Conversational Interfaces: Gimmick or Gateway?

In my line of work, you get to see a lot of product demos.

And I mean a lot of product demos.

I can’t begin to tell you how many agent-copilot and generative-insights-type demos I’ve seen in the past year. I won’t say too many, but it’s close.

It’s literally my job to think about what technologies contact center teams need to be successful, how new innovations in artificial intelligence will impact how teams work now and in the future, and what innovations have legs and are poised to stand the test of time.

It’s also literally my job to question not just the “what” and the “why” of these technologies, but also “how” they manifest in the user interfaces contact center teams use every day.

Any technological breakthrough carries the potential to revolutionize how we interact with the world around us. Conversational interfaces as a modality have been kicking around for a while, but it’s generative AI that put them in the spotlight.

In some ways, they’re a transformative “gateway” to more natural, accessible, and intuitive forms of human-computer interaction. But—and you had to know a “but” was coming—conversational interfaces are not the right choice for every use case and in some cases end up feeling like a gimmick.

I’m going to say something that some of you aren’t going to like: I really don’t think your agents should have conversations with generative bots. Their whole job is to have a conversation with your customers—does it really make sense for them to have a simultaneous side conversation with AI? This isn’t the way.

For tasks that require precision, speed, or complex multitasking, a conversational model can be unnecessarily tedious. The novelty of talking to a machine wears off pretty quickly and the need for a more appropriate interaction model becomes painfully obvious.

I understand how we got here, but I need to this call out: This is a crutch. This is temporary.

It’s familiar, and it’s easiest to slap a conversational interface on a product and call it a day. I understand the desire to capitalize on market trends. But we need product leaders to be willing to envision what interaction paradigms could exist beyond the conversational for this type of work if we are going to see meaningful adoption of generative AI in the contact center (beyond post-call summaries, of course).

Shifts are already on their way. You will begin to hear (perhaps more than you ever wanted to hear) about new “agentic AI” systems that will enable AI to take action and make complex decisions on its own. This evolution hints at a future where our interactions with AI might become less about chatting back and forth and more about seamless collaboration and decision-making support.

So are conversational interfaces a gateway or a gimmick?

Both, I’d argue.

In the world of contact centers, a feature that begins as a quick-fix gimmick can evolve into a gateway that unlocks new ways of working. Just as these gateways risk becoming yesterday’s gimmicks if innovation stagnates in the face of advancing capabilities and acknowledgement of user needs.

Advice? We have to keep questioning the tech we adopt, not just for its flash but for its function. Not all that glitters is gold, or whatever.

OK, that’s enough analogies for this columnist. See you next time!

Christina McAllister is senior analyst, Forrester Research, covering customer service and contact center technology, strategy, and operations.

Related Articles

Contact Center Analysts Aren’t Immune from Baffling Service Experiences

Knowing exactly where a company went wrong doesn't lessen the frustration.

When AI Is Ready, Will You Be?

The human touch will become a rare opportunity for differentiation.

Contact Center Catharsis

Here are some of the things callers want to vent about.