Your Highest-Intent Customers Are Being Let Down by Search

When a customer visits your website, they know what they want, i.e. they have come with a certain intent. They try to translate their intent into keywords, and what they often get in place of a solution is the dreaded No Results page. The problem, however, is not necessarily that the product they are looking for doesn't exist. In many cases, the product is there, but the website simply doesn't understand what the customer means.

This is becoming an increasingly important problem as the way people search for information, products and solutions changes. Customers are becoming accustomed to describing what they want in natural language, rather than trying to guess which keywords a search engine will understand. As AI-powered search becomes a normal part of how people research and shop, D2C brands need to consider whether their own search experience is keeping up with these changing expectations.

Cart vs. Search Abandonment

Cart Abandonment

Cart abandonment has long been one of the greatest issues for e-commerce brands, especially considering how often it occurs and how much is lost due to users abandoning their carts. Without citing any particular source for this, it is well-known that cart abandonment has been sitting at around 70% for the past 15 years. Discussions on how to reduce this number are constant, and brands have been trying different proactive and retroactive tactics to change it, but with little to no effect.

A still-unrecognised way to address cart abandonment is to implement Conversational Commerce. To resolve cart abandonment, most brands try fixing the checkout experience itself and, while it does solve part of the problem, with a large percentage of people reporting checkout-related issues as the cause for abandoning their shopping cart mid-purchase, it still doesn't address shoppers who abandon a cart due to insecurity. Conversational Commerce can help address some of this uncertainty by giving shoppers access to the information and reassurance they need before completing a purchase. 

Search Abandonment

While cart abandonment is a popular topic of discussion in e-commerce circles, search abandonment has not gained much attention, or at least not as much attention as it should be given. 

As much as ~50% of users turn to search as their preferred product-finding strategy, while only 44% of desktop/mobile e-commerce sites and apps have decent or good Search UX [2]. Additionally, nearly 9 out of 10 consumers, or 88% of them, consider search as one of the essential features of an e-commerce website. This gap between how people prefer to search and the fact that so many websites still have significant issues with their search experience is truly surprising. 

Image 1: On-site search performance tested on 183 e-commerce sites, Source: Baymard Institute [1]

Poor search UX not only results in unsatisfied and frustrated users, but it is also a serious cause of website abandonment. In fact, Baymard's research shows that 68% of e-commerce sites have a No Results implementation that is essentially a dead end, meaning that when a customer's search fails, they are often left with little more than a generic message asking them to try again. [3]

This is a particularly problematic experience because the customer has already given the brand an important piece of information: they have told it what they are looking for. Instead of using that information to help them find the right product, the website effectively puts the responsibility back on the customer and asks them to change their search.  

Image 2: Search abandonment statistics, Source: Google Cloud [4]

This accounts for an experience with devastating consequences for retailers, as search abandonment costs them $2 trillion annually globally [4].

Why Intent Matters More than Keywords

When a shopper enters “dress for a summer wedding” into a search bar, they are not necessarily looking for a product that has those exact words in its product title or description. They are communicating an intent. They are looking for a dress, but they are also telling the brand something about the occasion, the season and, indirectly, the type of product that would be appropriate. Similarly, a customer searching for “something for dry skin” may not know which product they need at all. They are describing a problem they want to solve.

This is where keyword-based search starts to fall short. The customer is thinking about what they want to achieve, while the search engine is looking for words that match the product catalogue. 

Baymard's research identifies several different types of e-commerce search queries, including product type, features, use cases, compatibility, symptoms and non-product searches, and shows that many websites struggle particularly with the more complex searches that require some interpretation of the customer's intent. [2]

For a D2C brand, this can create a particularly frustrating situation. A customer might search for “something to wear to a beach wedding” and receive no results because the website does not have a category called “beach wedding outfits”. That does not mean the brand doesn't sell anything suitable. It might have linen trousers, shirts, dresses and sandals that could all be relevant to the customer's needs, but the search engine is unable to make that connection.

The customer, however, doesn't know that. From their perspective, the website has just told them that it doesn't have what they are looking for.

AI-Powered Search Can Close the Intent Gap

This is where AI-powered search can make a significant difference to the D2C shopping experience. Rather than simply matching the words in a search query to words contained within product titles, descriptions and categories, AI can interpret the meaning behind the query and use that information to identify which products are actually relevant.

Similar to the example above, a customer searching for “a warm shirt for an outdoor wedding in autumn” is giving the website much more information than simply the product type. They have specified that they need a shirt, but also that it needs to be warm, suitable for an outdoor environment and appropriate for a wedding. 

Image 3: Etsy search results for complex search query, Screenshot

A traditional search engine might struggle to use all of this information, particularly if the exact combination of keywords doesn't exist anywhere in the product catalogue. An AI-powered search engine can instead understand the different parts of the request and use them to identify products that match the customer's overall intent.

This becomes particularly valuable when there is no exact match. An AI-powered search experience could identify relevant products across different categories. Instead of treating the absence of an exact keyword match as the absence of a solution, it can interpret the customer's intent and find products that could satisfy it.

This changes the role of the No Results page. Rather than marking the end of the customer's search journey, it can become an opportunity to recover that intent and keep the customer moving towards a purchase.

AI Has Changed What Customers Expect From Search 

With the rise of AI, online shopping behaviour has drastically changed, especially when it comes to researching and comparing. While, just a year ago, we could call this new shopper a Gen Z shopper, this has now become behaviour characteristic of all shoppers. 

Generative AI has taught us that we do not have to translate our complex thoughts into a couple of keywords hoping the results will match our intent. We can describe what we want in the same way we would explain it to another person - our very own personal shopping assistant. 

The reason this new conversational search has become so popular so quickly is, very logically, because it is more natural to us. We are no longer adapting to machines, but the machine has learned to adapt to us.

This is also changing what customers expect when they return to traditional e-commerce websites. After becoming accustomed to AI tools that can understand context, nuance and natural language, a search bar that only recognises exact keywords can feel increasingly limited.

Search Is Customer Intent Data

There is another reason D2C brands should pay more attention to search abandonment: the search bar is one of the clearest sources of first-party information about what customers actually want. 

Brands can analyse: 

• which queries produce the most searches, 
• which searches return no results, 
• which queries are repeatedly reformulated, and 
• which searches result in product views, carts and purchases.

A zero-result search for “gift for new mum”, for example, could indicate demand for a type of product that the brand doesn't currently offer. 

This data can therefore be used for much more than improving search. It can reveal gaps in merchandising, product development and content, while also showing brands how their customers naturally describe their needs. Customers are effectively telling brands what they want, how they describe it and which problems they are trying to solve.

Conclusion

When a customer uses the search bar, they are often demonstrating a relatively high level of intent: they have identified something they want and are actively trying to find it. If that search ends with a dead-end page, brands may be losing customers not because they don't have the right product, but because they failed to understand the customer's request.

AI-powered search offers a way to close this gap by moving e-commerce search away from simple keyword matching and towards understanding intent. Instead of forcing customers to translate their needs into the language of a product catalogue (and a new one for each particular brand), brands can allow them to describe what they actually want and use AI to connect that intent with the products that can satisfy it.

Resources

[1] Baymard Institute. "UX Benchmark." Baymard Institute. Accessed 2026. https://baymard.com/ux-benchmark

[2] Baymard Institute. "Ecommerce Search Query Types: The 12 Query Types Users Rely On." Baymard Institute. Accessed 2026. https://baymard.com/blog/ecommerce-search-query-types

[3] Baymard Institute. "No Results Pages: 18 UX Best Practices." Baymard Institute. Accessed 2026. https://baymard.com/blog/no-results-page

[4] Google Cloud. "New Research on Search Abandonment in Retail." Google Cloud Blog. 2023. https://cloud.google.com/blog/topics/retail/new-research-on-search-abandonment-in-retail