Improving Online Retail with Predictive Engagement

It’s fair to say that, as a broad industry vertical, the sheer impact of social distancing regulations took the retail industry by surprise. Ninety-five percent of the industry relies on customers visiting brick-and-mortar stores in volume to browse merchandise and make unplanned purchases. While some retailers had already made digital their primary engagement vehicle, most had to move very quickly to adapt. And that rush has created a less-than-desirable customer experience.

It’s difficult to find a customer who wasn’t affected by a poorly functioning website, delayed shipments because of wrong inventory counts or failed promises. While it appears that storefront shopping is rebounding somewhat as social distancing regulations are relaxed, the pandemic response has permanently siphoned some portion of retail traffic to a digital channel. As such, retailers must step up their games in delivering a better-than-bad online experience.

One immediately obvious component for improving the online customer experience is the item search. It can be one of the most frustrating experiences, quickly leading to an increased bounce rate. According to Stack Exchange, 50% of shoppers will use a search box instead of a hierarchical category selection; Marketing Charts reports that irrelevant and inaccurate results are a web user’s largest frustration. Consider the following:

  • A national pharmacy chain search for “oral thermometer” returns a line of cosmetics
  • A big box DIY store search for “5-inch galvanized nail” returns over 100 entries starting off with 1-1/2-inch nails
  • A sporting goods chain that lists men’s fashion shorts when the search was triathlon shorts

Each example is enough to leave the customer with the impression that, as a retailer, you don’t know your own business. And the search within your website ends there.

Building an effective in-site search engine isn’t easy, primarily because it involves more than IT; it’s multidisciplinary. Marketing has to think about how their customers name items, product management teams need to carefully construct product descriptions on the website and IT needs to continually use analytics to see how visitors are using the search engine — and how it fails.

Phase Out Online Customer Frustrations
More often than not, if a search doesn’t work, the customer will try other search words to achieve a better result. They might do this three, four or five times before giving up. Instead, Genesys Predictive Engagement lets you launch a two-part strategy to address this.

First, predictive engagement lets you learn what that persistence count is for different customers and when they bounce. Second, before that persistence threshold is reached, predictive engagement alerts an agent to support the customer.

Imagine that a customer enters keywords for a footwear style and gets results that clearly don’t match. A second attempt at defining the style is equally disappointing. As the customer begins to enter a third, and potentially last, frustrating attempt, the website pops a chat window that states, “It looks like we are having a problem matching your search criteria. Let me help.” At that point, an agent is engaged by voice or chat to bridge the language the company uses to the language the customer is using. This not only creates higher conversion rates, it also improves customer satisfaction.

Collectively, using predictive engagement in this manner:

  1. Reduces bounce
  2. Demonstrates the commitment to digital retail
  3. Delivers a differentiating capability

But it’s important that the screen pop is based on learning how often a different customer types, and different search types, will continue before being abandoned. It also involves engaging an agent at the right moment — when frustration begins to build. In a furnishings store for example, the search for a refrigerator/fridge might be abandoned if inappropriate responses are returned on only two attempts. In contrast, a search for desk lamp, reading lamp or office lamp might go through five iterations.

Predictive engagement learns to dynamically engage the agent after the first failure for a refrigerator but might only suggest a conversation after the third attempt at the lamp. Using an agent in these examples can maximize the customers’ likelihood for finding the right items with refined searches — and agents aren’t needlessly engaged.

Having a strong digital presence is critical in a post-pandemic world. So establishing a level of engagement that stands above your competitors is key. After all, your competitor is only a tab click away. A Genesys omnichannel engagement solution that leverages predictive engagement capabilities to heighten your performance keeps those competitors further away from your customers.

Read this ebook to learn how you can drive more sales through your website.

Share:

Recommended for you

Genesys ChatLive Assistance
Close Widget

Genesys ChatLive Assistance