Artificial Intelligence, Personalization, and Ethics: What You Need to Know

If you keep attuned to the latest technology news, then you’re likely familiar with the onslaught of headlines reporting new uses of machine learning (ML) and artificial intelligence (AI).

And if you’re like many organizations, you feel some inkling of “we should be doing that.”

Rest assured, however, that “today’s AI systems are decidedly not on par with human intelligence,” writes the acclaimed ML and AI expert Lance Eliot in an article for Forbes. As Eliot goes on to explain: “The computationally based AI of current times can do some rather tricky and impressive things, but it is at best ‘narrow AI’ that can only work in limited scope or domains, and furthermore lacks any semblance of common-sense reasoning that we associate with human intelligence.”

So the question remains: should your organization dabble in ML or AI? If so, how, and to what extent?

These questions are good ones—and not easy to answer. As we tell our clients: it depends. First off, we don’t ever advise implementing a new technology only for the sake of the technology. The use of technology always needs to follow the human need. In other words, as Jakob Nielsen of the Norman Nielsen Group puts it in a keynote talk, the human need must drive the choice and use of technology.

With these questions answered, you can then move on to assessing the array of technologies available—and how they might help you do a better job of, say, building more meaningful relationships with potential or current clients. Often, we find businesses, associations, and organizations asking these three basic questions: (1) How can we attract clients or members? (2) How can we help them realize the value we offer? (3) How can we continue to keep them interested?

These questions, of course, correlate with three critical points along the sales pipeline, with the first question (How can we attract clients or members?) relating to lead generation; the second question (How can they realize the value we offer?) dealing with lead nurturing; and the third question (How can we keep them interested?) involving customer retention.

Answers to all of these lie in an area that today’s AI and related technologies can, to a certain degree, address: personalization, or the effort to tailor your messaging and content to the individual. Of course, we’ve all been on the receiving end of utter failures in personalization, whether it’s the texts from the healthcare provider harassing us to schedule an appointment, long after we’ve done so, or the smarmy ads that shadow us as we browse the internet.

Today, personalization is far more than a trend. At Atstriata, we think of it as a movement—and work with clients to find and implement the right AI-based technologies to make it happen, from marketing automation and customer relationship management (CRM) platforms to chatbots and unique landing pages. And while we could write an entire book series on personalization across healthcare, association, B2B, education, and related areas of marketing, here, we’d like to focus on the ethics of personalization in marketing, an area that gets overlooked but is critical not only to your organization’s reputation but also to your return on investment and overall success.

We could spend hours sharing our thoughts on ethical personalization, but for now, suffice it to say that we will at least share some ideas to kickstart your thinking.

1. Understand the line between personalization and privacy.

Personalization is fueled by data collection, and in today’s Wild West of data collection, in which rules and regulations are ever-evolving, it is easy to cross the ethical line. Consider, for instance, that data collection comes with liabilities. If your organization opts to collect, then you must also opt to take responsibility for keeping data private and using it only at the behest of customer or client permission. Not doing so means you can be held legally liable.

The solution? Be transparent and honest with clients about how, exactly, you will use their data. Regardless of circumstances. Always. Period.

2. Beware of sensitive data and information.

Not all data and information are created equally. For example, people care more about the privacy of matters relating to their health and well-being, their intelligence, their income, and their greatest fears and challenges than, say, their favorite brand of socks or granola bars. “Health issues are one of those things people don’t want others to know about, and we will find that targeted ads [and other uses of personalization] violate our integrity or are intrusive,” writes Stockholm-based marketing expert Anna Loverus in an article, “Can Marketing Personalization Be Unethical?”

Given the amount of health data companies and organizations can acquire from users Googling symptoms and seeking online medical advice, many find themselves inundated with health data—and faced with the moral dilemma of where to draw the line. What can (or should) you do?

Let users drive and control what tailored content they receive—and not the other way around. For example, if someone searches on your site for “innovative cancer treatments,” don’t start sending them cancer-related advice willy-nilly. Give them a menu of options and always, always, always let them opt-in and choose what they receive.

3. Up your content wheelhouse.

Personalization in marketing isn’t only fueled by data. It’s also powered by content—a lot of content. From lead generation pieces to blog posts, email campaigns, case studies, e-books, white papers, social media, and yes, even prehistoric print pieces, your organization or association needs to shift to the mindset of being a content wheelhouse, if you haven’t yet done so. But keep in mind, too, that the last thing you want to do is push out a bunch of mediocre or substandard content. Quality matters. And if you’ve hired a team of what we call “clickbait writers,” then you won’t earn the respect you need to attract your clientele, especially if you work in a research-based, association, or B2B-oriented organization.

Now, don’t get us wrong: search-engine optimization (SEO) matters, but it must be used in tandem with a quality content team who can give your increasingly scrutinizing users the high-level, nuanced, and personalized information they seek.

4. Remember usability testing.

Plenty of organizations think they can bypass usability testing when they use a vetted AI-based technology to personalize the marketing experience. The logic makes sense, given that the content is already personalized, and a key purpose of user research and usability testing is personalization. Yet this logic overlooks the fact that efforts to personalize can backfire, leading to embarrassing and disastrous consequences, especially when not tested.

In addition, as countless studies have shown, once users have a negative experience with a product or technology, they are unlikely to give it a second chance. So, it is in your best interest to get it right the first time by conducting thorough user research and usability testing.

As personalization efforts continue to rise alongside the many AI-based technologies entering the marketing arena, remember that your credibility as an organization resides on user trust and the overall user experience. Keep these four ethical considerations in mind as you develop a marketing plan and strategy, and for tips on integrating AI, read “Use Artificial Intelligence to Build Relationships and Strengthen the User Experience.”