Establishing a successful contact center today is very different from what it was just a few years ago. Having worked with businesses to help them master the customer experience for two decades, I’ve never seen such a rapid transformation in how these centers need to operate. So, I spend a lot of time handling questions from business leaders about it.
Interestingly, startups and long established organizations often ask the same thing: “How do I set up a new contact center?” New businesses recognize they can no longer follow templates of the past. Established businesses, meanwhile, face what I call the “Frankenstein’s monster problem.”
Many businesses originally set up their contact centers years ago, when available technologies and customer expectations were very different. Over the years, they’ve added on new functionality -- for example, new tools that let them interact with customers through texting, apps, and video, or tools that give them AI capabilities. But, what they’ve ended up with are big, clunky tech stacks composed of different tools that don’t work well together. Ultimately, customers feel the difference.
So, I tell today’s businesses that, whether they’re setting up a new contact center or looking to redo an existing one, the steps to take are the same.
Outcomes first
In a workshop for Blackstone portfolio companies the other day, I explained the importance of where to begin: Concentrate on outcomes. Don’t focus on which technologies to use at first. Think through, in detail, what you want your contact center to achieve for individual customers who have specific intents.
For example, imagine that part of your company serves bereaved relatives who need to deal with a loved one’s affairs. Do you want someone mourning a loved one to call your company, only to discover that agentic AI is handling their call? Almost surely not. That is a time to ensure a live human being handles the call and provides the most empathetic, swift resolution for whatever the person needs.
On the flip side, if you’re a CPG (consumer packaged goods) company, a caller might be chasing up an order and just needs some basic information about its status. Do you want to make human agents handle that? I’d question your sanity if you did. In this era, that's the perfect example of something AI can generally handle, and customers are generally happy.
Come up with as many use cases as possible. Consider carefully what the results of each contact should be. At the end of a call or written exchange, what precisely will have happened? How will the person feel about the experience, and about the brand?
Seek a unified solution with top-notch segmentation
With this in mind, avoid Frankenstein from the beginning. Go for a single platform that combines all channels and functions into a holistic design. Make sure it’s one that can successfully and smoothly serve individual customers on very different journeys.
Choose an AI-powered solution that you can teach easily. In testing it out, set about introducing it to all of your different use cases. Tell it what your goals are for each interaction with any customer with each type of need.
Run tests to see how well it performs. Don’t cross your fingers and hope; put it through repeated exercises. Each time, tell the platform what it did well and poorly. It should keep learning, refining, and doing better each time.
Data breeds personalization
As you get it up and running to deal with real customers, be sure the tool has access to all of their available information. It should pull together the records of every interaction the customer has had with your brand across every touchpoint. It should use automatic transcription of calls and videos; LLMs (large language models) and NLP (natural language processing) to understand what the customer has said. And it should apply sentiment analysis to understand the emotions that the customer has indicated.
This is vital because, when a customer reaches out, the platform should instantly know who that person is and what their intent is most likely to be, and use that information to direct the call or message to the most appropriate place. This is a level of segmentation that was never possible before.
The customer record must also show how each individual has responded to previous interactions, and advise the human or AI agent on how to use the current interaction to improve over those experiences. This helps your brand achieve the hyper-personalization customers seek.
I will note that I’m proud to have a professional relationship with Nextiva, and recommend their UCXM (unified customer experience management) solution to organizations all the time. But my opinions are entirely my own.
Whether your company is large or small, the same strategy applies for setting up a contact center. By following these steps, you can experience results and success like never before.
About the author: Wayne Butterfield is Partner for AI, Automation, and Contact Center Transformation at ISG . An automation pioneer, he guides businesses through digital transformation, as well as cognitive and AI automation, in ways that transform customer service. An early adopter of RPA in 2010, he created one of the first enterprise-scale RPA operations. For more than 15 years, he’s been at the forefront of new technologies and pioneering solutions. He helps ensure organizations sidestep pitfalls and swiftly harness the power of modern technologies for efficiency and growth.
Edited by
Erik Linask
