Magazine Article | August 29, 2017

Chatbots In Retail: 7 Ways To Ensure Success

By Mila D’Antonio, Principal Analyst — Customer Engagement, Ovum

Chatbots look to transform the way retailers engage with customers. But success relies initially on thoughtful and strategic deployment.

In recent years, the retail industry has arguably been at the forefront in the adoption of digital technologies to meet the rapidly changing expectations of consumers. The adoption of AI-assisted chatbots is no exception. Chatbots that assist customer service and enable conversational commerce are poised to bring tremendous innovation in the way brands connect with customers.

Eventually and inevitably, chatbots will disrupt how retailers engage with consumers, triggering a long-term paradigm shift in how consumers will interact with machines. It’s important, therefore, for retailers to consider what degree of investment in the solution will provide the optimal outcome in customer satisfaction and deliver a superior customer experience. They also must ensure the chatbots are aligned to their overall business strategy to safeguard their enterprises economically.

As companies realize the quick ROI of AI-powered chatbots brought on by automation and instant response, the market will see rapid deployment of them, as companies substitute live support with virtual agents. Those ready to harness the disruptive impact of AI and enter the path to progress should learn to embrace and integrate AI into their current business models.

Before companies start building their chatbots, they should clearly identify the tasks that their intelligent virtual agents will perform. It’s also crucial to ensure that the other aspects and features of the AI platform make sense for a company’s business model and consumer base. Whether it’s business process automation through first-level queries or more in-depth customer engagement deployments where chatbots can predict intent and make transactions, chatbots should add value and deliver superior customer experiences. For customers, that translates to ease of business, frictionless experiences, and personalization. If the chatbot should encounter roadblocks and fail to deliver on these expectations, customers may become frustrated and defect. To ensure chatbot utilization and CX optimization, retailers should make considerations for the following:

1. OPTIMIZE THE CHATBOTS TO LOOP HUMANS INTO THE COMPUTING. If chatbots need to handle complex tasks, a blended AI model, in which agents are ready to step in, works best. This is especially important for complex transactions or nonstandard inquiries, which require additional consideration. Customers can direct their queries toward an agent, but if the question is simple, the bot can answer directly. If the chatbot has a high confidence of not being able to answer the question correctly, it can prompt agents with canned answers that they can use. Along the same lines, after the bot answers each question, it could ask the customer if they want to speak to an agent for more clarification. In either case, it’s important that the agent can access the previous chat history so that they can pick up where the bot left off.

2. STRATEGIZE PROPER ONBOARDING APPROACHES. Providing tailored suggestions to keep the customer engaged can be useful. Devising helpful tips and examples on how to use the tool can ease any potential customer frustration. Onboarding should be a guided experience to help customers understand how to use the bot in a friendly way. To get customers in the habit of conversing with their chatbots, retailers must ensure they proactively reach out to customers with information, insight, and advice — presented at the right time and location and based on predictive analysis of customers’ individual needs.

3. THINK STRATEGICALLY ABOUT HOW TO REORGANIZE THE CONTACT CENTER AGENTS. Companies should build their chatbots to focus on the most pervasive service issues and automate engagements without reducing customer satisfaction. Agents, in turn, should focus on more high-value tasks. Chatbots’ current role is to augment agents, not replace them, essentially acting as an extension of the agent. Chatbots will only replace some of the low-level tasks that agents do, like answering routine questions. This frees up agents to focus on more high-value, revenue-generating responsibilities.

"If chatbots need to handle complex tasks, a blended AI model, in which agents are ready to step in, works best."

Principal Analyst, Customer Engagement, Ovum


4. ENABLE THE CHATBOTS TO DELIVER CONTEXT AND PRECISE KNOWLEDGE. This entails delivering context through an open and connected platform that continuously integrates behavioral data with historical and transactional data. Finding a solution that easily integrates into existing business processes, systems, and consumer care solutions can provide a comprehensive repository of data that the AI solution can use to better understand consumer sentiment, as well as customers’ intent and nonverbal communication cues. H&M, for example, uses a chatbot that offers different outfits to consumers based on browsing cues and leads them to purchase through the messaging.

5. BUILD CHATBOTS TO INTELLIGENTLY FORMULATE A RESPONSE BASED ON CURRENT ONLINE BEHAVIORS OR ACTIONS. Incorporating information about customers’ digital behavior before, during, and after their chatbot interaction helps to improve the performance of the chatbot. A better-informed chatbot can deliver tailored answers directed at the user’s experience. Not only do chatbots excel at collecting customer data and delivering personalized experiences, they can assist the agent in also delivering tailored information with the data they collect.

6. ENSURE THAT THE CHATBOT MONITORS, TESTS, TRACKS, AND IMPROVES. AI-powered chatbots will become a critical part of the customer journey, as automation, omni-channel engagement, and personalization become key components in the path to purchase. Retailers must ensure the chatbot platform provides capabilities in areas such as integration, security, management, or monitoring, in addition to the ability to integrate with messaging platforms, enterprise systems, and NLU (natural language understanding) systems. Chatbots must provide monitoring capabilities to track different commands and responses of users. They should also monitor the customers’ behaviors and actions and build up natural language flows to use in processes, while submitting them to the knowledge management system.

7. CHATBOTS SHOULD INTEGRATE WITH MESSAGING APPS. Chatbots embedded into messaging platforms like Facebook messenger, Line, and Whatsapp lend themselves to the types of conversations customers are accustomed to having with their friends and families, which are asynchronous. Such integration also allows consumers to revisit past conversations, and agents can have a single chat history at their fingertips, no matter where the customer leaves off. Chatbots integrated into messaging platforms enable retailers to engage in conversational commerce, which involves the confluence of chat, messaging, artificial intelligence, and natural language understanding to enable shopping and transactions via messaging. It essentially brings the point of sale to the customer. For example, Sephora’s chatbot allows Kik users who message the retailer to provide more information about themselves through a quiz and then offers personalized beauty tips, product recommendations, and reviews. It then allows them to purchase the products they reference in chat without leaving the Kik app.

As AI-assisted chatbots position themselves to revolutionize the customer experience, enterprises must ensure they align tightly with their specific business goals and objectives. To be truly effective and thrive in the long term, retailers must strive for their chatbots to be agile, scalable, and omni-channel in nature. Eventually, chatbots will transform the way retailers interact with customers and become a vital component to the shopping experience and an essential element to staying relevant and in business.