The Right Way to Adopt an AI Strategy

AI in the collaborative space is not a new concept but its role and demand in the midst of a global pandemic have never been more relevant than today. However, having AI and having a strategy for how to use it and drive value, internally and externally, are two separate things. The term “garbage in garbage out” has never become more of an apt phrase than in relation to what data is driving this automated intelligence. Companies must adapt quickly to change, more so today than ever. Rigorously creating an answer and response database from scratch, or using a knowledge base that is not properly vetted, are what most companies are forced to choose from. The key is to understand that going from little or no automation to a full automation agent-less customer experience overnight is not the answer. Below are examples of core concepts for a successful AI journey.

Creating a hybrid model

Having the level of automation versus live agent organically change over time is a key step in this transformative journey. Allowing live agents to tune and correct AI assumptions doubles the expansion rate in terms of what AI can do for your business. Create confidence thresholds which force the AI to engage the agent only when its confidence is below a predefined level as opposed to just handing off the interaction. This way if the AI can handle 80% of the interaction and only 20% requires an agent, not only does your workforce efficiency improve so does the AI learning.

 

AdobeStock_282797991-[Converted]_ai

Provide multiple sources for the AI to reference

Have a variety of sources available as reference points in the consolidated indexed source of truth that the AI draws from is essential. Various departments should be able to continue contributing and on their terms to help ensure ongoing compliance.

Transform and create data models from existing sources

An example of this is using speech analytics to transcribe thousands or millions of hours of recorded audio to text. Use these new textual agent/customer conversations as models for your AI strategy in chatbots, advanced inbound and outbound voice IVRs.

Understand the scope of the AI solution and look for middleware

No one vendor will ever be a one stop shop for AI, and most vendors do not go out of their way to integrate to other solutions. Consider the AI capabilities and if integrations with multiple vendors will be seamless. The goal is to find a platform that facilitates bridging the integration gaps between AI solutions and your collaboration enterprise.

AI in Action

As a partner you need to not only offer product and services, but a roadmap and a strategy for how to take the customer from where there are today to where they want to be – showing how AI can facilitate these efforts. No customer will ever ask for AI, but they will ask for improved KPIs, for ways to do more with less, for ways to leverage the IT investments that they have already made and use those to facilitate something new. A modern example of this would be within the public sector when AI has been used to prevent the need for greater headcount. During this unemployment crisis, there was no way for organizations to simply staff up to handle the volume of interactions that they were suddenly forced to endure. They started by turning to AI to automate simple tasks and provide some level of service 24/7, and gradually expanded AIs role over time to enhance and improve their customer service.

Other examples that showcase the value of AI involve quality monitoring and customer service. AI solutions can learn from hundreds of previous agent evaluations and begin to score agents independent of a supervisor. Instead of a supervisor evaluating 5 out of 500 agent calls, 100% of the calls are scored and the supervisor simply confirms the outliers.

Customers are also using AI now to help with language barriers by employing things like real-time translation services right into the text-based analytics engine already running on all chat interactions. Entire chat sessions can transpire between to different language speakers without either even being aware that the transcription is taking place.

Like any adoption strategy, the goal is not to boil the ocean but to understand where you are headed, what you hope to achieve, and develop the roadmap to get there. All of these concepts and use cases have merit on their own, but the overall synergy that can be accomplished when combined is a staggering advancement to businesses not only in the area of customer service, but all types of workforce efficiency improvements. As a partner, your customer is looking for you to help them understand what is possible and open their eyes to how a well-executed AI strategy can truly transform the way they do business.


 

ABOUT THE AUTHOR
John Romeo brings 25 years of experience and has worked with fortune 100 companies to provide Unified Communications and Contact Center solutions that allow them to establish business continuity and improve business processes. He’s an accomplished Solution Architect, US Patent holder and his experience involves many years of Product Management, technical sales and design.  You can reach John at jromeo@tbicom.com or connect with him on LinkedIn.