Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

The diamond workflow for customer service AI

What consumers want most is consistency. Here's how to ensure your customer service AI is helping—not hurting—your business.

The diamond workflow for customer service AI

[Adobe Stock / Iryna]

Profile photo of Steve Pruden

Hey, have you heard about this thing called AI? Of course you have. We’re being beaten over the head with it. And I’ll be the first to raise my hand and say I’m one of the causes. But there’s a reason for it—and it’s not so that it can create a picture of you in space or write a poem for your spouse on their birthday.

As we move into the new year, the efficiencies and effectiveness of your operations provided by AI will be critical for scaling. This is especially true when it comes to customer service, as the resources and outputs needed to keep up with consumers’ heightened demands and broadening engagement continues its upward trajectory. 

There’s a reason to focus now. At some point in the near future, customer-facing AI, such as for customer service—in the operational sense we’ll talk about here—will become table stakes. And that’s when the returns dwindle. 

It’s the same for any marketing advancement. A decade ago, personalizing an email or web page with the customer’s name or an image of their city used to resonate as an added touch that built loyalty. Now, it is so assumed that a brand’s email will be personalized to you that it barely registers. Similarly, free two-day shipping was a differentiator about a decade ago, before the Amazon effect forced the majority of brands to adopt the same. 

The companies that can begin to leverage these technologies successfully are going to have a leg up on the competition, both in the bottom line (operational savings) and customer satisfaction (more beneficial experiences). AI should be treated like any other technology; it should either prove itself by saving you money or helping to grow revenue. If it doesn’t, you’re wasting your time. 

Maximizing your return requires a proven strategy. I call ours the diamond approach because the workflow it creates has to follow two paths but start and end at a centralized point to form a diamond.

Here’s what it looks like:

AT THE TOP: CUSTOMER NEEDS

Before you can start down your workflow paths, the entire organization has to align itself at a single starting point. And that starting point should be centered around your customers. 

Don’t buy any technology, integrate any systems, or launch any AI agents until you unpack the needs of whomever you are trying to serve. No one wants a chatbot for chatbot’s sake. Ask yourself: What are you trying to accomplish? The answer is whatever your customers need from you. Are they frustrated with how long it takes to resolve an issue? Are they struggling to find other products of yours? 

Use design thinking to solve for the path your consumers take to reach this point, their pain points along the way, and what incentivizes them to engage. 

LEFT PATH: CENTRALIZED DATA  

AI is only as good as the data that informs it. And again, your AI needs to actually do something to improve the customer experience.

Take what you’ve learned from step one—what makes your customers tick and how they want to engage—and make sure the data that informs those behaviors is being collected, is clean, and can be accessed and pushed out through all of your different channels. If a customer performs an action in one channel, every other channel needs to have that information to push out the most relevant experience possible. 

RIGHT PATH: AI INFRASTRUCTURE

A ton of work has to be done in the back end to have that seamless, interactive experience on the customer-facing side. Selecting the right technology vendor is going to be critical to your success. Like your data in the other path, AI has to sit on top of all of your platforms and work together as one in real time. 

Often, it’s easiest to leverage as few vendors as you can. For example, if you already have Salesforce’s Commerce Cloud and Service Cloud platforms for your e-commerce site and customer service software, it’s an easier experience to place Salesforce’s Data Cloud on top for AI that inherently connects. But that’s not a dealbreaker. It’s possible for integration between vendor software to work just as well, it just takes a few extra steps. What is important is that the AI is frictionless, because…

AT THE BOTTOM: CX ACTIVATION

What consumers want most of all is consistency. They want an interaction on a mobile app to feel the same as on the e-commerce site or in a physical store. Similarly, when it comes to customer service, talking to an agent online should feel the same as calling a phone number or getting help from an in-store associate. 

Along the same lines, the goal for AI service reps is that they feel like real people. Do you know the brands that are using AI customer reps the best? Probably not, because it should be difficult to tell when done well. Vice versa, your human customer reps need to be armed with the same information and resources as the AI bot. 

That’s why the centralized data and connected tech ecosystem are so critical. A customer’s history and profile should showcase the same information no matter who (human or machine) or where a company’s rep is accessing. 

AI is a place of exploration for organizations, and companies should absolutely give their employees the freedom to learn and test in this space. But you also need guardrails to keep from flying off the edge. The worst thing you can do is hire a group of data scientists and tell them to go. There’s a reason these roadmaps exist, and if you create the right plan and get in early enough, you’ll reap the benefits.