• by Kevin Lindsay , May 3, 2018

Artificial Intelligence is everywhere. And if pop culture’s to be believed, you might even be falling in love with your favorite device a la Joaquin Phoenix in “Her.” AI, it seems, is just that ubiquitous and compelling.

But even though it’s seemingly everywhere, AI still feels oddly elusive and the actual steps to effective adoption and integration of AI might feel a little murky. 

Regardless of where you are on the AI spectrum, the lowest-hanging opportunities to incorporate AI tend to fall under the marketing and customer experience umbrella. Within that universe, there’s a real connection between meaningful AI and service.

A good example is Swisscom, which recently realized there were some lags in its call center, specifically, from a visitor experience and efficiency perspective. By analyzing customer behavioral data, Swisscom could see exactly where customers were experiencing friction in the self-service process and leverage simple AI solutions to assist. This enabled Swisscom to improve delivery speed by 25% and lowered web management costs by 20%, all while boosting conversions by upwards of 150%.

Regardless of your course of action with AI, you do need to do two things to be successful: plan the work and work the plan.  

While AI has been hyped for years, it’s still a brave new world in terms of actually applying it to real-world scenarios. But early adoption gives companies a very wide, very longrunway in terms of experimentation without needing to prove heavy-duty ROI right off the bat. 

At the same time, companies need a plan. If they integrate AI because they feel backed into a corner, then they’re setting themselves up for disappointment at best. Companies should be experimental with AI. As long as they know the business problems they’re aiming to solve, they’ll be well-positioned for some early-on wins, even if they’re not quite all-in just yet.

My biggest piece of advice here: don’t just toss money at this latest-and-greatest buzzword. AI adoption is about making a deliberate, decisive move to improve workflows, customer experiences and, with it, the bottom line.

So, how do you start integrating AI into your business? Here are a few solid jumping-off points:

• The elusive marketing tasks. Think about marketing tasks that typically require heavy duty insights, know-how and subject matter expertise. Usually, these tasks, such as which offer is best for which type of customer, have been manual until now. Start with a very practical, very tangible ‘job’ and go from there.

• Clear-cut payoff tasks. When you can definitively show a clear-cut return on investment and effort with AI, you’re well-positioned for deeper support and deeper pockets. But, don’t lean on squishy metrics. This isn’t the time for how things feel or seem. Opt for concrete data points, such as lift, conversion rate or the fact that your AI efforts reduced calls to the call center driving cost savings. That’s powerful.
• Small-scale tasks. If there’s any trepidation around AI or if you’re simply scared it won’t work, launch your first efforts as a small-scale experiment. Don’t send all of your traffic to the AI experience – maybe one is more randomized, one is more curated, and one is full-on AI. This will ensure you aren’t rolling the dice too much while, at the same time, gaining comparison points for the different experiences.

And from there? The sky’s the limit. According to Harvard Business Review, 30% of early AI adopters say they’ve increased revenue leveraging AI to gain market share or expand their offerings. There’s a lot to be gained, but businesses should start small, consider true business goals and be deliberate, even in the experimental phases of AI adoption.