If you work in high tech, or even if you don’t, you would probably nod your head if asked whether the following is on the top five list of most asked questions. Is artificial intelligence (AI) a good thing or a bad thing?
AI is a hot topic and it’s probably because, as with any subject where potential impact is mostly unknown but highly anticipated, a broad spectrum of scenarios has been imagined resulting in an equally broad set of predicted outcomes. Whether it will be seen as benevolent or malevolent will probably boil down to how each of us perceives its immediate impact on our careers. Speaker of the United States House of Representatives Tip O'Neill in the 1980s coined the phrase, “All politics is local,” which to him meant that no matter how high up a piece of legislation is crafted and then made into law, its impact is at a local and personal level. Maybe a good way then to think about society’s collective anxiety regarding artificial intelligence today is to say that “All AI is personal.”
So, if your management team announced that AI was being factored into the strategic plan for the coming year and that its focus would be how it might help increase the ability for the organization to advance the needs of end users, what would you think? Would you be immediately worried about your job? Would your answer depend on the field in which you work? Should that matter? Maybe not.
AI Can Be a Friend or a Foe
I had a conversation with a friend the other day that reminded me that it shouldn’t really matter. She updated me about her husband’s startup business and how it’s been growing rapidly through hard work and also because he’s incorporated AI into the product. He’s secured business from a couple of major hospitals and because his product is focused on allowing physicians to do more with less, it offers a compelling value proposition to cash-strapped funding agencies that struggle to keep up with public demand for high-quality care while operating in an environment of low to no tax increases. His product is successfully demonstrating it can make a difference one hospital at a time. As he scales his business, and as AI matures, it’s not hard to imagine that his business and the sector that it plays in will explode. And that will happen despite the concern from doctors that they might be replaced in the hospital setting or even that much of their own clinical work might be handled by AI.
Cost, technology, and public expectations have been inexorable forces through history, especially in combination, and we should expect their continued influence on this critical sector of our society and economy. We should be optimistic that physicians who find they have more time available because of AI-driven programs will pursue related research or clinical work during that freed up time.
When we look at why AI is being adopted in the healthcare industry it’s relatively easy to identify the main goals as being cost savings, increased efficiencies, and (especially) improved health outcomes for patients. Studies have identified many of the tasks physicians perform that either lead toward or away from those goals. The argument then becomes the same as the one posed in high tech: If a task is deemed critical but repetitive (say, correlating a patient’s multiple and variable symptoms and making a diagnostic determination), and yet the speed of executing the task is detracting from those goals and constraining the organization’s ability to reach and treat more patients, why not accelerate things by offloading execution of those tasks to AI? After all, AI performs tasks much better, faster, more adroitly than humans. There’s no argument there. Where it falls far short, compared to humans, is in the application of common sense and human empathy that can only be communicated during human interaction and that’s where physicians offer the most impact from a patient’s perspective.
Caution: You’re About to Enter the Other Half of a Metaphor
Companies should examine Customer Success through a similar lens. One of the things that Customer Success Managers do is use tools that query databases in order to connect the signals that customers emit when they are using the product or otherwise engaging with the company’s services. CSMs, for example, look for patterns within the activities of a wide array of customers who might be experiencing certain challenges when they try to utilize specific features of a specific product. Some CSMs are able to do this well but the process is very slow, the scope of query often too narrow (and therefore the results are questionable), and they aren’t able to accomplish the analysis in anything approaching the milliseconds achieved by a machine learning and AI-driven program. So, we’ll repeat the same question… why not accelerate things by offloading execution of those tasks to AI? Would that mean then that CSMs won’t be required as much when AI takes on a larger role in Customer Success? The answer is that they won’t be required in the same way they were required before. Let’s look at another example.
When it comes to figuring out whether a customer is going to renew the subscription, an organization devoted to managing renewals look at all kinds of variables to determine a risk factor. Then they direct energy and attention to the kinds of accounts that provide the best opportunities for renewal and growth. And why not? Well, I’ll tell you why that’s not the best approach.
In order to have a much broader impact on the propensity of individual customers within their territory to renew, and a much larger impact on the growth prospects of their own company, the work CSMs should be focused on, if AI is able to support them and drive the efficiencies many of us envision, would be value-add responsibilities such as gaining a better understanding of their client’s business and the forces that the client is facing that might mean changes are required in the way they better leverage the products and solutions that the CSM supports. It would mean that the CSM could better examine the experience the client is having with the solution beyond just the implications of, say, the KPIs that are listed as means for measuring progress toward the goals detailed in a success plan.
Are the client’s teams properly enabled for success? What are the product adoption rates of the various teams? Do they differ from each other? If so, why? Would it improve the client’s business if all the teams adopted the product to the same degree? What are the observations the CSM has of the client’s use of the features of the product? What kind of direct feedback has the client provided that might benefit the product management team? How can the CSM encourage the client to join a community to share and to learn from industry peers? Should the CSM take a more active role in those communities? Knowing how the client has adopted one product, what can the CSM proactively do to educate the client about the advantages of an adjacent solution? If CSMs were to be able to operate in that manner, what would happen to the renewal rate? Through the moon, that’s what. And AI would enable that model to be executed at scale.
AI cannot perform in the same manner as the CSM described above, at least not yet. Much as a physician possesses the unique human quality of empathy and the ability to do a much better job of factoring into diagnoses and treatment plans the intangible quality of trust, so too do CSMs have the edge in those regards over AI. For the foreseeable future, only a human will be able to ponder how to deal effectively at a scale that humans relate to but they will need the assistance of AI if they want to make a broader impact across many humans, many territories, and many societies.
Listen to this Bloomberg Daybreak Asia podcast, AI Can Be Harnessed to Change World of Marketing, featuring Shashi Seth, our senior vice president of products.