Why do Shoppers use Virtual Assistants? We Analyzed Over 1,000,000 Conversations to Find Out

By
Sebastien Rankin
February 16, 2021

This article is part of Heyday’s Intel series, where I publish findings from our research around conversational commerce. We analyze how shoppers use conversational AI and how merchants can better leverage conversations, whether automated or human, to increase sales and customer satisfaction. Follow me for more research. 

“Ugh, I emailed support three days ago and still no reply!” 

I could see my significant other’s patience wearing thin. That’s what happens when you need to wait days to get answers to seemingly simple questions. 

“Do they really think this is a good customer experience?” she asked me. 

Hopefully not, but in an age where consumers are hard-wired to expect instant gratification, her all-too-common experience made me wonder: 

What are the most common questions people ask retailers? 

Giving shoppers immediate answers is the name of the game now, but you and I both know doing that at scale is easier said than done. 

But what if merchants knew which questions shoppers were asking and what percentage of total incoming questions they represented?

In theory, knowing which questions shoppers ask most often would help brands create automated responses and serve them faster (this is especially true if they use a virtual assistant.) 

So with that as my North Star, analyzed a sample of over 1,000,000 conversational intents detected by our AI chatbot to better understand the types of questions shoppers ask. 

And the results might surprise you. 

Sidebar: about this report 

Before we dive into the insights, here’s a quick breakdown of how we got our data: 

We analyzed a sample of conversations between shoppers and our AI-powered virtual assistant, looking specifically at the underlying intent of those conversations. Only intents that were detected with a degree of confidence of 70% or higher were considered conclusive. Then, we organized those insights into categories and subcategories to better understand how (and why) shoppers use chat. 

For this report, we analyzed over 1,000,000 intents detected by our AI across five retail verticals, specifically looking at the reasons why a shopper interacts with a virtual assistant.  

We bucketed “reasons” into two primary categories: sales intents and support intents.

We defined sales intents as anything a shopper asks that’s transactional in nature and related to either discovering and buying products (e.g., asking if your store carries a certain brand of jeans.)

We defined support intents as any question a shopper asks that’s informational in nature and related to FAQs like tracking orders, your returns and refund policy, store location and opening hours, etc.  

The goal of our report is to understand the reasons consumers use webchat, what they’re trying to accomplish via chat, and to what degree those tasks change depending on the merchant’s vertical (fashion and apparel, food and beverage, home decor, sporting goods, or beauty and cosmetics.) 

Now that you’re up to speed on our methodology, let’s get into the fun stuff. 

Shoppers mostly use chat for support 

If you haven’t set up your virtual assistant to automatically respond to FAQs, you’re making a big mistake and placing unnecessary burden on your support teams. 

As it turns out, 56% of the intents that our virtual assistant detected over the past year were related to customer support. 

Customers tend to use chat for customer service.

Of those support intents, the ones that came up most often were order tracking (40%), returns/refunds policy (20%), and store hours (15%). 

Just those three FAQs represent 75% of total support intents we detected. And each of those intents can be quickly resolved with automated responses.

Order tracking is the most frequently detected intent.

There's no doubt that fast first-response times are a pillar of exceptional customer service. 

A study from Arise that surveyed 1,500 U.S. consumers found that 80% expect support teams to respond within 24 hours, 37% within an hour, and 16% expect an immediate response. That, paired with our intent data, suggests that consumers are increasingly using virtual assistants to get immediate support, rather than submitting a ticket or calling a support line. 

Consumers are increasingly using virtual assistants to get immediate support, rather than submitting a ticket or calling a support line. 

It seems as though merchants using ticket or queue-based support systems are missing an opportunity to increase their NPS and CSAT scores. By using an AI-powered virtual assistant, they can automate responses to the majority of incoming support requests, resolve queries faster, and meet consumers’ growing desire for immediacy. 

Product discovery is the most common use-case

When you look at sales vs. support intents, consumers clearly use virtual assistants more for support. But when you dive deeper into the intent data, a different narrative emerges. 

As it turns out, 20% of the total intents we analyzed were related to product discovery, making it the most popular intent of all. 

20% of the total intents we analyzed were related to product discovery, making product search the most popular intent overall.

When it comes to shopping online, more consumers are turning to virtual assistants to help them quickly find the products they’re looking for. 

And the transition makes sense. Virtual assistants integrate with a merchant’s product catalog and can serve up products related to a shopper’s search with pinpoint accuracy.

The most popular intents detected (overall).

(Note: “small talk” is defined as expressions of courtesy or politeness. Things like asking a chatbot “how are you?” or saying “thank you” at the end of a conversation.)

Whether the shopper asked a generic question like “do you sell snowboard boots?” to more specific questions (e.g., “I’m looking for Burton snowboard boots”), shoppers are increasingly turning to virtual assistants to find what they’re looking for faster. 

Shopify data collected over Black Friday found that retailers that respond to a customer in five minutes or less were 70% more likely to convert that conversation into a sale. Paired with our data that shows more shoppers using chat as a purchasing tool, it seems clear that virtual assistants and live chat are viable options for merchants that want to increase their online store’s conversion rates and sales. 

And in the absence of in-person customer service, we see omni sales associates and distance selling using a blend of automation, live chat and video consultations being increasingly adopted as we move forward into retail A.C. (after COVID-19.)

How does the data differ from one retail vertical to another? 

While the data above paints broad strokes for how and why shoppers interact with virtual assistants as a whole, the data changes when you segment per vertical. 

With that, I set out to see if the dominant search intents changed depending on whether the retailer was in one of five possible verticals: 

  1. Fashion and apparel
  2. Sporting goods
  3. Food and beverage
  4. Beauty and cosmetics
  5. Home decor and furniture 

Let’s dive in! 

Fashion and apparel

As it turns out, the most common reason why shoppers use virtual assistants when they’re shopping for clothes is to find products, with 18% of all search intents in this vertical related to product search and discovery. 

But it’s what comes next that’s really interesting. 

What do returns and refunds, order tracking, and your shipping policy each have in common?

The top 5 detected intents with fashion and apparel retailers.

They’re all related to order fulfillment, which Shopify’s Future of Commerce 2021 report suggests is the largest pain point exacerbated by the surge in online orders brought on by COVID-19. 

Through virtual assistants, shoppers get instant, automated service whether it’s for shipping, tracking or returning their orders. 

Of course, making it easier for shoppers to find products online will help you close more sales, but smooth order fulfillment anchored on speed and complete transparency from the moment the order is processed till it's at a customer’s doorstep will be a key differentiator for retailers moving into the New Year.

Sporting goods 

Data in the sporting goods vertical suggests that shoppers use virtual assistants to find products more than anything else, commanding 20% of total intents. But while product search still reigns supreme, what’s interesting in this vertical is what came in at #5. 

Conversation routing is one of Heyday's most unique features. With it, shoppers can choose to either talk with a merchant’s virtual assistant, or they can request to talk to a sales associate in real-time via live chat or a video consultation — this is also known as virtual shopping

The top 5 detected intents for sporting goods retailers.

We’ve seen people embrace virtual shopping when shopping for luxury fashion and apparel, cars, and other big-ticket items. But increasingly, we’re seeing virtual shopping being used for technical products like sporting goods. 

When shoppers have specific needs and have to find products that match their requirements, virtual shopping is an excellent way for retail brands to share their technical know-how and give shoppers the confidence to buy online. 

Food and beverage

This vertical is interesting because from the outset of the pandemic, food retail sales in the U.S. grew by nearly 85% year-over-year and, although sales have tapered off from the initial surge in late March, “they remain significantly elevated compared to the previous year,” as per Statistica

The top 5 detected intents for food and beverage retailers.

In our intent data, we similarly saw a correlation between total sales and the volume of online grocery shoppers using our virtual assistant’s product search. When you consider the number of SKUs that grocery stores carry, it makes sense that consumers are using virtual assistants to find products faster.

With our customers in the food and beverage industry, however, we saw an even greater percentage of shoppers choose conversation routing, which suggests a stronger desire for people to talk to a real human when they shop for food. 

Beauty and cosmetics 

Similar to the other verticals we analyzed, people use our virtual assistant’s product search more than anything else. 

What really jumped out to me, though, was the emergence of an intent entirely unique to the beauty and cosmetics vertical: ingredient info. 

The top 5 detected intents for beauty and cosmetics retailers.

When people are shopping for hygienic products cosmetics, they care about what they’re made of and how they’re made. Everyone’s skin is different and, depending on their skin tone, sensitivity and type (dry, oily, combination, etc), not all products will be right for them. 

Our data suggests that, for beauty and cosmetics retailers, assuring that your virtual assistant can give shoppers accurate information on each product’s ingredients can be a differentiating factor in closing more sales.

Furniture and home decor 

For furniture and home decor retailers, we see the trend of shoppers using virtual assistants for product search continue. In fact, we noticed that the more SKUs a merchant carries, the more shoppers use a virtual assistant to find products faster. 

But one data point was particularly interesting: the popularity of shipping policy intents. 

The top 5 detected intents for furniture and home decor retailers.

For the most part, furniture and home decor products are relatively big (think of that new desk and chair you bought for your WFH setup), and there are usually extended delays between when you buy the product and when you receive it. 

Knowing this, shoppers seem to be proactively asking virtual assistants about shipping prior to making a purchase. 

Consumers use virtual assistants as a portal of product discovery 

Based on the sample we analyzed, it seems clear that more shoppers are using AI chat for a mix of sales and support. While support queries are the most common intent category, product search is by far the most popular intent overall, with order tracking coming in close second.

But what really stood out to me is how many consumers use AI chat to find products faster. It raises the question: do shoppers prefer using AI chatbots over traditional search bars? 

We’ll have to run more analysis to know for sure, but as of right now, the data certainly suggests that shoppers use chat as a portal to quick product discovery. 

To recap, here are the main takeaways from our analysis: 

  • Across all retail verticals, shoppers used virtual assistants mostly for product search (20%). This suggests shoppers see chat as a channel for product discovery. 
  • Order tracking made up 40% of all support queries over 2020. This intent is ripe for automation and teams that do so have an opportunity to lower their backlog of customer service requests. 
  • The majority of conversations via chat are related to support requests (56%). 
  • The three most popular support queries are 1) order tracking, 2) returns and exchanges, and 3) store location and hours, which account for 75% of all incoming support requests. 
  • The vast majority of detected intents could be resolved by AI chat, which suggests that brands who leverage conversational AI can drastically improve first response times and customer satisfaction, and online sales. 

Use these findings and your own internal data to find areas of your digital customer experience that are ripe for automation. When done well, automation can result in more online sales, more efficient customer service, and a more convenient digital CX as a whole.

Want to learn more about virtual assistants and virtual shopping in ecommerce? Talk to our conversational commerce experts today to see how Heyday can help you build an exceptional online experience for your customers.