SiteSpect Synopsis: AI and the Customer Experience

February 19, 2018

Share

We’ve been talking about AI for a long time now, but in the past month as articles rank business and marketing trends for 2018 and the new year has us collectively looking forward, there has been even more Internet talk of how AI will affect business practices — especially your customer journey and road to personalization. This week, I’ve collected recent articles about AI A/B testing and optimization for marketers, given you the gist, and a final analysis.

Forbes, What Artificial Intelligence Means To You and Your Business

Contributor Ian Altman outlines ways that businesses can use AI to improve customer experience. His primary argument hinges on the assertion that rather than chilling and homogenizing customer interactions, AI can create warmer and more personal experiences. He notes potential uses for AI in employee training, in understanding sales tactics, and in customer service. Altman finally warns that AI must be implemented properly, and that it is not useful or beneficial on its own.

The Takeaway: “AI technology in the workplace can be a positive or negative development. If used poorly, it can be a waste of money that only changes the image of your business without providing sincere value. If used well, it can drastically improve your employee training and customer interactions and provide value to everyone involved.”

MIT Sloane Business Review, How Big Data and AI Are Driving Business Innovation in 2018

Contributor Randy Bean reports that 97.2% of executives surveyed in New Vantage Partner’s annual executive survey are investing in AI initiatives for their companies. The corporate trend is moving toward using AI to interpret big data and to understand customer behavior. Bean writes, “For the first time, large corporations report that they have direct access to meaningful volumes and sources of data that can feed AI algorithms to detect patterns and understand behaviors. No longer dependent on subsets of data to conduct analyses, these companies combine big data, AI algorithms, and computing power to produce a range of business benefits from real-time consumer credit approval to new product offers.”

The Takeaway: Brands are overwhelmingly excited about and drawn to AI to automate data interpretation and hopefully provide more personalized and targeted Customer Experiences.

Harvard Business Review, As AI Meets the Reputation Economy, We’re All Being Silently Judged

Writer Sophie Kleber discusses here the risks of using AI algorithms to interpret social data on behalf of business and organizations. Our social data impacts us in contexts from hire-ability, to health care eligibility, to social standing. The major risks she outlines are “Bad Data” and “Bad Math.” Without human checks on data we ask AI to interpret, as well as checks on how they interpret it, a lot of the information AI systems provide can be false and unreliable. Kebler writes, “Algorithms don’t have a conscience; they repeat what they learn. When algorithms repeat and perpetuate bias or opinion, we need to consider mathwashing.”

The Takeaway: We must be wary of how and when we use AI, especially when it comes to interpreting social data.

DMN, Alexa May Give Voice to Targeted Marketing

Contributor Ariella Brown reports on Amazon’s recent research into modes of advertising through Alexa and speculates that targeted marketing might appear in Alexa’s answers to user inquiries. This would be in the form of sponsored results, or answers to questions that include particular brand names. These ads pose a huge opportunity for brands as they increase awareness in an actionable context. But, Brown writes, “While more relevant ads may be considered a good thing, both for the brands looking to target effectively, and the consumers who are looking for relevant messages, there are always people who just don’t like the idea of ads creeping in everywhere.”

The Takeaway: Watch for Amazon Alexa using AI to deliver targeted ads, and the potentially negative reaction those targeted ads prompt from consumers.

Context

Brands laud AI as a panacea for marketers, automating the heavy lifting required to put A/B testing data to good use and creating the optimal, personalized, customer experience. Meanwhile, this excitement is underscored by a vague distrust of the accuracy and discretion of machine learning. So, what does this mean for your A/B testing and optimization scheme? Here are a few points to consider: All of these articles — and others not mentioned — generally agree that AI powered features like live chats for customers and trainings for employees can work wonders and provide measurable benefit to the customer experience. However, this depends entirely on the accuracy and unification of the data. Once companies start using AI for too many different tasks, its efficacy becomes much more difficult to attain. For instance, in companies the use of AI for marketing and customer experience optimization requires multiple stages of data collection — meaning multiple points of interpretation — which means your picture of the customer journey is fragmented, unclear, and silo’d. In A/B testing and personalization, it’s crucial to understand the customers experience all the way through their journey, which means both gathering data at each stage of the process and understanding the process as a whole.

Necessity

The discussion around the AI trend further raises the question: Where does my organization need it to get the best out of our digital channels? If you do have an AI implementation it can be leveraged to your benefit in A/B testing and personalization. That said, the best, most efficient approach to A/B and multivariate testing does not need machine learning. Instead, it requires fast, accurate, and actionable information production which fits into your martech ecosystem. If you already do make use of AI, you need an A/B testing solution that can leverage it to enhance your results; an independent view that can automate the entire A/B testing process but realizes work still needs to be done around building the variations, innovating, and driving ROI, without sacrificing control, creativity, speed, and insight on your end. Besides, at the end of the day, even with AI you need a human check on its reporting.

Take Away

AI implementation is exciting and suggests huge changes, tremendous benefits and improvements to the way we run our businesses. But, over reliance in the short term or use at too many distinct points sacrifices a unified understanding of your customer experience. So, if you have AI, take advantage of it. But do it carefully and with a realistic eye toward your A/B testing and personalization needs. Make sure you employ A/B testing on all your digital channels that does not depend on AI, so that it can promise accuracy, speed, and reliability.

To learn more about SiteSpect, visit our website.

Catch Up on the SiteSpect Synopsis

 

Share

Suggested Posts

on-site-searchbp3.jpg

On-Site Search: A/B Test the Functional

bp4.jpg

On-Site Search: A/B Test the Back-End

SiteSpect dashboard

Got Too Much Content? Consider A/B Testing Content Carousels!

Subscribe to our blog: