When you think of using artificial intelligence (AI) to improve customer experience, what comes to mind?

Some people might (mistakenly) think of it as trying to replace human beings with Siri or Alexa, or reducing individual customer needs to a faceless demographic. In reality, AI and machine learning are starting to have some surprisingly practical applications for improving customer experience. Let’s look at a few that people are starting to talk about nowadays.

1. Building a better chatbot with artificial intelligence

Years ago, so-called virtual assistants were touted as a way that businesses could save money and reduce the need for human interaction with customers. And while they have succeeded to some extent, some also developed a reputation for mis-interpreting customer requests, giving wrong answers, and standing as an impediment between customers and their needs. This article from uxdesign.cc does a good job of explaining many of the reasons why the chatbot-as-human-replacement strategy often fails.

Today, experts predict that a new generation of chatbots will be focused on dramatically increasing the reach and productivity of human agents. Imagine, for example, an AI-guided application that predictively suggests solutions and pushes information to customers under an agent’s control. In this article from CMO Australia, contributor Vanessa Mitchell bluntly predicts that in the future, the purpose of AI chatbots will be to augment human interaction, not replace it.

2. AI and predictive marketing

Competitive pressures are forcing marketing teams to target their efforts more efficiently than ever. This means that AI can serve as a natural extension to better targeted marketing efforts which, in turn, can lead to happier and more engaged customers. According to CIO Magazine contributor Philip Kushmaro, artificial intelligence now has the potential to predictively improve market segmentation, targeting accuracy, and improved interactions with actual customers.

3. Better content curation

How many of you out there love getting marketing emails?

The answer might be higher than you think when the content truly benefits people. When a shopper gets targeted discounts based on the things he actually purchases, a consumer gets content or offers keyed to a recent life event, or a psychotherapist gets information on new research or publications targeted to her clinical specialty, AI has the potential to help marketing actually make customers happier. In a recent article in MarTech Advisor, Lucidworks’ Lasya Marla notes that by learning patterns in content consumption, artificial intelligence can improve the marketing content you receive as well as your online experience as a customer.

One caveat for AI: garbage in, garbage out

As futuristic as AI sounds, its value is based on a relatively simple premise: AI applications fundamentally learn from data rather than programmed instructions. When a chatbot suggests a solution, or an AI application targets a market, their responses are based on the analysis of large amounts of past data. This is where the fields of AI, machine learning and data quality have begun to intersect.

The promise of artificial intelligence starts to fall apart when bad data leads to flawed predictions. For example, incorrect or bogus contact data – which represents as much as 25% of the average contact database, according to SiriusDecisions – can completely change the interpretation of a demographic and its purchasing behavior. (Of course, we can help with that.) The same is true for all of the other inputs that drive machine learning processes, making formal oversight of data quality more important than ever.

We’ve seen bold predictions about AI and customers in the past – and many of the areas discussed above still remain in the realm of future trends – but there is a renewed sense nowadays that AI now has real practical applications in the customer experience space. More important, capabilities such as these have the potential to become competitive requirements. It is an exciting time in AI, and the field is well worth keeping an eye on from here.

What do you think of when you hear the term “artificial intelligence” (or AI for short)? For many people, it conjures up images of robots, science fiction, and movies like “2001 – A Space Odyssey,” where an evil computer wouldn’t let the hero back on his spaceship to preserve itself.

Real AI is a little less dramatic than that, but still pretty exciting. At its root, it involves using machine learning – often based on large samples of big data – to automate decision-making processes. Some of the more public examples of AI are when computers square off against human chess masters or diagnose complex problems with machinery. And you already use AI every time you ask your phone for directions or a spam filter keeps junk mail from reaching your inbox.

In the area of contact marketing and customer relationship management, some experts are now talking about using AI for applications such as predictive marketing, automated targeting, and personalized content creation. Many of these applications are still in the future, but product introductions aimed at early adopters are already making their way to the market.

Data quality is key in AI

One thing nearly everyone agrees on, however, is that data quality is a potential roadblock for AI. Even a small amount of bad data can easily steer a machine learning algorithm wrong. Imagine, for example, you are trying to do demographic targeting – but given the percentage of contact data that normally goes bad in the course of a year, your AI engine may soon be pitching winter coats to prospects in Miami.

Here are what some leadership voices in the industry are saying about the data quality problem in AI:

• Speaking at a recent Salesforce conference, Leadspace CEO Doug Bewsher described data quality as “AI’s Achilles heel,” going on to note that its effectiveness is crippled if you try using it with static CRM contact data or purchased datasets.
• Information Week columnist Jessica Davis states in an opinion piece that “Data quality is really the foundation of your data and analytics program, whether it’s being used for reports and business intelligence or for more advanced AI and related technologies.”
• A recent Compliance Week article calls data quality “the fuel that makes AI run,” noting that centralized data management will increasingly become a key issue in preventing “silos” of incompatible information.

The ROI of accurate and up-to-date contact data is larger than ever

Naturally, this issue lies right in our wheelhouse. For years, we have been preaching the importance of data quality and data governance for contact data – particularly given the costs of bad data in time, human effort, marketing effectiveness, and customer reputation. But in an era where automation continues to march on, the ROI of good contact data is now growing larger than ever.

We aren’t predicting a world where your marketing efforts will be taken over by a robot – not anytime soon, at least. But AI is a very real trend, one which deserves your attention from here. Some exciting developments are on the horizon in marketing automation, and we are looking forward to what evolves over the next few years.

Find out more about how data quality and contact validation can help your business by visiting the Solutions section of our website.