3 Guiding Principles for Digital Commerce Search in the Future

Customers want brand experiences that understand them beyond a superficial level. We know that keywords and classic identifiers do not necessarily paint the entire picture of who the customer is and what they want to do.

Imagine if you went into a furniture store and told the associate you’re looking for an armchair and they only brought you to the chairs that have ‘armchair’ in the product name. There are likely many more products that would have suited your needs—what about recliners, swivel chairs, or adirondack chairs? Digital commerce experiences need to go beyond superficial indicators and deliver the same breadth of offerings that a great associate that understands your intent would point you to in-store.

Doing this is easier said than done, but the market is showing clear signs that this is the way forward. Lucidworks CEO Will Hayes and GM of Digital Commerce Peter Curran sat down with the CommerceTomorrow podcast to discuss the current state of digital commerce search, why the old lexical methods of search are on their way out, and why powerful technology like semantic vector search is on its way in.

Listen to the full recording above and read on for a few key takeaways:

Search needs to consider concepts, goals, and intent.

When customers log into a website, they are not thinking about the exact query needed to get the exact result they want from their search. They are more often coming with a goal in mind – to stock up on toilet paper, learn something new, or simply to get a sense of their options. Search has traditionally not handled this sort of ambiguity very well. What most often happens is users end up bending to the way the application works, and mapping their intention to the results. Moving forward, this needs to be the other way around. Every brand and application will be better served by understanding who the individual shopper is and what their intention is.

For example, say an office manager searches for “small conference room table and chairs.” Conceptually, this likely means that they intend to outfit a room that fits three to six people, with each person having a chair and matching table. While this concept may seem straightforward to the average person, manually adding rules to deliver the right results for that type of goal-focused search through keywords alone can become incredibly complex. Cue semantic search.

Lexical search is out, semantic vector search is in.

Most search engines rely on the use of keywords, or lexical search. Given the demands of the past year and the push toward better, more personalized digital experiences, this method is not sufficient in parsing the intentions of customers. Search engines need to get better at dealing with ambiguity and, as mentioned above, concepts. Semantic vector search bypasses the barriers to understanding intent that lexical search struggles with by mapping out many more factors in determining the intent of the customer’s search. This allows for some incredible things to happen—such as searching in a language that is different from the website, and still receiving results. Or searching for a product that isn’t in the catalogue, and still receiving something close to what would satisfy the intent of the customer.

The idea is that semantic vector search can allow brands to more accurately deliver meaningful search results that don’t rely on the customer having to do the work or on employees to manually tune the search results. One of the top five retailers deployed Lucidworks semantic vector search solution, Never Null, and decreased null results by a whopping 91% compared to the previous year—that translates into hundreds of millions in sales.

Signals are key to democratizing digital experience insights.

Signals are the combination of data points and interactions unique to an individual customer. This includes search history, previous additions to their cart, and more. Most importantly, signals help brands understand and predict their customers’ intentions and goals. Signals provide valuable data that, when applied across a brand’s multiple channels, can further strengthen all touchpoints of the digital commerce experience.

For example, insights gained from signals can be utilized not only by marketing teams, but also by product development or logistics teams. When insights are democratized across all parts of the brand, the connected experience can be more quickly optimized to match the signals in real time. Brands should capture signals wherever customers are and apply those insights everywhere to continuously improve the customer’s experience every time they return.

Want to learn more about how semantic vector search can help your brand create personalized digital experiences that make your customers feel understood? Contact us today for a demo.