EP208: Balancing Relationships and Efficiency in AI Sales

In our brave new AI-augmented world, navigating tech integration while retaining that human trust factor remains a tricky balancing act.

So who better to provide expert guidance than our sage of sales, Chris Beall? Today our very own ChatGPT steps in as co-host for Corey to pepper Chris across the AI-sales trust landscape. Should we unleash these bots to comb leads? How do we mitigate client skepticism? Does transparency enhance trust? Chris distills hard-earned wisdom on these questions and more.

From specific use cases in training and process efficiency to ethical dilemmas around transparency, you’ll gain critical insights for smoothly integrating AI without severing those all-important human connections. Chris even gazes into the future, weighing engaging versus alienating roles for our robot friends.

Chris offers actionable advice so we can deploy AI judiciously while cementing bonds of respect and rapport. Time to bridge that tech-touch gap in this episode, Balancing Relationships and Efficiency in AI Sales

FULL EPISODE TRANSCRIPT Below:

Chris Beall  (00:14):

Hey everybody, this is Chris Beall, CEO of ConnectAndSell. And today I am not joined by Corey Frank, my co-host. Corey runs the awesome company, branch 49, in addition to many other things, finishing school for future CEOs by providing folks who want to be great professionals in the world of business with the opportunity to learn how to talk to strangers by cold calling them. So I asked chat GPT today to come up with 10 questions to play the role of Corey and kind of interview me Now, poor chat, GPT doesn’t get to listen to the answers. I could probably have it do that, but I’m not going to do that. But I asked it to give us an opportunity to talk about a topic that is near and dear to me, which is the relationship between trust and efficiency in B2B sales in the era of chat GBT.

Chris Beall  (01:12):

And there’s a lot of speculation out there as to where all this is going. I’ve been fairly immersed in AI for a long time, since 1992 in fact, and certainly am delighted that we have the ability to get to the next level with ChatGPT and all the GPT-4 stuff. And it does lots and lots of neat things. Many people speculate on what it might be able to do in sales, replace salespeople and so forth. I thought I’d explore just this one question, which is how about trust and efficiency? Are they actually at opposite ends of the spectrum? That is if we ramp up efficiency too far in B2B sales, do we ramp down trust or is there a case to be made for trust and efficiency actually working together in B2B sales. Now, just to remind my view of B2B sales, pretty simple. We need to establish ourselves with somebody else, another human being as an expert in a matter that’s important to them and as on their side.

Chris Beall  (02:18):

If we can do that, we become what’s called a trusted advisor. And a trusted advisor has a lot of freedom to get good information, quality information from another party so that they can bring their own expertise to bear and guide the process of coming up with a potential solution to a business problem that can result in a commercial transaction between the seller’s business they work for and the buyer business, the one they work for. So trust is super important, but efficiency is important too because there’s a question of, well, who are we going to sell to? There’s a lot of folks out there we could talk to. Most of them will not be ready to

Chris Beall (02:58):

Buy in the near term, even if they fundamentally need our offering, and most of them probably don’t need our offering if we’re at all specialized. I happen to work for a company called ConnectAndSell that provides something that lots of B2B companies need, but not all of them need it to the same degree. And so we have an efficiency problem just like everybody else. How do you efficiently go through the market and find the folks that are worth exploring with further, this is why the end of every sales interaction or the outcome of every sales interaction is the same. It answers the question, should we mutually decide to go down the road further together? So I asked chat GPT to play host Corey. It is standing in for you right now. And here’s its first question.

• Chris, could you start by giving us a brief overview of how chat GT and similar AI technologies are currently being integrated into B2B sales processes? 

Chris Beall (03:56):

Well, one place we know they’re being integrated into B2B sales processes is in conversational intelligence. Almost every conversational intelligence technology now, and in technologies like the one I’m using here Zoom video, I’ve got the ability to summarize conversations using a GPT model. What does that really give us in B2B sales? Well, one thing it gives us is time compression. If I’m a manager and I’m trying to understand how are all the deals going, I can talk to my reps and have one-on-ones, I can have one-on manys where we do some pipeline review or deal review, all of those things are available, but if I could take every 30-minute conversation and have it compressed down to me into 10 relevant points that I can rely on, I can actually learn what’s going well, and also where it might be worth bringing some other resources to bear, whether they’re technical resources, whether it’s a senior executive, whether it’s some brainstorming, other resources might be worth bringing in at a certain point.

Chris Beall (05:07):

And these technologies are awfully good at this. So they’re helping us be managers to participate more in the sales process without interfering in it. And it’s really an efficiency play. Now the efficiency comes from taking a 15, 20, 30 minute conversation, maybe longer compressing it down to 10 relevant points that can be read in about one to two minutes. So that’s call it 30 to one compression of time. And that means as a manager, if I had time for say 10% of all the conversations that are happening to come into my brain, so to speak, I now have time for a hundred percent and it only takes 30% of my day. So it’s quite an amazing breakthrough. And how does it speak to trust? Well, it’s kind of interesting. We need to trust that GPT compressor that took the information from a 30-minute conversation and the transcription engine that gave that information to the GPT engine for analysis.

Chris Beall (06:14):

We need to trust that it’s kind of getting the whole thing that is, we wouldn’t want to find out, oh, it gave me kind of a vanilla compression or a vanilla list and there’s a disaster going on in there and I didn’t notice it because once we rely on something like a piece of technology to help us out, it can often introduce blind spots. This happens quite literally when you’re driving old-fashioned car mirrors, you have a blind spot right next to you, but you’re relying on the mirror to tell you who’s right next to you and you can make certain kinds of errors more frequently. So that’s a place anyway where these technologies are being used in a very efficient way. They’re also being used to write, I’ll call it better emails. And I don’t mean these automated emails that are being sent out effectively as spam, let’s just face it too many, many folks.

Chris Beall (07:10):

But really just writing better than most reps. Let’s face it, most sales reps did not grow up as readers and writers. Some did, but most are better talkers and listeners. And so having an assistant, a co-pilot as they’re called to actually draft an email for you to do it in a professional way and to do it in what it thinks is your voice, it’s dynamite, saves a lot of time for the rep, but that’s not mostly the time. The rep can maybe get to the point of trusting that technology to frankly write better than the rep does. It’s certainly more efficient. So let’s go into the second question. So ChatGPT or Corey in stand-in mode here is asking me.

(2) In your experience, how has the incorporation of AI like ChatGPT affected the level of trust between sales professionals and their B2B clients? 

I would say not at all in the case where trust is actually being established, but I think it has kind of fooled a lot of sales professionals into thinking that they can generate trust in a mechanical way by having some AI write lots and lots of very clever emails, very wonderful emails and just send ’em out to folks.

Chris Beall (08:29):

And somehow, that will generate trust-level relationships. I don’t believe it. I still believe that there are those seven seconds that we have in an ambush call or a cold call to get trust, and that has to do with how people work on the inside. So if I read something that may have been generated by a bot, and I know that’s true of everything that comes my way at this point, it affects the level of trust immediately because I think, well, why would I trust a vendor’s bot? I know what its job is. Oddly enough, I will actually be more inclined to trust a salesperson that I know is a person, but how do I know that that’s a person? Well, I can’t know it through a bunch of texts that’s on the screen an email. I need to actually be speaking with that person and they need to be acting like themselves, so to speak.

Chris Beall (09:20):

So even if they were replaced with a bot, it’d have to be a really, really good bot. It’s not enough to fool me into think it’s a person. It has to enable me to comfortably say, yes, this is this person and by the way, get me to the point where I believe they’re an expert and they’re on my side.

So how about efficiency improvements?

This is question number three that chat GPT came up with playing the role of Corey Frank,

3) Can you provide specific examples of how AI tools made B2B sales processes more efficient coaching?

No difficulty coming up with an example or two. I’ve already given one of them. I think there are more to come. I believe we’ll actually get to the point where AI tools will help us look for how a whole team is doing or how a campaign is doing right now.

Chris Beall (10:10):

It seems to be very much one at a time. Here’s a conversation, analyze the conversation, take a look. What about analyzing all the conversations or a bunch of conversations that fits some criteria or finding the challenges that are thematic within a team’s efforts to reach out to the market? Do we really know who our message is resonating with in ways that one of these AI models can pay attention to? These things are tremendous at finding patterns within vast amounts of detail. So that would make it more efficient to find out who to talk to or more efficient to find out where our own team is challenged or those kinds of things. Bigger chunks I would call it. So those are some examples. By today. Most of it is conversation at a time or doing some writing and also doing some research for us. So if we are out trying to do some research before a conversation, a meeting that we’re having with somebody, certainly AI can help us a lot because it reads faster than we do.

Chris Beall (11:17):

It’s probably already read about this person. But one thing for sure, and this is about trust, remember GPT technologies hallucinate facts. So if you rely on facts that have come back from an interaction, a chat with a GPT technology, you probably are going to be brought up short someday. This is why lawyers have been caught recently trying to use ChatGPT and similar technologies to generate citations, legal citations, and they’ll use them confidently and it turns out well, it wasn’t a citation, it was a hallucination.

Alright, so next question from our virtual bot named Corey Frank,

(4 )How do you see companies balancing the use of AI with the need for human interaction in sales? Is there a sweet spot? 

Well, let’s work the other way around on this one. The need for human interaction in sales is absolute because of what I already said, which is we need to establish ourselves as the seller, as an expert who is on the other person’s side.

Chris Beall (12:24):

We can afford to be on their side. We’re the expert after all, they’re the buyer. They’re a generalist about what we are an expert on. So is there a sweet spot with regard to using AI? All of research is a sweet spot because the AI can lead us to understanding things about the world that would be too expensive for us to go find out, like who should we be talking to or what’s going on in this particular field? Or there’s some new developments that we should be paying attention to and so forth. So AI is great for research, great for drafting, but we have to be a little careful in final form. So proposals, sure, draft a proposal, you GPT model, but am I going to s end it off without taking a look at it? No. It actually implies that there’s going to be other human interactions kind of to check that what the AI is producing is accurate and compliant, which is also a big issue.

Chris Beall (13:22):

Well, how about in sales training? How can, this is question number five.

(5) How can Che GPT and similar technologies be used in training sales teams to enhance their efficiency while maintaining trust with clients? 

Well, one thing you can do with the GPT model is make a simulator that plays the role of a client. And certainly, if you’re working off an awful lot of client responses, which are all out there for almost every kind of product in the world, getting a simulator out of a GPT model is pretty straightforward and chat is a very natural way to do it. Now I’ll give you the caveat on this, which is when it comes to establishing trust with the human voice, the human voice is a human voice, not a human keyboard. So you can simulate by going back and forth using your fingers. You can also do it using speech-to-text, but it’s the sound of your voice that is crucial, especially in those first few seconds. It’s actually always crucial, and it may well be that, fairly soon, these AI models can listen to the sound of our voice and give us feedback as to how trustworthy we sound. That would be a wonderful thing. Right now, they’re not quite there. They’re fairly easy to kind of fool into thinking that we sound wonderful. It’s very subtle what human beings pick up on with regard to trust. I’ve often said people are to trust or

Chris Beall (14:51):

As wolves are to fear, they can smell it. And if you’re not trustworthy folks, they can so to speak, smell it in your voice. I know that sounds weird, I’m not talking about some synthesia either. So how about ethical considerations? What are this?

(6) Question number six, What are some ethical considerations the companies should keep in mind when deploying AI in their sales process? 

There is one, keep this in mind. If your bot is talking to a client, it needs to say it’s a bot. That’s pretty simple. Transparency is always the way to get to a high level of ethics, and I have a bad feeling there’s going to be a lot of folks saying, well, that wasn’t me. That was my bot after the fact. Instead of saying it upfront, you got to be upfront. So that’s a big, big ethical consideration. Another one is, if your bot or your AI is giving advice, expert advice, you better make sure it’s actually expert advice because you take on a little bit of what I would call kind of a Hippocratic stance.

Chris Beall (16:01):

That is, first you do no harm. When you advise somebody, you have to be conscious of the harm you might be potentially doing to them when they follow your advice. And so, just blindly saying, Hey, the advice from my GPT model is just go ahead and follow it without having it checked. It’s not just a problem that you might not be trusted. You might be committing a pretty serious ethical breach.

Alright, so how about question number seven.

(7) From your perspective, how do B2B clients generally perceive the use of AI and sales interactions? Does it enhance or diminish trust? 

Chris Beall (16:44):

It depends. If you give me something that you say was generated by your AI, by your chatbot or by your GPT model and you say, Hey, I had the GPT model whip this up for you and I took a look at it, it looks pretty good to me. Let’s discuss. No worries at all. In fact, it might enhance trust because you’re being open and transparent about using your technology, this technology to be more efficient and to bring in maybe other points of view or to digest more data that might be useful to the client. So in that case, I think you can enhance trust. I think destroying trust is really, really easy. I think I just hit on it in the previous answer, which is, hey, ask somebody to trust a piece of automation that you’re providing and don’t tell them that it’s automation. That one is going to cause not an erosion of trust, but a step function down from some number, probably not quite one if one is the highest number but to zero, and once trust is lost, it is not something you can go and get back.

Chris Beall (17:52):

So how about skepticism this question number eight,

(8) If there’s skepticism or resistance from sales teams or clients towards AI tools like ChatGPT, how do you recommend addressing these concerns?

Well, a couple of different ways. There will always be skepticism about new technology. There should be. New technology means new, it’s unproven. We’re not sure what it does. Really, really cool technology that makes you think something that’s not actually true. Like, hey, ChatGPT is a person who’s talking to me. That’s not actually true. Hey, ChatGPT is thinking about this. That’s not actually true. It’s a next engine. It knows a lot of stuff that it’s read so to speak, and it knows how to spit out the next token think token word, very similar concepts. And so when ChatGPT is talking to us so to speak, it’s really just going next, next, next. Now maybe our minds work like this too, and our voices work like this.

Chris Beall (18:52):

I have a feeling we do a lot of next, next, next, ourselves. It’s just the way the world is. We love to think that we’re really brilliant, having deep, deep thoughts and all that. Probably not. We’re probably just spitting out the next word that comes to mind. That’s why we call it comes to mind. And so it’s natural skepticism is natural. The way skepticism is overcome is through two things. One is transparency, it let’s be open about things. And the other is track record. So if the track record is good, and we note that over time some particular function, and I’ll go back to those conversations where you take a discovery meeting, and the AI compresses it down to 10 points. Well, the first time I read one I might think maybe it’s missing some things. By the time I’ve read a hundred of them, which only takes me a hundred minutes that I might’ve spread over a couple of weeks, I’m starting to go, Hey, I think this is pretty good.

Chris Beall (19:47):

I don’t have any big misses yet. But it just takes time. It takes time, and it takes experience. This is why the most skeptical people who are really, really smart plunge in and start experiencing what I call in anger or with an intent for a meaningful outcome, a new technology that is clearly as powerful as ChatGPT and all the things that are like it. So if you haven’t gotten in there and put your hands on it, so to speak, and made some mistakes, try writing some prompts and seeing what happens. Don’t just do the same thing over and over. Your skepticism only will go away appropriately and correctly with experience that leads to track record because now you kind of know where the landmines are.

How about question number nine, future trends.

(9) Looking ahead, how do you think the role of AI and B2B sales will evolve? What new developments should we expect?

I think the main developments we should expect are around efficiency. We will get much, much more efficient in B2B sales when two things happen. One is we have a better idea of who to talk to, who to go out and have conversations with that is very efficient. Secondly, we will be able to efficiently assess new people coming onto our team, assess performance of people on our team and help them get better. There’s a lot of detailed information and patterns that can be detected, and those can be detected in meetings so they can be detected in targeting or in messages and we will be able to bring that stuff together much, much more efficiently so that we can generate trust more frequently. And if you think about B2B from a sales perspective, it’s essentially the manufacturing and management of a portfolio of trust. So first, you have to be worthy of trust; okay, that might be the hard part.

Chris Beall (21:48):

But once you’re worthy of trust, are you being trusted by enough folks who might be able to take advantage in a good way of what it is that you provide? Probably not. I see AI, this GPT type, as being able to help us sift through, make sense out of detect patterns in larger amounts of data with different data sources so that we’ll just do a better job of having the right people talk to the right people often enough and generate more trust because that’s how it works over time and do so efficiently. So, allowing the generation of trust efficiently I think is going to be the main role of AI for quite a while in the world of B2B. And then finally, advice for sales leaders number 10. So Corey, here you are being played by ChatGPT who asked the final question.

(10) Finally, what advice would you give to sales leaders looking to integrate AI into their strategies without compromising on the trust and personal connection that’s vital in B2B sales? 

Chris Beall (22:53):

Well, my advice is kind of always the same, and I know people might get a little bored with it, but it is what it is; that is you’re trying to essentially pave a market with trust. A market is a list of people that if any one of them decides to do business with you, it reduces your cost and risk of doing business with every single other person. And therefore, their are companies in that market. Now that is not an oversimplification, but it hides some things like the flow of people through that list. Some come on, some go off, they retire, they go to new companies, they do all sorts of things. So there’s a lot that goes on. And usually your market is not defined very, I’ll say, coherently. That is markets normally are simply made up. Oh, we’re going to sell to all industrial companies who have more than a thousand employees who are located in these states or something like that.

Chris Beall (23:51):

That’s a kind of target market that you can think of, but it’s not a true target market because it doesn’t follow the rule that we were just talking about. So okay, so I got this market out there. Now what’s my advice? Well, you had to pave the market with trust, but you have to do that both in a targeted way and a little bit of a promiscuous way. That is you have to take the current opportunity and you have to identify it targeted, not just among those who might need what you offer, but among those who are more likely to need it. So we’re going to have more application of AI to determine different levels of intent. Now we tend to determine intent from something super simple, which is, oh, they visited a website or they went to G2Crowd or whatever it happens to be, right?

Chris Beall (24:41):

Action indicates intent. Well, there’s a lot of other ways that folks express intent, mostly through investment and mostly maybe about investment in people. So there’s going to be a lot to be done with regard to paying attention to how you can identify the folks that you need to pay attention to in the market while also helping the 11 twelfths of your market that’s not in market right now be more accessible and that you understand more about them and more favorably inclined to think that you are somebody they can trust to be an expert in an area they care about and that you’re on their side.

So just by sticking to those principles, then we are very free to use these technologies in creative ways. When we abandon those principles, we enter the danger zone and we’re probably not going to do too well. So I just want to thank my virtual Corey Frank here, ChatGPT for asking.

Chris Beall (25:42):

I answered these questions. I think they’re pretty sharp questions. Did ChatGPT think them up per se? No, it responded to a prompt.

Here’s the prompt,

please play the role of Corey Frank, co-host of the podcast Market Dominance Guys, you will be interviewing Chris Beall on the topic “the relationship between trust and efficiency and B2B sales and the era of ChatGPT”. Please generate 10 questions for Chris to answer. So Corey may well be a next engine himself and spit out questions like this one word at a time. I think he probably goes into a little bit deeper conceptual well than ChatGPT. But ChatGPT, you did a pretty darn good job there. I really appreciate it. And until next time, for Corey Frank, this is Chris Beall, Market Dominance Guys.