Inning Two Of A Nine Inning Game: What’s Actually Working In Contact Center AI

Leaders in every industry are fielding the same question right now, “What’s our AI strategy?”

In the contact center, that question gets concrete fast. This is where customers actually touch your company, so whatever you decide shows up in the experience right away.

Getting a yes on the budget is the easy part. Teams get stuck on what comes after it:

We put that to a panel in our recent webinar, Beyond the Hype: Practical AI Strategies for Modern Customer Experience.

Fred Stacey, CEO of Cloud Tech Gurus, sits between the buyers, sellers, and builders of contact center technology. He joined LinkLive’s Charlie Goll, a Senior Technical Account Manager, along with our VP of Solution Consulting, Erin Stewart.

There’s real value on the field already, and most of the game is still ahead of us. Here are four things that stuck with us.

Skip the Home Run: Your First AI Win Probably Isn’t a Customer-Facing Chatbot

The instinct is to put a chatbot in front of every customer, but that’s rarely where the first win comes from.

“It doesn’t always mean containment,” Charlie said. “It doesn’t always mean AI is going to do everything for every interaction.”

Teams tend to get further by putting AI somewhere less visible first. There are four places worth looking:

Take the Base Hit: Automated QA Reviews Every Call, Not a Sample

If you want the quickest return, it isn’t the chatbot.

Every conversation gets scored instead of a handful. You catch the coaching gaps sampling would miss, like the agent who handles billing beautifully and freezes on retention offers. You also get trend data worth handing to product and marketing, like the complaint that spiked the week after a fee change.

Read the Scoreboard Right: Containment Isn’t the Stat That Matters

Plenty of teams still grade their virtual agent on containment, meaning how many conversations never reached a human. That number tells you nothing about whether the customer got what they came for. A caller who gives up and hangs up counts as contained.

Measure resolution instead, the same way you’d measure an agent.

One myth also needs retiring. “There won’t be any human agents left. That’s not the case,” Fred said. Automation absorbs the repetitive volume while your people move up to the harder conversations.

Spring Training Comes First: If Your Agents Can’t Find the Answer, Neither Can AI

Ask why an AI project stalls and it’s almost never the technology.

That’s the whole problem in one line. Most centers have twenty versions of the same process, and the best answers live in the heads of the longest-tenured people. Charlie watched schedulers dig through OneNote just to find how each individual doctor wants appointments booked. None of that is written down anywhere a machine can reach.

Feed that to an AI and you get exactly what you’d expect, which is garbage in and garbage out.

Get Your Reps In: Three Things to Do This Quarter

The good news is that no vendor can fix the knowledge problem for you, which means you can start on it today. All three of these happen in your own house, before anyone signs anything.

Play the Long Game

Every theme from the hour pointed the same way. The wins that hold up are the unglamorous ones, and the obvious metric is usually the wrong one.

The work that decides whether any of it lands happens before a vendor is ever involved.

Second inning is a good place to be. There’s plenty of time to get this right, and a lot less time to keep pretending the groundwork isn’t the work.

Want the whole conversation? The session covers the parts we couldn’t fit here, including legacy phone systems, who owns AI now, and what a token actually costs you.

Watch the webinar on demand