5 Guidelines for Using Generative AI in Customer-Facing Roles

ChatGPT is here… yay! All our problems are solved!

While this tech may not be inherently evil, there exists a considerable propensity for misuse and abuse by individuals and organizations. How likely is this tech to supplant your employees, rather than supporting them? How willing will they be to learn more about your product and educational materials when the robots can do it for them? Employees could become obsolete and out of a job, but perhaps you don’t care because you see the end-game here as a drastically reduced human workforce anyway.?

As I outline below, that way of thinking is mistaken both from an ethical perspective as well as a practical one for businesses. I include 5 guidelines for using Generative AI like ChatGPT, especially for customer-facing roles like sales, customer success, and support. Because it’s ok to “move fast and break things”, as long as that doesn’t include your customer’s hearts.?

First a quick story…?

Last night at bedtime, my 5 year old daughter was crying about how unfair it was that she doesn’t have a computer or phone like mommy and daddy. (Her screen time is limited to video chats with remote family members and the very occasional viewing of Mickey Mouse Funhouse… Miska, Muska, Mickeeeey Mouse!)?

I felt so deeply in that moment how I am misusing and abusing my devices, and failing to model a healthy relationship with them. And I told her that I have been irresponsible about how much screen time I allow for myself. I told her the truth, that while I use them for entertainment, information, and connection, they have become a prison. She wants what I have, but I wish I could go back to her experience of living life directly and not via the trappings of technology.?

So, given that generative AI comes with its own set of risks if we are not mindful of how we use the tech, I sat down this morning to write this post…?

TL;DR How many times have you become frustrated at a customer service bot or raised your voice and asked to speak with a human repeatedly? We are not merely looking for a resolution to a problem when we reach out to support, we are often looking to be understood, appreciated, and cared for.?

1. In Customer Support, ensure that ChatGPT is being used as a tool, not a replacement.?

What a benefit to provide immediate resolution for customer issues! But over time, as their confidence in the tech grows, reps will undoubtedly send generated responses to customers without deeply analyzing or even screening them first. I don’t know about you, but if I suspect that I was just forwarded an automatic response—no matter how helpful—I experience that as a dismissal and will likely terminate my relationship with that vendor.?

This behavior puts brand affinity in jeopardy. How you convey responses to customer issues is just as critical as the information shared. Brands are more than an amalgam of features and benefits, they convey a feeling (which is likely ultimately why the customer went with you in the first place). Don’t sacrifice affinity towards your human-centered brand for things like speed and efficiency.

2. Provide continued training to Service reps and CSMs on your educational materials and the product knowledge-base that ChatGPT is searching on their behalf.?

Nobody understands exactly how this tech works, which means that we are unsure how creative it can be at assessing and solving novel and unique problems for specific customers. Are they fed enough of the right information and do they have the proper parameters for doing this? How does a ChatGPT response compare with suggestions by skilled, experienced veterans who understand complex problems and can suggest workarounds that may seem counterintuitive? My guess is that for the time being at least, these people will still outperform the AI. But newer employees may never get there due to over-reliance on AI early in their tenure.?

3. Train people to know what each customer or prospect is really looking for.

Relationships, especially ones where you are being asked to advise, are based on deep listening. Not for information alone, but also for what is not being said. What does a customer or client really want? Why? What are the goals? Determining that takes intuition, which thankfully so far has not been programmed into a machine. I hope that it never will be, because if AI becomes effective at that, it will call into question our very humanity as nothing more than predictable behaviors based on data points. But I digress…

4. Continue to hire for soft skills and natural abilities, not just being a handler for an AI.?

Is the customer angry or frustrated? It takes natural gifts like empathy and patience to respond to the emotions that the customer is feeling. You may be able to quickly resolve a problem, but how you respond to the emotional experience is just as important. Over time, as frustrations with products and services compound, customers may just choose a competitor with similar benefits and features because of what the brand stands for and how every interaction makes them feel. There’s a huge risk here in terms of dehumanizing your brand.?

5. Give every employee time to get into nature and to take time away from their devices. (This was a great idea before AI emerged on the scene.)

Obviously, we can’t be proscriptive about this. Businesses are not cults. But leaders can model this and train managers, set aside time or subsidies for these experiences, or even create regularly scheduled team gatherings outside. You can provide resources about why tech downtime helps with personal benefits like improved health and happiness and professional ones like pausing to gain perspective on the businesses to be able to iterate and create what’s next.?

This is an example of where you can thoughtfully use generative ai and tech in general to create efficiencies. This can provide more downtime to get into nature, build human relationships, etc. But be vigilant. We tend to fill the time created by efficiencies with more tasks and staying productive.?

With all due respect and admiration for business owners and leaders who are trying to survive in a difficult economic climate and think they need to exploit every competitive advantage, we cannot look at this new technology through the myopic lens of economic efficiency. Especially when the future of our species is at risk.?

I’d love to hear your thoughts in the comments. Where do you agree? Where am I off base? Oh, and I wonder… has someone asked ChatGPT for ethical guidelines for its use in different contexts? I’m still not ready to engage with the tech, so if you feel like it, let me know what they say. Wait, is it “they”? Ask it for its pronouns while you’re at it.?

#generativeai? #chatgpt?