growth. In this hyper-online world, personalization is the next frontier when it comes to search.
As such, it will be critical for businesses to master the art of recommendation to maintain customer loyalty. That said, it’s unrealistic to expect new technology to immediately be able to guess what a user wants — and companies can’t risk continuing to sacrifice consumer privacy for personalization.
So as the power of the third-party cookie declines even further, companies must begin deploying interactive, dynamic, privacy-conscious systems so that customers can “collaborate” with recommendation technology to efficiently navigate all available options. If companies can keep up, the search bar’s evolution is poised to change the way consumers interact with brands for good.
Dr. Emile Contal is Co-founder and CTO of Crossing Minds, a smart recommendation platform for ecommerce and content delivering recommendations to nearly one hundred million people. He completed his PhD in Machine Learning from ENS Paris-Saclay, the most prestigious French university for Applied Mathematics. Contal has a dozen academic publications, with the impact of his research witnessed by hundreds of citations from prestigious conferences on machine learning and other technical industries.