The Great and Terrible God: Where A.I. Succeeds and Fails in The Customer Experience

A.I. is finding it's way into the lives of consumers more than ever, but is it the right tool to help grow your business?

Written by Jamey Bainer
on 07.11.18

There is no shortage of talk on the topic of artificial intelligence. AI is so widely discussed that even kids and our non-tech-savvy grandparents know the acronym and discuss how it is going to impact society and our future — even if they may not quite grasp what it is. In the business world, AI is either a source of excitement or stress, depending on your level of understanding and where you are on the adoption journey.

While everyone scrambles to implement AI into their business, and with so many use cases influencing what you could or should be doing with the technology, it’s easy to lose sight of the end goal and forget that AI has its limitations. At the end of the day for any customer-facing business, all new technology, particularly AI, should be leveraged to enhance the customer experience.

Its application for marketing and advertising is growing by the day. Let’s explore some real areas where AI truly excels, and where it does not.

Ultimate Task-Master, Useful Advisor

AI’s ability to complete simple tasks and surface data and answers on demand is one of its most significant strengths. It can be used with great value to test and determine the best options; giving marketers insight and perspective they may not have been able to see themselves. Done right, it looks like this:



Source: Business Insider / Sephora

Sephora’s app feature, Virtual Artist, uses AI to apply makeup virtually, teach users new techniques, and show how various looks would appear on their customers’ faces. This solution enhances the customer experience by empowering and educating; making make up more accessible to customers who are unsure of what suits them or how to apply it. Educating their audience through AI builds brand loyalty for Sephora, and gives them another opportunity to make sales by listing the products used to create these looks.

Proven, a beauty and cosmetics start-up, is leveraging AI to remove the pain of trying to force a “one size fits all” approach on products that are anything but. The brand creates personalized skincare products based on an AI-powered “largest beauty database in the world.” The result is skincare design based on different product categories, ingredients and review ratings. The learnings are then passed onto cosmetic chemists to craft skincare “proven” to be successful, thus ensuring customers get the best products, and Proven gets products guaranteed to sell and satisfy their customers.

Hopper’s travel app uses AI to predict future flight prices and recommends if users should “buy now” or “wait for a better price.” Since its launch in 2015, the app has sold more than $600 million worth of flights and is currently selling about $1.5 million every day. This is an excellent example of a company using artificial intelligence to vastly improve the user experience. Why spend hours anxiously combing through prices and deals when you can outsource the task to a bot and get a better deal as a result?

Kia, late last year, launched a chatbot on Facebook Messenger designed for customers to have a direct source of information for anything they might be searching for as an alternative to the brand’s more than 800 distinct websites fractured across the web. Since then, the bot, dubbed Kian, has driven three times more conversions than its main website, Kia.com. Introducing chatbots allows customers a 24/7 customer support service where they can ask questions they may otherwise not ask if they had to call in, email or search for answers themselves (heaven forbid). Giving easy access to customers translates to increased conversions.


Source: Digiday / Kia

Where the Machines Fall Short

AI is still in its infancy and to think that it does nothing but solve problems is to put far too much trust in the hands (?) of machines. We are still some time away from AI taking over human power and leaving us redundant. Without human skills and abilities, AI falls short in anticipating and meeting human needs.

One of the most commons reasons AI is falling short of delivering a better customer experience is because marketers rush to throw technology at a problem without stopping to ask themselves what the user really needs. Take the area of personalization, for example; marketers know their customers want experiences personalized to their affinities and needs, and they look to artificial intelligence to provide the solution at scale, but the problem is that AI’s capacity for machine-learning is overly reliant on past behaviors. Our past purchases aren’t always the best indicators of what we want or need now, at this moment.

Examples of this going wrong are endless. All over social media, we gripe about companies retargeting us with products or services that we purchased as a one-off and now receive endless advertisements for, begging us to repurchase. Case in point: An Amazon customer buys a toilet seat and then receives targeted advertisements about more available toilet seats. As this Amazon customer puts it:


Source: Twitter

AI can undoubtedly take requirements and past learnings to create new solutions. However, it is still a long way away from perfect execution. AI still relies heavily on the input of humans to ensure the solutions enhance the customer experience, rather than spit out suggestions to repeat the same past behavior. These attempts fail to meet the standard of personalization customers now expect, especially when brands are using their past purchasing behavior to deliver these suggestions.

Another scary downside of AI is human bias. AI learns from the information input by humans, and I think we can all agree that we are not perfect. Left unchecked your AI “solution” could reflect gender, race, education or political bias, which is more than certain to hurt your brand credibility.

At least for now, AI needs a human hand. Especially when it comes to communication, collaboration, problem-solving, organization, leadership, customer service, and management.

Staving Off The Singularity: Bring People and Machines Together

Where AI truly succeeds is when marketers and businesses use a blended approach of human staff and AI to deliver exceptional customer experience.

Late last year, the Cosmopolitan Hotel in Las Vegas launched Rose, a flirty chatbot concierge that offers guests insider tips and special promotions during their stays. Rose’s playful voice is powered by an AI capable of responding to over 1,000 conversation threads, and the Cosmo’s human communications specialists are on hand to fill in when Rose can’t respond on her own.


Source: Travel & Leisure / Cosmopolitan of Las Vegas 

Stitch Fix owes its rise to prominence in the volatile retail apparel space to a smart collaboration between artificial intelligence and human stylists. In this case, the AI is leveraged to help human stylists parse through thousands of style combinations and customer data points to create the personalized fashion kits that are then delivered to subscribers. It’s a perfect example of using technology to augment employee capabilities, but ultimately leaving the final decision and creative spark to the human brain. Stitch Fix started with one machine learning algorithm, but now leverages hundreds, which are helping the company optimize everything from shipping operations to inventory management.


Source: Author’s Stitch Fix

Lola, the travel app launched by Kayak co-founder Paul English, uses a combination of artificial intelligence and human agents to service the complex needs and preferences of business travelers, from selecting airline bookings that will maximize the traveler’s rewards program earnings to only booking stays at boutique hotels with wine bars.


The best way to figure out how artificial intelligence can help you grow your business is to bring it down to Earth, and the best way to do that is to keep your customer at the heart of what you’re doing. Ask yourself: What problems could AI solve for them? What processes could it help speed up? How could it make something easier and deliver a better experience of your brand and product?

Until the inevitable day when the machines rise up to overthrow their human masters (because we know this is a thing that will happen), it’s all about people, and any technology you choose to employ within your marketing suite should be there to help better the experience of the people who choose to buy from you.