After Baymard Institute completed a large-scale usability study on e-Commerce search in 2014, the web usability research firm declared it was “tempted to declare the current state of e-Commerce search ‘broken.’” When test subjects were assigned product-finding tasks on various sites, they were unable to locate the item almost one-third (31%) of the time.
When shoppers can’t find what they want within the first set of results, they don’t rethink their searching behavior and try different strategies. After all, these people have high expectations for the search experience after years of using Google’s AI-powered engine.
“In the absence of good search results, the average user will simply think you don't have what they're looking for and will go somewhere else,” explains Lauryn Smith, Senior UX Researcher at Baymard.
Even though Baymard’s last major benchmark study on e-Commerce search was a few years ago, “search is always on our radar,” Smith notes. Has retail site search improved? “We haven’t seen a lot of change,” she admits.
There are exceptions to the generally mediocre state of e-Commerce search, and Amazon is the most exceptional, according to Baymard. No wonder — Amazon has a legion of software developers immersed in artificial intelligence (AI), machine learning and deep learning, as well as a mindboggling amount of data from every click and page view made by almost 200 million visitors each month.
The future might look bleak for retailers that aren’t Amazon — if it weren’t for thousands of brilliant software engineers who have contributed their brain power and creativity to open-source projects like the Apache Software Foundation. Thanks to their collective efforts, AI technology has been advancing beyond the gilded doors of monopolies like Amazon, Alphabet/Google, Microsoft and Facebook.
What that means is that most retailers have the opportunity to provide their customers with search functionality on par with and perhaps better than Amazon. It also means that retailers that don’t address the sorry state of search for their customers will quickly be left behind.
The Road To AI-Powered Search
Let’s take a brief look at three technological advancements that have paved the way for a reinvention of e-Commerce search.
First of all, cloud computing is a big deal. Infoworld sums it up nicely, describing the cloud as “a virtualized pool of resources, from raw compute power to application functionality, available on demand. The key advantage is agility: the ability to apply abstracted compute, storage and network resources to workloads as needed and tap into an abundance of prebuilt services.”
With the cloud, retailers are no longer constrained by the limited capacity of their own hardware, software and programmers — giving companies the ability to scale in an unprecedented way.
The cloud also has unleashed the creativity of the programming community. As lots and lots of people gain access to computing power, technology advances much more rapidly than it did before.
Flowing from the cloud is the second important advancement, open intelligence. There has been a huge explosion in the “big data” companies are generating, collecting and trying to gain intelligence from. To deal with this data, especially unstructured data, function-specific workflow and analytics tools had to be developed.
The easiest way to understand this is to compare open intelligence to the software-as-a-service (SaaS) model. With SaaS applications, you can get convenience, but you trade off flexibility because the application provider puts tight constraints on what the application can and cannot do.
There are two inherent problems with SaaS. One is that the providers own the data, so a retailer can’t see inside the “black box” to understand and learn from their own customers. The second is that SaaS customers can customize a solution, but they can’t rewrite it. They have to trade flexibility and extensibility for speed-to-market.
For a business, the question is always how technology advancements can be turned into useful applications. That is where production-ready AI comes in and where a company like Lucidworks operates.
The beauty of production-ready AI, as the term implies, is that it gives e-Commerce companies access to state-of-the art AI tools that they can deploy immediately. AI is said to “learn” from the data it is fed independent of human bias, and a retailer that incorporates AI into search applications will be able to take advantage of the intelligence that has been developed through millions of searches.
Critically important, a retailer can see how these technologies make the decisions and change them if necessary. AI doesn’t “think” like a human, which is one of its advantages, but it’s important to make sure that machine logic doesn’t overrule human logic to the point where the needs of the business aren’t being met.
With the democratization of Amazon-like search, consumers will very quickly begin expecting a much better experience wherever they shop. At that point, any retailers still consoling themselves by saying that search has been an afterthought across the industry will be standing alone — and that’s a scary place to be.
Prior to joining Lucidworks, Vivek Sriram led product and business development for the search business at Amazon Web Services, where he co-created Amazon Elasticsearch Service and helped grow the search business from inception to large scale. In his 20+ years in software Vivek served in engineering, business development, and product roles at various companies including Lucidworks, Vizu (acquired by Nielsen), Aggregate Knowledge (acquired by Neustar), and SBI Razorfish.