A recent study by Juniper Networks indicates that retail sales from chatbots will nearly double annually, reaching $112 billion by 2023, which is a lucrative outcome of the automation of customer sales and support processes. Chatbots are just one example of how technology is being used to create intelligent content and perhaps what is becoming the future of e-commerce.
Online retailers are also investing big in artificial intelligence-driven systems, such as smart CMS, that allow them to deliver individualized experiences, an increasingly crucial part of customer retention and driving sales.
Rasmus Skjoldan, chief marketing officer of Magnolia CMS, discusses with Digital Journal how new AI/ML features in modern technology are changing the way retailers create an online customer experience.
Digital Journal: How is artificial intelligence impacting on retail?
Rasmus Skjoldan: Retail is a seasonal industry, and because of that, there can be large peaks and valleys in customer questions and associated ticket volume throughout the year. This is a perfect use case for where digital transformation can have a massive impact, because digital communication channels and automation in the form of chatbots and AI can help alleviate the pain associated with seasonal upticks in volume.
It can be hard to scale live, phone-based support systems for such variations in traffic, and because of this, customers often have to wait on hold to get help — a terrible experience. Messaging, on the other hand, allows conversations to continue without the customer having to wait around for a live agent, and bot-based automation can take on a lot of the heavy lifting for customers with routine requests like checking on a shipping status.
These two pillars of digital transformation — messaging and automation — are critical considerations for retail. By deprioritizing phone volume in favor of digital interactions, agents will have additional bandwidth to respond to the messages that can't be resolved through bots, and also have multiple conversations at a time. Because of this increase in agent productivity, retailers can offer a better customer experience, without having to hire as many seasonal agents. To take it one step further, embedding phone right within digital channels can really be a differentiator for brands moving forward. For example, when there is phone backlog, customers can be offered the option to continue the conversation over messaging without having to start over.
DJ: How important are chatbots becoming for retail, and what advantages are retailers seeking to leverage from chatbots?
Skjoldan: The biggest retailers are able to automate so much more in this era of digital transformation because they can integrate bots with back-end systems, like an existing CRM or shipping provider. These integrations allow them to remove the agent from a lot of conversations. By assigning common inquiries to chatbots, retailers can remove some of the burden from live agents. In terms of staffing around the holidays, bots can handle a lot of the increased volume, so that retailers don't have to hire as many seasonal agents.
DJ: Generally, how do customers react to chatbots?
Skjoldan: Customers want their issues to be resolved quickly and effortlessly. If a chatbot is able to accomplish that, then most customers are happy. Frustrations arise when chatbots misinterpret the issue at hand. It is important for brands to use chatbots to cater to the customer's desire for convenience, while also offering the option to speak to a real agent when the chatbot becomes an inconvenience.
DJ: What are the main limitations with chatbots?
Skjoldan: Many chatbot vendors today are relying on a purely conversational bot experience, and it can be very difficult and time consuming to train these AI models. To circumnavigate this issue, organizations can use conversational AI technology to classify the ticket, coupled with decision trees that are deterministic and far more effective in resolving use-case specific issues.
DJ: How can technology help to overcome these limitations?
Skjoldan: If retailers keep their automation and bots decision tree-based, then the retailers are controlling the conversation. This is a simpler form of technology that is easier to manage. If AI determines that the customer inquiry can be handled through an existing bot workflow, then your customer moves through a set of predetermined tasks instead of the customer trying to have a conversation with a bot. If the customer's question does not match a predetermined workflow, the back-end software will connect them with an agent.
DJ: What will customer service look like five years from now?
Skjoldan: The customer service industry is already seeing massive improvements in the efficiency of CX through automation, and I believe in the next five years, that progress will only be magnified. Brands will automate more than 90% of their customer service and reserve agents for their most complex issues. To get there, the industry will need more data scientists, engineers and analysts to maintain models and create bots that will ensure a great customer experience. Automation will work towards improving efficiency of service — and more significantly increase revenue for brands as a result.