AI Sales Coaching Solutions: The Ethical Considerations

Good morning, everyone, and welcome to day 2 of the AI Sales Summit. We are going to be talking about a subject that has been covered yesterday. It's going to be covered today. It's going to be covered tomorrow. AI sales coaching that happens a lot of times before the call, during the call, as Gara just mentioned, and after the call. But we're going to be taking a different look at that, and we're going to be looking at it from the governance, ethical, and moral and legal implications of these extremely fast-changing technologies. There's, there's new AI coaching solutions come out, coming out, you know, daily. And I feel extremely fortunate to be joined by 2 fantastic minds in AI right now. Viera Orbanich is a human behavior and AI specialist, responsible AI advisor, founder of Helmet, and partner at Ethical Intelligence. And Dr. Andrea Rotulo is a senior global AI operations and governance executive delivering AI automation and advanced analytics into industrial, regulated, and energy environments. And thank you both for joining us today. Let's get started. So in the AI sales coaching realm, there's a lot of talk about upskill— upskilling the sales reps. And, and that's as advertised versus maybe as proven. So I want to get your take and maybe start with you. Andrea, kind of what's your take on upskilling versus downskilling, short-term versus long-term? How is it measured? So take it away.

Thank you, Jeff. And thank you so much for having me. I think this is a critical discussion that we need to have. So thank you for creating this space. Before I start on upskilling and deskilling, I would like to bring a bit of context of what I'm seeing today. Which is that most sales leaders buying these AI coaching tools are focusing mostly on the numbers. And I think that's not the right first question. I understand that, that, that's an important question. I think that the right first questions need to be focused on judgment, need to be focused on people, need to be focused on the law and bias. Because the tools that fail to answer those questions are the ones that will end up in the news, in court, or actually quietly deskilling your team while the dashboard is showing impressive numbers. So that's, that's the first thing. Gerhard has talked about mindset and Mike as well. Well, I think that's the right mindset we need to have when looking at these tools. And then regarding upskilling or deskilling, I think it depends entirely on whether the tool builds judgment or is substituting it. There are— I think these tools need to be focused primarily on simulation and role-play. So basically helping the sales representative to build the skill. Now, real-time prompting during a live call, that's the opposite because actually the rep executes without ever encoding their reasoning. So it's the opposite effect on capability. And let me just give you a bit of a study that was conducted by Carnegie Mellon University. And Microsoft Research, where they assessed more than 1,000 real examples of AI-assisted work. So what they found was that higher confidence in the AI was associated with less critical thinking, and higher confidence in your own ability was actually associated with more. And something quite interesting that was shared by those workers is that reliance on the tool actually discouraged them from critical reflection. So that's why, right, we need workers actually. So what's the process that I think you should worry or should worry as a sales leader is that workers stop checking the AI precise when they lack the skill to inspect and improve what it gives them. Then the less you know, the less you check, and the less you check, the less you learn. You see how this is a downward spiral?

Right.

So my suggestion to check what's happening with your sales team is run the unplugged test. Turn it off for a few weeks and see how your team is is responding. Has this tool actually improved them? Has it actually coached them in a way that has upskilled them, or what has happened? So my suggestion is that you run those tests. You once in a while unplug from the technology and see whether the tool has been helping your sales team or actually deskilling your sales team.

Right. Yeah, Vera, what would you— how about you weighing in?

Yeah, so I am aligned with what Andrea said. And I think what's important, first of all, to say as context is, and when we're looking at responsible AI, we're looking at legality, right? Being legal means you're compliant with the law and all of that. But if we go one step further and we look about the governance of AI within a business, within an organization, it's very much looking at what is good business practice. Right? So what is going to mitigate the risks and what is also going to gain us more long-term gains?

Yeah.

So looking at that, it's really important that we actually look at what would be good practice here as well, right? And if we look at AI sales coaching, we need to look at the fact that these solutions are coaching solutions. And that in itself is something that should develop the team, right? So we're actually giving an opportunity to the salespeople to develop themselves further. So looking at that, I, again, as I said, what I resonate with what Andrea said, because there is a practical, a practicing element to it that it's really great if it's unplugged, if people have the, you know, psychological safety to fail, because in order for you to develop, You're going to need to be able to also fail in order to grow and giving people the opportunity to do that. And then actually, after giving them that opportunity, evaluating that. If they're constantly being evaluated at the same time as they're being developed, that becomes a clash because then people are not going to feel like that they can fail openly and, you know, like they have the opportunity to actually grow. So, Good business practice here would be to allow them the opportunity to develop and then actually assess them. So to, to kind of also have that data be separated, right? The, the data of, of their rehearsals and the mocking that they do with the AI separated from actually how they are performing.

Right. So, I mean, And what I'm kind of hearing from both of you too is that it's— this is— these tools should be, whether they are or not, should be, and the ones that will be the best and the most successful long-term will be those that are elevating the team's skills, elevating the team's critical thinking, elevating the team's individual abilities. not those that are acting as a replacement for— in other words, turning them into kind of sort of human bots where they're echoing what the AI is telling them to say, um, versus the AI helping them to develop, you know, through practice, a higher level of skills.

Yeah.

And I think the other thing too I'd like you both to weigh in on is that I don't know how much, um, you know, we're introducing these tools, a lot of, and they do affect the mind, the mindset. And a lot of these, we're not going to know what the true outcomes are possibly for years, if not, you know, a decade or so of research. So I think the question is, should Where, and maybe both, should it be on the solution providers or the people that are getting the solutions to institute these kinds of what you were mentioning, the on-off kind of testing over time to see what it's actually doing? Because it may look good at first and then look bad in the long run. So do you have an idea, like Andrea, where that responsibility lies? Is it, is it both the solution provider and the— or is it buyer beware and they need to do it? Or where does that exist, do you think?

In an ideal world, it should be on both sides of the, of the table. But we also need to be very clear that there is a line between coaching and surveillance, and some of these tools, whether they may not be intending to do this, but when they are looking at continuous scoring of tone, of pace, and of presence, that sits quite closely to being a surveillance tool. So this is where it gets really complicated, and corporations may want to be surveilling their people. That might be part of their objectives. So hopefully they— we will have more data, more information. There's already quite a few studies showing that Actually, when you are focused more on surveillance versus than actually the coaching and upskilling of your people, you will be harmed in the long term because you are just having a parrot of a human rather than having a person that has more tools to better connect with your client. So, yeah, I think we need to assess why are these tools being integrated into an organization, what's the actual goal, and what would be really great is that you buy tools where you can provide feedback coming directly from your sales team. So one of the metrics that would tell me that you're actually looking to empower your sales team is that you are asking them for your feedback, their honest feedback on how this tool is working for them, and you're taking that extremely seriously, and you are ensuring that they can have a conversation with the tool developer to ensure that the tool developer adjusts the functionalities and uses of the tools according to the values of your organization. That would, for me, would be a signal that you're looking to help your people rather than do surveillance on your people.

Right. And Biaira, what do you—

I think it sits with both. I think that when it comes to the vendors, there is a responsibility there to actually build and deploy AI systems that are responsible and that take those things in consideration, the kind of impact that they're going to have on the users. And I think when it comes to the businesses, it's also their responsibility to ask the right questions before buying. And I think that there is also a really important point to make on AI literacy for the people who are using it. So in this case, the salespeople, To actually understand how the AI works in order to also be able to understand which feedback is relevant, which feedback is not relevant, to actually understand what the— how that model has been trained, what good looks like, and if there are specific things that that specific AI model is, you know, evaluating them against to tell them, to give them feedback. To actually know when that's relevant to them, because it might not be based on different use cases, based on who they are, the cultural, you know, context in which they're operating, based on their background and so on. So I think that the AI literacy part is also really, really crucial in order for people to be able to apply their own critical thinking and, you know, their own judgment on these things. But I think another just thing that I want to highlight is that there is a difference when we're leveraging or kind of using these tools as crutches. So they're making us more confident, but they might not give us more capability, right? So there is a distinction to be made between confidence and capability when it comes to upskilling our salespeople. And the goal here should be not to create salespeople who are dependable on these tools, but actually people who are independent and that they can at some point stop using the tool and actually apply what they learned. So using it really like a bit like as we're using a gym, right? You're going in, you're using, you know, your dumbbells and so on. But then you can see that results in your life when you're, I don't know, you're going hiking or you're doing other activities, right? And this is a bit of the same. We need to use them as, as developmental tools.

I'm hearing a couple of things. Number one, so Andrea, you had mentioned empowerment, which I think is something that stuck with me. And then you had mentioned feedback, and that works both ways because it's right now it's like, oh, the sales reps are getting feedback on their performance in a particular call setting or what have you. But the critical thing here and the solution providers that are going to do well and that you should look for are the ones that are actively making sure and working with the clients that the actual users, the salespeople, have full understandability of the AI that is being used, how it works, full disclosure, and that they are a critical part of the feedback loop that goes back to the not only management but also the solution provider so that it's a kind of a continuous, oh, this is kind of going the wrong way, we need to adjust here. You've got to keep adjusting tuning to make sure that you're not, you know, you're, you're never going to totally eliminate, but you need to minimize biases and that kind of thing. So Yes.

And Jeff, if you don't mind adding a little bit to that.

Yeah.

Is you hired your sales team, you hired them. So you see value in having these people in your organization. Right. And so now just trusting a tool because it looks shiny and it's presenting a lot of data and not trusting the feedback of the people you hired, I would find that to be strange. And what you don't want to do is to raise the stress of your people. What you don't want to do is to lower the job satisfaction. So yes, it's clear that a tool that's working well should raise the bar for everybody, but get the feedback from your team. The whole point of you getting this tool is that you're helping them. So that you can do better as an organization. So if they are saying, look, functionality A, B, and C is working great for me, but I want to disconnect D because it's stressing me, it's distracting me, listen to your people. At the end of the day, that was the whole point in the first place. You got this to help them.

No, and I agree with that. And it brings up whether maybe you start out with test environments first? Or like, one of the ethical questions that kind of puzzles me, or I think about, is that what happens, say you have a 100-person sales team, are any of them allowed to opt out if they don't want to use it? And then how are they judged if they opt out? Are they judged just by their performance financially? Are they meeting their quota, or are they judged because now negatively, because that may lead other people to want to opt out and therefore kind of undermine what the management is trying to do? So I think that's— those are questions that are not necessarily answerable, but it's interesting to ponder as far as what happens in those situations. And that also leads to You and both of you have already touched on it, but, but possibly how you may be in the long term. Are you helping or harming emotionally or intellectually your, your sales team with, with these tools? And a lot of times you're probably not going to see that right away if, if it is in fact Harming them. Viera, you want to take—

Yeah, I think so. I think human well-being, it's something that should be taken very seriously and should be top priority here. Also because sales in itself is such a relational, it's such a relational thing, right? And I think that the well-being is not only mental, but it's also relational, right? So it's actually looking at if you're constantly feeling like you're being surveilled and you're being evaluated by these systems, then what does that do to your mind, to your mindset? There is, you know, a risk that your sales team might start performing in a certain way to rather towards the AI system because they might know, okay, when I say this, it scores me higher or it gives me good feedback. So then instead of actually being with a person and leveraging their knowledge and their capabilities to connect with a human being on the other side. They might actually just perform for the AI system if there is a fear that that system is constantly evaluating them. So that's one thing, and that straight away erodes trust that needs to be there in when it comes to sales between a buyer and a rep. And I also think that on a relational side, we're We're buying these solutions in order to help the sales team to develop and obviously to sell more. So we're looking at this relationship between the salesperson and the buyer, but then indirectly, this also affects the relationship that a salesperson might have with their manager, right? When right now the manager might not be the one evaluating them, or it might just be somebody who is auditing. How they got evaluated by the system. So there is an erosion also within those relationships. And then furthermore, the relationship they might have to self, right, where they have the agency and they have the capability. And it's very much how then they also perceive themselves. So I think that that's really important also to take into account, not to say that these solutions cannot provide a lot of really great possibilities, but it's just for us and for people who work in sales to really think of, okay, how can we actually apply these tools within our specific organizations? Where do we have to have a human being in the loop? And if this actually harms our people, who is actually going to own that and what are we going to do about it, right? Those are things that needs to be thought of when we're actually using these tools. So this is not to say not to use them. It's just to actually look at how can we mitigate those risks and the kind of questions we need to ask ourselves.

Yes. Andrea, do you want to?

Yes, I agree with all your points, Vera. And going back to a point that I made before, there is a very line between coaching and surveillance.

Yeah.

And so let's just remember that, that being rated on your quota is a performance conversation. Now, being rated on your voice, on your facial movements, it's a completely different psychological contract. And they need, like, we need to have that honest, open conversation on what's the information that's being captured and how it's going to be used. There's evidence from 2 studies, one that studied the effects, the effect on electronic monitoring, and what they found is that It raises stress. As I mentioned before, I talked a little bit about lower job satisfaction. This lowers job satisfaction. And then this feedback, continuous feedback, actually makes these effects worse.

It's like a spiral.

Right. But having— now there's a different way of tackling this information, which is larger meta-analysis, which could be constructive one actually. So when you look at this information and you do an analysis rather than real-time feedback that you're giving the person, then the invasiveness has a moderate effect and workers actually feel significantly less stress in this case versus when they perceive that they are being monitored real-time just to, to surveil them, right?

Right.

In this case, when the larger meta-analysis is done and they are offered the feedback for their own growth and development, and that's clear, they take that information positively. So basically you have here the same technology, but it can have the opposite psychological outcome.

Yeah, it makes you wonder. Like if at some point, for example, in construction or different kinds of things, you have OSHA, which regulates, or at least in the United States, and I'm sure every country has their own regulations as far as worker safety. And you wonder if eventually they're gonna have the same thing for AI for emotional or mental safety. For workers because of how invasive AI is into the mindset. But so that, that, that's kind of interesting. So it's, it sounds like the, the AI coaching tools that are really going to be the best are the ones that treat the sales people as partners in a process that is an ongoing process and that puts them, kind of elevates them into a very critical level of being kind of, they should be driving the process really as far as because they're the ones that know what they need and what they don't need and what would make their job better and what not better. It's not a, oh, here's a solution we thought of to make money. So Here you all use it.

That's actually a really important point, Jeff, because there is a— there are a lot of AI solutions that might not actually solve the problem that your organization has. So the most important thing for each team and organization that is buying AI tools, it's being very clear on what is your specific need, and then actually looking and choosing tools that can solve for that. Right. So that's if there is one thing for people to get out of this session, like knowing your need would be a really, a really important one.

Right.

Yeah.

Yeah. And the way I would run this would be very simple. You bring different tools that are available in the market and you have these tools, the technology providers come and demo these tools. To your sales team, and then you have your sales team provide feedback on what they like, what they didn't like, and you as a leader clarify the whole point of this is to support you, to coach you. So with that mindset, which one do you think is going to serve that goal? Now, if I come as a leader and say, now you will have to use this tool, I don't ask for your feedback, Uh, most likely people are going to feel— and by the way, this is going to have an impact on your performance— well, that clearly doesn't sound like you're trying to empower your team, but you are just monitoring them. And there are many studies, again I mentioned a few of them, but that's going to have a detrimental effect over time, and probably you're going to lose your people.

Right. And I think, and that's one of the things too, I just wanted to, to quickly say that if, if anybody wants to connect with me, we had put together kind of a PDF of questions that you should be asking for solution providers based on some of the things that we talked about. We only, we were at the tip of the iceberg. There's a lot of legal and moral issues that we didn't even get to as far as consent and different things with different countries and, and who needs to be notified as far as whether you're using an AI tool, who doesn't. But yeah, just connect with me and we have one because the key is, and I think Andrea put it perfectly, it's like, what do the salespeople think can help them the most of these solutions? Not just, oh, this looks cool here, you have to use this. So we have a PDF, just, just hit me up and then, and then, or connect with me, um, and I'll send it to you. Hi, Gerhard. Do you have anything? Well, I wanna hear Michael, Michael's take because, um, uh, Michael, you are in, uh, uh, you know, that business where you help, uh, salespeople with simulations and you use professional actors versus AI. And, um, how did what Andrea and Vieira said resonate with you?

A lot of thoughts, actually. I do want to ask a question, and I'll lead into an explanation here. How do you create the feeling of a safe environment, a safe place to practice? What's the best approach to do that?

Yeah, so my, my 2 cents here is bring In your people from the beginning.

Yeah.

That's what— if you really care about them, you would care about what they think. So again, bring multiple tools to the table, have them demo to your people, and explain from the get-go what's your vision and your commitment as a leader. And if it's for coaching, say, take those lenses of I want to empower you I want to hear your honest opinion, and you're not going to be punished if you say, I don't want to use this tool. We will work through this together. That would be, I think, and based on my experience making these things happen, these transformations happen, that's the right way to approach it. Do not push a tool. Sit, get their feedback from the beginning. Do not make decisions without them.

Yeah.

Yes, I agree with what Andrea said. It's almost like team in the loop, right? Get your team to be there and actually be transparent with them. But also another thing I want to add is allow for failure, right? When you're developing people, that's really important. Again, this is not about evaluating their performance. It's about actually developing them. And, you know, imagine if I would give fins to a person who never went for a swim. And I said, now you're going to like, here you have this really cool tool, but you have to swim 5 kilometers. That person is most likely not going to be able to do that, right? So we're actually having to develop people in a way that they can also progress. And then also look at how these tools are going to develop with the progress of the salespeople, right? So it's really important that we allow people to fail and grow from From their failures.

Yeah. Yeah, I would call it professional play in terms of the conversation you're having with the practice platform or even with a real person and role player in a scene-related conversation. It's professional play. Have fun with it. It is a safe place to try new things, to be bold, to develop a new technique, and even do it in repetition and have fun with it. Do it in play. And when that conversation is light, the salesperson has an opportunity to allow themselves to feel. What are they feeling from the other person? And that's where you now become more emotionally acute to the conversation, where judgment is still there and available for the salesperson to use. Yeah, absolutely. Don't take the practice of judgment away from them. That's a skill they need to have and something they need to work on constantly because everyone feels differently at different stages of the conversation.

So, and I would also add that particularly when you're talking to a solution provider and they are AI forward, to me, a red flag would be those that say everything is perfect and we've got it all worked out. Those that I would tend to lean towards, those that say, you know, AI is new, we're working on this, we're going to be working with you. It is a process. We, you know, what they have, what do they have in place to monitor what the issues are and correct issues that may pop up? So I would look at those. It was refreshing. We were talking to somebody a week ago that had an AI solution tool and they said, well, you know, AI, it's, you know, something we have to watch out for because it does make mistakes. So it was refreshing to hear. It's like, oh, you know, sometimes you hear our AI has got it all worked out. Or we have no bias. And that's, those are red flags.

Good discussion.

Well, I think that the takeaway is that you wanna empower salespeople to step out of their comfort zone, and you don't wanna overpower people with AI by pushing them into uncomfortable situations. So it goes back to Using your emotional intelligence and that Andrea said, get them involved from the beginning.

Share a vision with them, include them, let them see it working for them, and then get them excited about it. And then when they go and experience growth, they feel the experience of growth from the session.

They can let them help and let them help develop the vision.

Yeah.

Yes. These are grown-up professionals. You need to trust your, trust your people, please. Right. They will give you good feedback.

Well, thanks everybody.

Thank you so much everyone for having us.

Thanks, Andrea. Thanks, Chiara. Thank you very much. Thank you.