MarTech Interview with Antanas Bakšys, Head of Search at Nosto

Antanas, can you tell us a little bit about your background and how you got into the world of search?
Before I founded SearchNode back in 2013, I had an ecommerce site. The search we had access to was really poor – with many visitors leaving the site because they couldn’t find products even though we were selling them, they’d been searching for the wrong keywords or simply that it was slow and inaccurate. That led me to create SearchNode, based on a passion to make ecommerce search solutions available to all ecommerce brands with cloud-based infrastructure, spellcheck, autocomplete, and other big innovations back in the day. The idea was to take a leap forward from first generation ecommerce search solutions which were big and clunky. We’re now at another tipping point in the market—brands today expect search solutions to be automated and constantly self-learning, understanding the real meaning behind the words shoppers enter in the search bar, not just blindly matching keywords, without needing constant attention. That’s why SearchNode joined forces with Nosto and I took on my role there as Head of Search.

Why is onsite search increasingly important to ecommerce businesses?
Essentially, poor search drives customers away and harms revenue for merchants. Our 2023 research with consumers found that 69% of them go straight to the search bar when they visit online retail sites, but 80% say they’ve left a site because of a poor on-site search experience. Losing customers has a direct impact on revenue. Retailers understand that, and attribute around 39% of all traffic bouncing (shoppers leaving immediately) to poorly performing search—or because shoppers can’t find relevant products.

Consumers expect a seamless search experience, yet the majority simply aren’t receiving one. At a time when competition is increasing and the economic outlook is unclear, the shopping experience becomes business-critical. Companies simply can’t afford to leave money on the table through poor search performance. Consumer behavior is also changing—68% said that, with living costs rising, they want to compare products from different online stores, and good ecommerce search makes this easier.

So, search has moved up the agenda for ecommerce brands and is something they know they need to improve on.

Why do you think ecommerce search is still a painpoint for merchants?
Obviously, ecommerce search has been around for the last 20-30 years and brands understand the importance of a positive site search experience. 81% of those we questioned acknowledged that good, fast, accurate site search is essential for a successful ecommerce store and 84% said that they plan to invest in continually improving in this area.

However, as sites have become more complex and customer behavior has changed, the technology has not evolved adequately. This means that managing search has involved a lot of manual work and effort to get right, which leads to the challenges of delivering effective search not being solved. Instead, companies have spent time and resources on other areas, such as their site’s user interface.

Ecommerce brands acknowledge that there’s a problem; 95% say they experience painpoints around their search solution. It’s just that, up until now, it has been too difficult to solve them with existing technology. The combination of the importance of search, issues with current solutions and today’s economic pressures are among the reasons we, at Nosto, have invested heavily in this area.

What particular failings in search drive customers away?
We asked consumers what frustrated them most about site search and the number one issue was irrelevant results. 69% of consumers say they see irrelevant results often, and real world site testing reveals 81% of brand websites return irrelevant results for two-word search queries.

Relevance is the foundation of successful search. However, achieving relevancy is challenging due to the volume of data and the complexity of queries that retailers must manage. Rather than entering one-word search terms and wading through the results, searches are now much more specific. For example, a consumer might search for “blue striped shirt with button cuffs”, rather than simply using the term “shirt”. Shoppers expect that results will be relevant but are often disappointed.

Linked to relevancy, companies need to better understand the intent behind the search terms that are entered. For example, consumers might mistype their query (‘smartpone’ for ‘smartphone’, for example,) or use different forms of a term (e.g. ‘couch’ for ‘sofa’), or use compound words. Internet search engines like Google and Bing are able to cope with these behaviors, so shoppers naturally expect site search to be equally strong when it comes to delivering relevant results.

A second set of frustrations center around wasting the customer’s time. How many times have you been unable to set filters to narrow down the search results, or even clicked on what looks like a relevant result only to find that the product is out-of-stock? These all add friction to the customer journey, irritating consumers and causing them to leave your site—as does slow loading.

Is there a gap between what sites provide and what customers expect? If so, do you know why this is?
There is a gap. In many cases, brands think they’re delivering a good experience, but consumers disagree. Essentially, search isn’t meeting shoppers’ expectations and potential customers are leaving sites that don’t provide them with the right results.

The reality is, merchants are often frustrated prisoners of their current technology—they recognize the importance of search, but their solutions prevent them from meeting customer needs. There’s a pressing requirement to bridge this gap to increase revenue in ever-more competitive markets.

Where should ecommerce sites start when it comes to improving search?
They need to start by putting the right foundations in place and that requires a focus on relevancy. Companies need to understand the intent behind search and that means using solutions that embrace AI and natural language processing technology to make results more relevant. Unlike previous systems that required heavy manual intervention, these automatically learn from customer behavior, removing the load from ecommerce teams.

Our research identifies big gaps between how companies think they are performing, and shoppers’ experiences. That points to a requirement to listen more closely to customers and perhaps spend time carrying out mystery shopper research on your site to see how it really performs in the real world.

Finally, site search improvement is a journey, not a destination. You need to track metrics and continually learn from the results. This isn’t currently happening on the majority of sites—they don’t track crucial metrics (such as the most popular query term that delivers zero results, or queries that deliver the highest revenue, to name a few). Acting on these insights helps ecommerce brands maximize performance and sales.

What do customers expect from site search results?
As a minimum, they want accurate, relevant results! However, their expectations go way beyond that.

Customers want a personalized experience across ecommerce sites, and that includes search. They expect data on their on-site behavior to be used to personalize their search experience—and their search data to personalize their wider on-site experience too. So, if they’ve been browsing a particular type of product, then retailers need to deliver results that match their interests, for example.

They also want more than just product results. Consumers, particularly early in the buying journey, or if they’re purchasing something complex, want reassurance. That means providing them with relevant content in search results such as articles, FAQs, and collection pages, or showing user-generated content such as images or videos amongst product results.

It’s worth noting that younger generations have higher expectations, such as around personalization and being provided with relevant content. So, while this is important to all ecommerce brands, it is business-critical to those targeting younger audiences.

How can retailers combine search and personalization to improve the experience for customers?
70% of consumers said they are likely to complete a purchase if search results are personalized to their purchase history, or specific brands, colors, and sizes they like. So, if a shopper is browsing extra-large clothes, then when they decide to search for jackets, the results should prioritize items with XL sizes available.

Similarly, if they have been browsing on-sale or discounted items, then reduced price products should be more visible (or rank higher) in their search results. Combining search and personalization provides the chance to use consumer data to guide results and drive revenue—yet is under exploited. 66% of ecommerce brands admit they don’t offer personalized search. This again means they’re losing potential sales without realizing it.

Finally, can you talk about any exciting projects or developments that you’re currently working on at Nosto?
Of course! In case you hadn’t heard enough about AI at the moment, I’ll let you in on this one… our product team’s actually working on introducing generative AI features and functionality into our Commerce Experience Platform. There’s going to be some pretty cool features released soon, including: dynamic AI-generated copy for any text fields in campaigns and widgets; AI-assisted image labeling/tagging to automatically identify and organize visual content; and—the one I’m most excited about—AI-enhanced search relevancy through ChatGPT’s large language model. We’ve had a lot of interest since opening out sign-ups for the beta program of this, so are really looking forward to seeing the impact for our clients.

Tune in to MTC Podcast for visionary Martech Trends, Martech News, and quick updates by business experts and leaders!!!