How Bots Can Generate Value For Enterprises

Omer Biran

While enterprises have been closely watching the latest machine learning (ML) innovations over the past few years, the time has arrived to stop watching and start implementing this transformative technology. A recent study found that 48% of companies using ML has said that the top benefit is profitability, but the benefits go even beyond just making additional revenue! ML applications like chatbots can also improve the customer experience and reduce employee workload.

When we think about how the nature of communication has changed over the years, apps like Facebook Messenger, WhatsApp, and Slack instantly come to mind. There’s clearly a shift in the way we share things with others — messaging apps surpassed social media in active users as early as 2015! We’re getting more comfortable with communicating like this in our personal lives. And now, we’re starting to expect the same services from businesses we buy from. However, corporations often aren’t up-to-date with new trends.

It’s time for enterprises to embrace this trend and get comfortable with conversational AI or risk losing customers. For example, take the challenge of customer support optimization. With the advancement of mobile technologies allowing customers to chat while they’re cooking dinner or out at the store, it is now easier than ever for unhappy customers or customers with questions to get the help they need. This means customer-support inquiries across industries have significantly increased in recent years, causing many representatives to become overwhelmed and unable to provide the type of support customers expect. This puts the company and individual representatives at a great risk of customer churn, which doesn’t look good for anyone.

With advances in machine learning, can the technology be used to help customer support teams more reasonably manage their requests? The answer is yes. Tools like chatbots are very efficient at automating easy conversations and simple requests and can reduce some of the burden customer teams face. To get started with chatbots, enterprises must first consider the following:

  • Determine your business value: Before building a chatbot, teams need to ask: “What is the business value of this?” Only once you know the problems the business is trying to solve can you begin to think about the development of the bot. Once you have answered this question, you can take out the drawing board and build a concept.
  • Treat chatbots like the products they are: A chatbot in itself is a product with an interface. If you want to use data in your bot that you don’t have in your back-end system, it won’t work. Chatbots are not an all-knowing source of a data; they are simply an interface. Developer teams need to establish what data already exists in the back-end system before expecting outputs from a bot.
  • Decide on your connection: For customers to actually use chatbots, you need to understand which platforms customers actually use already. If you don’t know where your customers are, you won’t be able to make your chatbot available to them. Once you’ve done this, you can decide if your chatbot will be online, in emails, in text messages, or on messenger applications. It’s critical to put your bot where the people are.
  • Chatbots cannot be compared to humans: While a chatbot can help manage customer support, it is very different from a human instructing you to do something. Chatbots for customer support are capable of handling very specific conversations and should not be expected to know what your mother wants for her birthday, or what you would like to eat for dinner that night. There is a global misconception that machine learning is capable of learning things and developing a consciousness, but really, it can only accomplish what we have taught it. No chatbot can surpass the human brain. As humans, we’re aware of everything around us, and chatbots are not.

Once you have thought about these aspects of chatbots, your organization can then implement this technology to enhance the customer experience. In customer support specifically, chatbots can:

  • Autonomously manage certain topics (invoice-related questions)
  • Prequalify inquiries (the chatbot prequalifies the lead and asks for key contact information and then forwards to the human to take over)
  • Dispatch requests (the chatbot receives every customer request and dispatches it to the correct service)
  • Facilitate automatic replies (the chatbot sends a basic reply redirecting the customer to a website page or resource for their specific question)

By enabling fast and specific support for customers, bots can ultimately increase customer satisfaction, which leads to enhanced customer loyalty, increased productivity, and revenue. Businesses across industries should be looking at how chatbots can not only improve their bottom line but improve customer and employee experience. By implementing this technology in one area, companies may just see the results they want to spread across the organization.

Learn more

For more information on SAP Conversational AI and how to automate your customer support with bots, please reach out to me at

This article originally appeared in DZone and is republished by permission.

Omer Biran

About Omer Biran

Omer Biran is managing director of SAP Conversational AI (former CTO of Recast.AI). Omer joined SAP through the acquisition of startup Recast.AI in early 2018, and today manages the SAP Conversational AI unit of the SAP Leonardo Machine Learning portfolio. With a MSc in IT and computer science as well as a PhD in mathematics, Omer worked as a software engineer at Intel and as a solution manager at Sungard before joining Recast.AI.