This post was originally published at insideBIGDATA.
More is better—more data, more insights, more happy customers and employees. More can also feel like ‘too much’ if you don’t know where to start. Here are four strategies you should consider to help you focus and improve the digital experience for your users.
1. Balance Precision and Recall
Whoever owns site search at your company probably agrees that delivering exactly what the shopper is asking for in a search query is a great shopper experience. And I don’t disagree! However, new technologies have made it possible to not only deliver what they’re searching for, but fulfill their goals.
Let’s use the example of searching for a “wild king salmon fillet” on Whole Foods website. Yes, the search should return the precise results for this precise query. But what if it also recalled results that are relevant to your intent? Your intent isn’t just to find and buy a piece of fish, maybe you’re hoping to cook the fish with garlic and olive oil and garnish with thinly sliced lemons. If the website returns all of those results – no, ‘garlic’ is not the same as a ‘salmon fillet’ – you’re getting what you’re searching for and your intent or goal is fulfilled.
This new way of thinking may require some consideration and negotiation internally. Balancing these two things and widening the net you cast with every search is a fundamental change to today’s definition of precision, recall, and relevance. Semantic vector search, for example, can enrich the results with additional related products to reduce friction in the product discovery journey, shorten the path to purchase, and show your shoppers you understand their goals. In this case, it’s making a delicious, well-seasoned piece of salmon.
2. Apply AI to Your Zero Results Report
It’s a rare and loyal shopper who continues to plug away in the search box after they experience a dead-end in a “no results” page. Wondering how many lost sales can be attributed to null results search queries? Identify the number of sessions that exit from a zero results page. Multiply that number by your conversion rate and AOV and you have a general idea about the money you’re leaving on the table. For most websites that number is quite high. For example, one of the top 5 retailers used semantic vector search and reduced null results by 91%, resulting in a 20% increase in clickthrough rates from search and a 28% increase in AOV.
When reviewing the zero results report, you’ll notice there are a multitude of zero result problems, including misspellings, in stock/out of stock discrepancies, vocabulary mismatch, and products that are commonly searched for but not carried. Semantic vector search will pair up these dead ends with persistent and successful shopping experiences to improve recall.
3. Connect Employees and Customers
A recent Forrester report found that companies are struggling to connect employee experience (EX) and customer experience (CX), but those that do report greater revenue, operational efficiency, and customer satisfaction. Building that connective tissue isn’t so easy—skip ahead to the next section for the “how”; keep reading to learn the “why.”
Let’s consider the relationship between customers and digital merchandisers. The merchandiser has to ask another team member to figure out what’s trending in customer queries, add-to-carts, and conversions. Once they get the answer, they go to IT to get an engineer to update the website and pin popular products so customers can find them easily. Those customer insights inform merchandisers who can curate a more relevant experience. Unfortunately, the back and forth of multiple parties creates barriers and lag time between employees and customers.
Data travels faster than people. Connecting customers and employees means that merchandisers have real-time customer insights available in a way that they can readily understand, and they can immediately upgrade CX in response. Customers win because they enjoy a personalized omnichannel experience that delivers on their goals. Employees win because they’re making educated choices based on customer signals without having to wait for the back and forth of multiple teams. Better CX feeds better EX, which feeds better CX—and this rising tide lifts all boats.
4. Personalize Every Session with Machine Learning
Omnichannel personalization is the best tool you have to make customers feel like they’re browsing with a personal shopper by their side. You could be burning out your merchandisers with manual rule curation to achieve this level of individualization. Customer profiles and in-session signals are two key data sources to create strong and effective personalization. Retailers can leverage NLP (natural language processing) capabilities, signals (collecting, processing, storage), and content/user analytics, to create that special “you-know-me” experience.
With machine learning capabilities merchandisers can easily identify and approve suggestions for low-performing queries. When ML is woven within the applications and user interfaces that employees use every day, they spend more time smiling and less time frowning at work. Empower your search and merchandising teams to get back to tackling larger initiatives that optimize your product discovery experiences, while delivering on omnichannel personalization for shoppers.
Over 50% of shoppers say they changed their shopping habits during the pandemic – buying online when they wouldn’t normally. This is the moment to stay focused on your goals and keep being creative with how you reach them. Retailers that invest in digital experiences across all channels will ride the waves of whatever’s next and win loyal customers.
Contact us today to learn how Lucidworks can help your team create powerful search and discovery applications for your customers and employees.