AI for Customer Experience: 5 Things That Hold Brands Back

It wasn’t too long ago that artificial intelligence (AI) was the buzzword for all kinds of amazing customer experiences. With AI machine learning, intelligent chatbots, and predictive modeling, the impossible is now possible. So why aren’t more brands riding that bandwagon all the way to the bank?

At our recent Customer Response Summit session on AI, a show of hands revealed a surprising number of senior CX executives are still stuck at square one or two on the path to full implementation. Here are the five biggest stumbling blocks they cited (hint: it all comes down to #5).


1. Absence of a clear long-term strategic plan

In their enthusiasm for new tech, many brands plowed ahead on small-scale applications without creating a long-term blueprint for success. “Our approach started out very piecemeal,” said one executive. “We’re looking to bring various elements into a more cohesive strategy.”

2. Lack of ownership and internal collaboration

Too often, the teams tasked with developing AI products are working at loggerheads with the departments their programs are supposedly designed to support. “Sometimes our IT group doesn’t bring us in to ask us how something should work because they don’t like what we have to say,” said another attendee. “They assume, ‘customer service will just figure it out.’”

“We went to chatbots to drive our costs down, but it hasn’t worked out as we’d hoped,” said another. “It was a result of too many cooks in the kitchen.”

3. High levels of complexity within the existing organization

Many brands cited concerns that AI could further muddy some already highly complex organizational structures. “When you’re a Tier-3 vendor, you have 40 different systems to navigate,” said a vendor in the CX space. “We need AI to have these systems talk to each other, but we’re afraid to exacerbate existing problems.”

Said another, “In many cases AI seems to be adding to the complexity, not reducing it.”

4. Predicating resource allocation on achieving short-term goals

“How do you get your organization to be patient with AI?” an attendee asked. “Implementation is a long-term process.”

“The leadership obviously wants a return on its investment sooner rather than later, so it’s a question of managing expectations and providing clear and specific metrics to capture incremental improvement,” another suggested. “But that’s easier said than done.”

“It’s been hard for us to secure the funding to clean and integrate our data without a near-term ROI,” another agreed.

5. Poor foundational data and data governance practices

“How do we get from here to there?” one attendee asked. “Our challenge is the capacity – systems in the support organization were not in place. We’re tasked with laying the data foundation system we need to implement AI.”

“Our customer service organization has a pretty good data foundation but not all departments do,” said another senior executive. “There needs to be more integration. Who ‘owns’ the data?”

“You need to be thinking about your underlying data sources,” another added. “Without that piece, your AI will never function as you envisioned. You must have a strong foundational data element to build on.”


The Bottom Line

Given the importance of building on a strong data foundation for any AI strategy, square one isn’t necessarily a bad place to be. It’s far easier to begin your AI journey by implementing a data governance framework than it is to go back and clean your data and data-storage structures after AI products are already in place.

No single CX Leader can go at it alone. If armed with a long-term strategic plan, a solid support network of cross-functional stakeholders, and clear expectations for incremental improvement, you can gain buy-in for taking that critical first step.