23 Jul, 2018 A.I. Improves CX and Sales – If You Do This
You may be wondering if Artificial intelligence (A.I.), along with the merging of humans and machines, can improve the customer experience, agent productivity, generate revenue, or possibly lower the cost associated with supporting customers.
Netflix suggests movies based on your watching habits, Amazon recommends products based on prior purchases and companies use it to follow up on potential leads. Let’s face it, if you ever use Google for anything, you have engaged some form of A.I.
Most people embrace A.I. as it makes their lives easier. The questions for many organizations include: is A.I. a good solution for our need? Will it improve account management and / or customer experience (CX)? If so, how do you effectively deploy it? According to Humach CEO, Tim Houlne, you can absolutely utilize A.I. in most settings, if you understand the role of the human in the process.
“A.I. is the automation component that can simplify and streamline your processes. To make it friendly and ultimately effective, requires leveraging humans to deliver the personal characteristics and creativity,” says Houlne. “Humans must provide an editorial as well as ethical review to ensure the A.I. provides the desired outcomes.”
Humans are required to tune the machine, whether it is aligning the natural language processing or matching analytics and algorithms to achieve desired outcomes. Natural language processing is already widely used through simpler applications, such as chat bots and SMS messaging. When deciding to use A.I. for meatier projects, such as account management and improving customer service, it’s important to incorporate building a brand persona.
Who Is Your A.I. Representing?
Humach, like many organizations, utilizes an automated outreach platform that receives and interprets an interaction, makes a decision on the appropriate response, and reacts accordingly. For example, if the interaction is one where routine information can be disseminated, the A.I. responds accordingly. If the interaction requires an escalation, the A.I. performs said escalation or requests confirmation on a response.
“When deploying A.I. on more interactive situations, human intervention is crucial. The A.I. is trained to escalate based on a yellow or red scenario, looking for feedback,” says Houlne. “If it receives the same answer for an escalation say ten times, on the eleventh it knows how to respond, and the interaction becomes green, meaning the system has learned the appropriate reply. This is impossible without intervention from educated agents who know the best answers and can support the A.I.”
Another critical component to ensure success of your A.I. is to develop a persona to ensure the A.I. properly represents your brand.
“We worked diligently to not only ensure that the answers are appropriate, but that the responses are done in a tone that properly represents our brand,” says Houlne. “Our team is very friendly and helpful, and now with the persona we have developed, so is our A.I.”
Houlne says the A.I. works so well, people actually call and ask to speak to her.
“Building a persona for your A.I. is crucial and should not be glossed over. If you don’t give it a name and build a personality, then the interaction will not be positive. The idea is to improve account management by automating appropriate processes, but in such a way as to also improve customer experience. They are not mutually exclusive.”
Design with the End in Mind – The Customer Experience
Incorporating A.I. into any area of your business requires a well thought out design process. It’s not a set it and forget it scenario, and it’s crucial to consider the impact to CX.
“We frequently hear stories of organizations that were given marching orders to check out automation and move toward incorporating it with the idea of 100 percent containment, meaning that automation will handle 100 percent of the issues it receives,” says Houlne. “This is both unrealistic and dangerous as it will negatively impact the customer experience.”
To ensure your A.I. is deployed effectively requires a deep dive into the transactions it will handle, and the expected outcomes needed for positive CX. If there is a customer challenge that occurs that will take multiple steps to resolve or require creative problem solving on the fly, these should never be automated.
“Poor design is a major contributor to bad customer experiences. It’s crucial to uncover the challenges and ensure that the user interface supports a positive experience via any A.I. application. Think Amazon or Uber. They make it easy,” says Houlne. “They deliver on their promised brand experience.”
Poor design is also a reason many A.I. deployments are abandoned. It’s poor utilization, rather than broken technology.
“It’s interesting that we are still discussing the importance of escalations and customer experience when it comes to any transaction, automated or otherwise. This is not a new issue. Think of voicemail jail, where people just started hitting zero to get out of a voice tree and to a person,” says Houlne. “The premise is the same, but the application is slightly different in that it’s technology in the form of cloud-based platforms, internet and apps, rather than the telephone.”
Technology will enhance a well-designed program, allowing it to move more quickly and improve CX. Or it can exacerbate a poorly designed program, leading to customer frustration and even loss of revenue. It’s important that organizations keep in mind that there are four components to any A.I. rollout – technology, process, design and human intervention. Without all four, your A.I. will most likely cost you customers.
Find Your Ideal Branded Customer Experience
Adding technology, particularly one with the sophistication of A.I., to an existing infrastructure is difficult. As more organizations move to the cloud, there are less internal subject matter experts to deploy on these types of initiatives. Fortunately, there is a way to incubate and learn without negatively impacting your current environment.
“We built Humach Labs to allow companies a simple way to incubate new technologies within their ecosystem. Additionally, we intentionally designed it to reduce professional services that are often required for these types of deployments,” says Houlne. “This creates the fastest path to a long-term strategy and roadmap aligned with corporate goals.”
Any organization that has deployed technology realizes that during the roll out, you uncover a bevy of processes that don’t work, challenges that were unforeseen and internal people who, intentionally or not, bring a negative light to the situation.
“Our services are built to ensure that a company is ready to go through the re-engineering required to deploy a technology platform like A.I. Additionally, we are able to identify those snipers within an organization that can disrupt the success deployment of the technology,” says Houlne. “Plus, you uncover challenges without any negative impact to your existing ecosystem or team members. It’s definitely a winning way to develop complex technology deployment strategies.”
To learn more about Humach Labs, contact us today.