As consumers, we interact with artificial intelligence daily, likely without giving it much thought. Netflix recommends your next binge based on the show you watched last night. Amazon convinces you to buy three more items that align with your buying habits. And ride-share apps like Uber or Lyft analyze your location and connect you with a driver heading your direction. When AI makes everyday tasks easier, the consumer experience feels seamless.
But for many business and marketing leaders AI is the buzzword du juor keeping them up at night. Executives often feel like they should be doing more in the space but struggle with knowing where to start. If you're considering whether and to what extent to weave AI technology into your business strategy, here are some factors to consider.
Researcher Johannes Weber inspects an industrial robot at the Fraunhofer Center for Cognitive ... [+]
1) Weigh ROI before building anything.
"There’s this assumption that you have to build highly specialized, custom machine learning models on your own. In fact, that’s usually not the best thing to do," says Will Thompson, managing director of thought leadership and publisher of the award-winning Forbes AI. For most business needs, partnering with a cloud service provider to leverage pre-trained models is the better bet. Considering return on investment can help guide your strategy. Ask yourself, "How much more accurate do you think you're going to get with your predictions about your customers using AI, at what cost and for what return?" Thompson recommends. "You really need to keep that idea in plain sight or else you risk throwing a lot of money down a rat hole."
2) Tackle process automation first. When deciding where to start, Alorica Chief Marketing Officer Colson Hillier recommends focusing on back-office processes first. Identify the repeatable tasks that can benefit from a level of automation to make employees' jobs more efficient, he says. Using machine learning to automate tasks can make a significant impact on your business, whereas user-facing applications such as voice interactions with customers are much more complex to implement.
3) Test, learn and refine. "It's rare that you open up or deploy an AI system and see an immediate benefit," Hillier says. Start small with a defined use case, apply metrics to measure success and refine your approach, he advises. Hillier—whose company provides outsourcing services to help businesses scale and run more efficiently—says AI needs to be supported within the executive ranks and at a strategic level for the long term. Rather than evaluating the technology as a trend that will fade or a box to check off, "you really need to take a long view on AI." Though building out highly specialized talent, expensive systems, and organizing huge data sets can feel overwhelming, "it is a core foundational capability that businesses in the next century need to embrace," Hillier says.
4) Recognize consumer concerns. A recent research report from Pega revealed that consumers lack trust in AI technology's ability to make sound and ethical decisions. Less than 30% of respondents said they felt comfortable engaging with AI in customer service transactions. To overcome this challenge, businesses must ensure the value their offerings provide outweighs the trade-offs. “Consumers will provide rights to utilize information and intelligence about themselves when they see that directly providing a benefit that outweighs their perceived risks,” Alorica's Hillier says.
Marketers within categories such as healthcare and financial services, which have heightened privacy sensitivities, should take extra caution when using consumer data to create AI solutions, Thompson says. Both Hillier and Thompson agree that algorithms shouldn't be left to run amok—there has to be human intervention to limit potential problems. (Read: Salesforce Explores Blockchain To Stop Robot From Ruining $1 Trillion Opportunity)
Ultimately convenience will win consumers over. "People don't buy products because of the label that's put on them or how the sausage is made," Thompson says. "They buy something that is faster, cheaper or better."
Like some of today's tech industry front-runners, "The company that can find a way to use this technology to revolutionize their business model is going to win, and they’re going to have an exponential advantage over their competitors," Thompson says. "It ain’t easy to get there, though.”
Read: AI Is Predicting The Future Of Online Fraud Protection
