Retailers invest massive amounts of money and effort to attract shoppers to their websites, but underinvest in the search functionality on their sites. Improving site search offers incredible return on investment, both to delight customers and to increase revenue.
According to eConsultancy, nearly a third of visitors to retail sites use the search box. Some studies estimate that two-thirds of these visitors will bounce if the search is poor. For large retailers, this means a loss of tens of millions of dollars every month just from underinvesting in search.
Investing in site search isn't just about recovering lost sales, however. Great site search is a foundation to building inspiring and personalized customer journeys. It offers retailers a unique opportunity to make a strong first impression on shoppers, establishing a framework for long-term relationships. Conversely, a mediocre or poor search experience leaves money on the table and even encourages consumers to check out the competition.
The good news is that advances in artificial intelligence (AI) and deep learning are changing the retail experience in every way, and search is no exception. While e-commerce giants like Amazon.com and eBay have used these technologies behind the scenes for years, AI is becoming more and more “democratized.”
Here are five things every retailer can do today to optimize their site search and dramatically improve their customer experience:
- Combine analytics with human evaluation. How can retailers tell whether their search is effective? Measuring good search is critical and should be done using both objective measurements and human evaluation. Analytics are the most scalable way of measuring how customers are experiencing search, however, human beings can provide an additional — and important — layer of quality control.
- Use semantic search. Semantic search allows customers to type or speak naturally and receive relevant results — the way they might shop in an actual brick-and-mortar store. A semantic search engine can intuitively understand the shopper's intent. This dramatically improves visibility into the store catalog and allows sites to show relevant results even on long-tail queries (e.g., "women’s sneakers size 7.5 for under $50"). On average, sites with semantic search bars experience a 2 percent cart abandonment rate compared to the 40 percent cart abandonment rate reported by sites that use a text-based search bar.
- Invest in autocomplete. Autocomplete (i.e., finishing a query a shopper has begun typing) is more than just a way to reduce the consumer’s search effort; it also guides shoppers toward queries that perform well. This helps to ensure a successful — and profitable — search experience.
- Evaluation of result set size. One indicator of effective search is the size of the overall results. Search that returns no results points to a problem with recall — a good match is likely out there, but nothing shows. On the other side of the coin, queries that return huge result sets point to a lack of precision. In both cases, the underlying problem is a failure of understanding user intent. Great search doesn’t over provide or under provide returns results; it brings back everything that is relevant — nothing more, nothing less.
- Invest in the overall search experience. When you’re working on hard problems like ranking and query understanding, it’s easy to lose sight of details like the size of the search box and the layout of the search results page. Don’t assume you know what is and isn’t important. Try to treat each decision about your search experience as a testable hypothesis.
So, this holiday season, don’t leave money on the table. Give yourself — and your customers — the gift of great site search. Your customers and your bottom line will thank you.
Amir Konigsberg is the CEO of Twiggle, a company that helps e-commerce search engines think the way human beings do.
Related story: Why Chatbots Today Are Being Built Incorrectly by Businesses
![](https://www.mytotalretail.com/thumb/?src=/wp-content/uploads/sites/14/2017/12/Amir-Konigsberg-CEO-Twiggle.jpg&w=51&h=51&c=true)