Transcription
Good morning, everybody, and welcome to this episode of 30 Minutes to Presidents Club. My name is Arwan Farouk, and I'm here with my co-host, Nick Siegelski. And folks, we are live in studio with a multiple-time Hall of Fame guest, the one and only Jake Dunlap. He is the CEO over at Scaled. He is a former VP of sales multiple times, and he is giving you probably the most tactical episode you have ever heard on how to use AI in sales. Nick, why should people listen?
Correct. This was the most tactical episode on how to use AI in sales. Jake goes deep into how you should be prompting ChatGPT to inform your messaging. He talks about the triangle of things that you need to ask it about to inform your messaging. He gives us a number of different prompts that you can use to tweak what it gives you. He talks about using it to inform the questions that you ask. I was always afraid of how you would actually use ChatGPT to tell you what questions to ask, because that just seems weird, but he has a very, very smart strategy for how to do that. So this is one where if you're thinking about using AI in your sales process, I think you want to listen to Jake.
Welcome to the show. You might remember we start every single episode with your top three actionable takeaways. So, let's get to your three.
Uh, excited to be back, fellas. Alright, so I've got three generative AI top tips for you. Tip number one is around prospecting and my tip and advice is stop going to Google first. If you want to get access to like everything and summarize a company's link, you can go, "Hey, here's a link to the company, here's a product that I sell. Summarize for me the top three ways that this company makes money, top ways that they do business, their prospects, et cetera, and how we might be able to help." So if you're prospecting, you want to book a meeting with them, make sure you can go insert links, ask it to triangulate, and it's going to help you to book more meetings and help you to prospect more effectively.
Beautiful. What's number two?
All right, number two is closing more deals. And this is one that a lot of people don't think about, which is scenario planning. And so scenario planning is something that ChatGPT is very, very good at. So give it all the information, like how you would go to your boss and say, "Hey, I'm in a competitive deal. We're competing against Zendesk. Here's what my company does. I'm selling to the CFO. What are the top three things that I should make sure to include in the proposal? And then what are maybe the top three things I should not include in the proposal?" So if you want to use it to help you close more deals, get later stage, give it more context, do some of that scenario planning. It's a fun one.
Very nice. Round us out. What's number three?
Number three is skill development. Right? For a lot of you out there, you know, you're waiting around for the training or enablement to happen. Uh, but you should be taking your own skill development into your own hands. Right? And ChatGPT can be your own little wingman, your own little enablement person. So imagine this. Your boss keeps harping on you, "Jake, you need to be better at executive communication. Right. I need you to have more presence." So I can take that to ChatGPT. "Hey, I'm a rep at this company. I've got X years of experience. This is the type of deal size I sell. This is the dollar amount. It's a mid-market sale. I want to build a six-week training program with 30-minute detailed agendas that will help me to level up my executive presence." It gives you the agenda and then you can go back and say, "Hey, for this one, create a 10-question trivia game, asking me questions one by one that will help me to go through and actually learn." There's a whole new world opening up where you can take personal development from your sales career into your own hands, which is awesome.
So Jake, you talked about plugging in the ChatGPT when you're researching an account, website, what you do, "Show me the top three ways that they make money. Top three ways that we can support them as a product." I'm curious, why go that route versus what I've heard other people do, which is, "Hey, ChatGPT, I'm trying to research this account. Can you show me these top five research triggers that I know are proved things that lead someone to buy?" Why leave it more open-ended like that?
It's a really good question, and that's kind of the beginning of the prompt. What we try to do a lot of is, "Here's what the company does," and kind of the addition to that prompt would be, "I'm trying to book a meeting with this job title." So based on a person in this industry, or we call it like sub-industry, and this job title, and what we do, what are maybe the top one or two priorities? So I would take it a step further. What cuts through the noise still today is relevancy, and that's what most reps aren't doing right now. They're not saying, "I actually know the trends in not manufacturing, but not industrial manufacturing, and heavy equipment industrial manufacturing." If you send an email to a VP of Ops that says, "Hey, this is the trend or this is what I know is happening in your space and where I think we might be able to help." Dude, that's 0.1% of what's getting sent right now. So that's the way to do it around prospecting. You got to kind of try, we call it the triangle: industry, job title, what you do.
Well, if I just add to that a little bit, a lot of times what reps will do is they'll give you irrelevant research. For example, you might be reaching out to a company that does professional services in Milwaukee, and you're like, "Oh, they were growing in headcount, that meant that they had a problem that I could solve." Right. And then you try to take that same employee headcount growth trigger and apply the same messaging to what it means for a manufacturing company.
Yeah, exactly. And the problem is, what happens to a manufacturing company when their employees are growing might be radically different from what happens to a professional services company when they're growing, but you might not have a verticalized territory where you know everything about both of those industries.
Yeah, exactly. So what would be like an example of an output that you would get from ChatGPT? Maybe like one example of like how they would make money, and one example of how we could potentially help?
At least for me, what I'm looking for is again, when you say, "You know, let's say, how do they make money?" Right? What I'm looking for is the "how." How a company makes money. It used to be the first question that I asked every single rep. So whenever I was a frontline manager, reps would come to me and say, "Jake, you know, here's a company. We're trying to prospect him." I'd say, "How do they make money?" If you don't know how a company makes money, how can you help them? Because if you don't know, like, "Okay, they're in this industry, it's really competitive, da, da, da," like, to your point, these are the competitive pressures, I can't really appreciate the nuances of what somebody in that role is going through. I have to know how that company makes money to be able to triangulate what we can do and put it in their words. "Hey, hey, in your industry, these are the trends. These are things that are happening here. So I'm going to help help you to stay a step ahead." The other thing I'll call out is a lot of the messaging I'm talking about is for the leadership level. A lot of people's messaging is built for directors or managers about use cases and end-users. Why VPs or why CEOs take meetings is to get smarter. A VP of a big department, a CEO, they take meetings to get smarter. If I'm trying to set a meeting with executives, that's why I do this type of outreach. That's why I ask, "How do you make money?" Because I need to know how they make money to understand what they're probably focused on, industry pressures and trends, etc. So that's why I ask the "how do they make more money?"
Yeah. That's why it's so critical.
Yeah, I mean, it actually reminds me of, we had Kyle Coleman on the show recently, and he was talking about how you might consider using AI with prospecting, and one of the things that he said was, anytime you can find an initiative that the company has, that is close to impacting revenue, the closer something is to impacting how they make money, the more it is going to resonate when you reach out to an executive. My guess is, what ChatGPT is spitting out for you, it's pretty damn good. But I'm sure there is some last-mile refinement that you are doing. So what are the top two to three places that you see in that refinement to make that message ready to ship?
I want to zoom out just a little bit and talk about the concept of custom GPTs. Okay. The future of generative AI is not prompting. Okay. Prompting, you two are probably too young. You used to, it used to be called MS-DOS. It's your computer. Okay. You would be like, "forward slash run Oregon Trail," and it would like run the computer for you. That's where we're at with ChatGPT, right? That is where we're at with generative AI. Generative AI now relies on you prompting it. Okay. Custom GPTs are agents that prompt you. They have very specific use cases. So again, like, we've built a series of them. Like, "Discovery Call Prep." You click it, it says, "I'm ready to get started." It says, "Great, what's the name of the company? Give me the link," and it does this. Custom GPTs are the future, and I would encourage every single rep out there to go start experimenting with this. You can go. All you do is you, you go to, it says, "Explore Custom GPTs." You go to the top right, you hit "Create GPT." And to get the last-mile refinement, that's where you're adding in custom instructions. What we're doing with a lot of our clients, Dialpad, do some other ones, is basically building out for their team the last mile. So imagine on the backend, I'm not just programming custom instructions about what goes into good discovery call prep. I'm putting in your products. I'm putting in your buyer personas. You never need to tell it what company you work for. So imagine you're doing discovery call prep, you click the button, it says, "I'm ready to get started." Literally, you give it a link and the job title or the LinkedIn PDF, and it immediately does your prep for you, in real time, because it already knows all of those things. And again, I would really encourage all of you to go do this. The beautiful part about this new world, you don't need to be a tech nerd, you just need to be able to understand like a call scorecard. Like, I'll shout out my, you know, my good buddy KD, he had some of his leaders go through our AI-powered seller class. One of his sales managers' tactical application was like, "I think I could put my call scorecard in here." Yeah, you can. So he went, created, uploaded his call scorecard, gave it the instructions, "This is what I look for, dah, dah, dah, dah, dah, etc." His leader is doing two to three X more call, you know, refinement with his reps. The reps can now do it themselves. And it's scored the exact same way. So the thing that I would call out for a lot of people is this is where the ball is headed. So I would highly encourage all of you to take the initiative, go create one. You just type in instructions. "What do you think's in good prospecting?" or whatever. It doesn't really matter. If you are doing prospecting, I'll give you a couple of very tactical tips. A couple of things I like to do at the end of every prompt. "What are the two things I should have asked you that I didn't?" And then it will give me two more. Right? And I'll be like, "Oh, that's good." The other thing is I tell it to be more out of the box or whatever. I'll say like, "Okay, this is an interesting value prop. Give me a wildcard variation that's that uses industry." Oh, that's another one. "Use industry jargon." Telechat, DBT, use all their weird industry jargon. Give me as many acronyms as you can in those value prop messages, and it'll give you this like, "Itad world, blah, blah," like all this stuff. So, and then I'll say, "Great, give me a trend slash, you know, kind of like value propaganda." I don't have it, right? The email, right? We'll get into that in a second. Probably. I'll say, "Make it crazier." Like, "Make it more like if I was an industry peer, I was speaking at a conference that they wanted to attend the session. What would those three lines look like?" So the more specific you are, the more you tell it to get jargony, the ChatGPT responds to that. It's an LLM, right? If you tell it to work harder and give you a better answer, it does. Or it gives you like a different answer.
So how would I know that I got it right? What would be an example of a good answer of, "This is how they make money. Here's how you can support it," versus one where you look at it and you're like, "Oh God, I'm not going to send that."
Oh, that's a good. That's a really good question. I think you look for anecdotally, meaning like, for me at this point, how often do reps get it right now with Google? You know, how often do reps get it right with like going and doing their own research? So I trust it most of the time. Candidly, like if there's something intuitively that doesn't line up for me where I'm like, "Okay, it says that they do this. This seems really generic." I would say that that's probably the more common thing that I see is like, "The way that we can help is too generic." It's like, "We can help them with this." And then if that's the case, don't quit.
What would that sound like? It would be an example trigger of how they make money. Yeah, and uh, fluffy. How we, yeah. Let's say if like Scaled, let's say if we were doing it for a tech company, right? And it's like, "Scaled could probably help them increase sales." Mm-hmm. Right? Or "Scaled, you know, Scaled could help them with revenue operations." What we would probably do is then say, "Get more detailed and specific. I want you to really hone in on trends in their industry." Just tell it to get more specific. You just gotta keep pushing it. And I think that's why a lot of people give up on Gen AI is we're just not used to communicating with machines like this. If we give too much information to Google, it quits. It's like, "No results." Whereas this is the opposite. So we are reprogramming, you know, for me, 20 years of professional experience in terms of how to solve problems. You have to start to get used to going deeper. So you don't like the answer? No problem. Just keep hammering. Be more specific. "Hey, that's generic." Tell it that. "Specifically why this industry, why this job title, and why Scaled could help around like this problem point." So I'd say it's more of like a feel. I wish I could give you a more black and white answer because it gives different answers every time. You can type the same question in and we all get different answers. I think it's more of a feel of, does that feel generic? Is that something that anybody else could say? And then just tell it to be better.
This reminds me a lot of a prompt that my other co-host, Mark Cosiglo, was punching into ChatGPT, which is we were pretending that we were a marketing agency that was selling to Home Depot. And he was like, "What were the top three things that Home Depot believes that they can do, based on their 10K, to drive revenue growth up, down, etc.?" And it spat out three, like, pretty damn good things. It's like, "Oh, there's a softening in outdoor sports, so they're not focusing as much there. But they're really focused on some of these, like, more in-person type of activities centralized around some of the retail stores." And then what it said is like, "As a marketing team, what you can do is you can focus on helping them with their in-person events and activating those in-person events." And I was like, "That's pretty damn good. If Home Depot is trying to like get all these people playing around with outdoor sports stuff inside their stores and we're a marketing agency that specializes in event management, that's a pretty damn good way that we can support something that is literally written inside of their 10K." So my question for you, Jake, is you said you are not having it write the prospecting email for you. Two-part question: Why? Second part is, how do you turn that output into a good prospecting email, and what does that actually sound like?
Yeah, great question. Well, usually the problem is, first, ChatGPT starts every email off with "Dear." A lot still. Like, I literally tried to tell ChatGPT, I was like, "Get meaner." And it's like, "Or get more," and it's like, "I cannot do that." So that's, you know, problem one. Well, ChatGPT is not there yet. It's okay at making those correlations. It's not great, and that's why the human is still very much in the loop in sales for the foreseeable future. ChatGPT can make some inferences. It can say, "Oh, yeah, in industrial manufacturing, this is what a VP of Ops is doing. This is probably how you can help." But the human needs to say, "Oh, I know that's an issue. Oh, that's the one," or like, "This is what's going to be a hot button for them." We have to realize it's not there yet at being better than good. Good sellers are great sellers at doing the nuances like that. So it gets you 80 or 90 percent of the way there in two minutes, versus 20. And that's how you really have to view these tools. We've got this, this fallacy happening in sales, where something has to be automated to be scalable. That's insane. You know, your iPhone has Q&A, your Tesla has Q&A on their production lines. Not a lot of people, but some. We have to understand that AI's job is not to automate. AI's job is to help us be more scalable and to do higher quality activities in a higher volume, which is amazing. It's really the only technology that lets me do more and higher quality. So it's about getting you there and getting me those insights that might have taken me 45 minutes to research in 25 seconds.
Generally, when I'm prospecting to get a meeting with the CIO, I'm not just going to send one email. And if I don't book a meeting, call it a day. When I think about best practices for sequencing a contact, five years ago, my first email, that would be the one that I put all the muscle into it, right? A really, really highly personalized email. And then everything after that is like bumps, bubble-ups, or more like, more like, uh, broader stroke messaging specific to like, "Oh, their industry." It's more canned stuff. How are you applying AI and ChatGPT to the entire sequence or cadence when I'm reaching out to somebody?
There's two parts in there. One, I'll talk about non-ChatGPT and then I'll talk about ChatGPT. You know, the first is, so we have a sister company called RevOptics and RevOptics, we API into the largest sales engagement platform. So I can just tell you all very definitively, I have seen Touchpoint 1 be the number one converting part of a cadence in like one out of thousands of cadences. What does that tell you? You have to keep doing value consistently. Then I'll consistently see Touchpoint 6 having a higher conversion rate than touch points. Number seven. So number two is, there's a lot more noise today, so we have to have more omni-channel touch points. That's the data. This isn't my opinion. Go to RevOptics, you can see all the data like that. That would just be more of like, anyone out there prospecting, you have to keep adding value throughout and you have to be doing multiple channels. We're seeing a lot of our clients, 50-plus percent of the meetings are coming from the call. It's like, "Cold calls are dead," blah, blah, blah. We can get into that. It's about coming up with new variations of the value prop, which at GPT can help me to do is like, "I don't know anything about regional banks. Create a 15-question trivia game to teach me about the latest in cyber regulations for regional banks in the Midwest." And it can teach me that, and then I could go, "Oh, hey, did you know blah, blah, blah, blah, blah?" It can help you to actually continue to create variations of those value props. The first prompt set I get is gonna give you two or three trends. Just keep playing with those. Pull on the next thread. Pull on the next thread. That's the other thing. The reply to the reply, just so you guys know, data says like that's not as successful anymore either. So like the triple reply. So if I'm a rep and I have a choice, I can either have my 10 emails that I'm going to send over the next two months and I can personalize all 10 of those emails using ChatGPT, or I can say, "I'm going to do the traditional thing, which is personalize the first email and then go spend my time instead personalizing the first email on another account versus going and personalizing nine on one account." How should I think about that trade-off?
I just told you the data. You're just wasting time. Number one doesn't convert. I'm just telling you guys. The first touch point is the lowest converting. Not necessarily the lowest, lowest, but almost always lowest. Again, I think we've gotten away from, and again, I'll get tactical, but guys, the goal is to book a meeting with this person. You're giving a book of accounts that says, "My job, your job as an SDR, as a salesperson, is to book a meeting with ABC Co. That is your job. It doesn't matter how you get there. Get there." We've moved to this idea of account coverage as the goal, versus getting a meeting with somebody as the goal. We're seeing companies already move away from that. We've got a lot of our clients that are trying to unanchor from activity metrics, and we're really trying to push them to unanchor from like, "Just go customize another one, and then do that. Send all." It's not working. You see the data. You all miss outbound numbers. Like, why could we not see this? The data is right there.
What I'm hearing is touch points four through ten. It sounds oh, that has all day. My question for you though is, and I don't know if you know what the data actually shows, should touch points four through ten be personalized touches that I am manually putting effort into? Because I would assume I get more meetings if I do more personalized touchpoints four through ten. But I know I could send three X the number of messages if these are more broad-brush messages. So I'm wondering, does the increase in response rate, positive response rate, outweigh the volume? And I know it's a good question.
Yeah, like that's an opportunity cost question where it's going to be different for each business. So I can't tell you for your business. And I want to call out the difference and is there's a difference between personalization and relevancy. I'm just saying be relevant in every email. Add value in every email. I'm not saying every email has to be some weird new nuance of "Nick and I saw this thing." But I am saying, are you adding value in each touch point? And that doesn't require some type of level of crazy personalization. And like I said before, to do this type of additional relevancy thing, I'm literally talking about spending another one or two minutes in chat. "Give me another idea. Make it invincible. Make it to where this VP of Operations would have to meet with me, like his CEO was like, 'You got to go meet with this guy.'" And it'll come up with some crazy shit. It always does. So what we have in these tools is for the first time, I can do more activities, and it's higher quality. Yeah, everything else was, I do more activities and quality degrades. Correct. We don't now want to use that free time to just go do more shitty activity. Like, use the free time to do more relevancy. It doesn't have to be hyper-personalized. And then look, as you become more of an expert in your space, you kind of know what the trends are and you move faster and faster, just like any rep that's ramping.
This is all talking about how you research, find relevance, right? A message to one account. Are you doing anything with your entire patch where you say, "Here's a list of a hundred accounts. Help me prioritize which ones are the best?"
Yeah, that's a good call. So it's not there yet, right? So ChatGPT is not there yet. They do have a new model called the Strawberry model. It came out in September. That's better at that type of data, but you still would need to give it some input. So it's not quite there yet in terms of being able to be like, "This is it," like, you know, like SixSense on steroids or something like that. So no, you can't, you can't do that yet with it. We just built a GPT that does your account plan for you though. So it's about, it takes you about 20 minutes and it just walks through like, "Where's the spin today? Who are the stakeholders? What are the additional products you want to sell?" and just asks you questions and then prints your account plan. That's a pretty cool use case for it. It's not good at lots of data. "Find me the 55." If it's like, "Read an annual report, look for these trends," it's gotcha. But mashing up seven annual reports and tell me the trends, like it'll start to make stuff up.
So to recap before we go into ways to use this during your deal cycle, is there are almost like three levels to using ChatGPT when you're doing personalized outreach to an account. So the first level was one that really only I was familiar with before I even had this conversation with you, which is, "Hey, take this 10K, look for these three trends. I know these three trends matter." The second level is what you brought to the table, which is, "Independent of the 10K, show me how they make money in three ways. Show me how we can support them in three ways." And that's more open-ended. Also helps with above-the-line messaging. And then I can take all of that stuff and be like, "Hey, these are the three email templates that I usually use in emails one, four, and eight. Can you find a way to weave all of that stuff into these three email templates?" And it's sort of starting with a base of like, "This is how I write human emails," versus literally asking it to create emails from scratch. Do I have that?
Like, I love that. Yeah, I love that. Now, what I want you to do is I want you to think about all the things that go into what you think goes into a good email template, and I want you to create a custom GPT that says in the instructions, "This is what goes in." You upload 10 examples of what you think, and then maybe it can get closer to writing emails that are specific to your tone. And that's the next piece is we're seeing it. Reps are building their own stuff. Leaders. So then you're not having to do all that prompting every time. It'll prompt you. That's where this is going is even more exciting. I think we've written four or five, six different newsletters on how to write a good cold email, multiple email templates. And I literally just plugged in links to six 30MPC newsletters and I was like, "Pretend you are Arwan Farouk. You've written these six newsletters on how to write a good cold email. Now write me a cold email using this personalization trigger, selling this product to this company." And I was like, "Holy smokes, the difference between just using those six links versus just saying, 'Write me a cold email using any advice on the internet.'" It was a massive, massive difference, right? Again, the future is these productivity things where you just do that with every single one. You're like, "This is the topic and the flavor I want you to do for this one," and it already knows all that for every single where you have.
So let's assume now we are booking meetings with people and we're starting to get a pretty full calendar. One of the main use cases you shared that you're using ChatGPT for, for when you have a deal in cycle, is for scenario planning. So could you share the most common two to three scenarios that you will have it help you with, and what that actually looks like for a rep?
I'll talk about the two I think are the most common. One is a competitive deal. And in a competitive deal, you probably know you're up against somebody. I'm actually shocked by how little people know about the competition. You know that your boss told you, "We're better than them," but do you know why? Do you know why a CFO would care or not care? In a competitive deal, I can literally have it analyze. I can say, "Hey, here are the products that they, here's how they talk about themselves, here's their pricing page. I'm in this deal with this company," and you describe it. Like I said, give it the information. It's a CFO. They're in this industry. Here's my concerns. Here's my last call with them. Here's what we do. One, two, three. Give them the same links. Again, going back to that kind of quick example, "What are the top two things I should make sure to talk about? And if I'm worried the CFO is going to care about this, how could I better differentiate us in the proposal versus the competition?"
Yeah, I mean, that's the type of stuff I used to sit there and spend like, if I was putting together a big proposal, I'm spending like an hour, two hours just kind of thinking through it, moving things around. And now I'm like, "Oh, and then then you use the next thing I talked about. What are two more reasons? What did I miss?" And then I almost always, when you do that, it gives you some like little extra. You're like, "Oh, that's good. Like, I like that one." So competitive deal, I love because it could just help you to do that kind of scenario planning around that persona. What are the things you're going to care about against the competition? And then the other side of it, I like to use it in complex deals, which is you can upload multiple call transcripts. Let's say if you have multiple divisions, you know, CTOs love to greenlight deals. They love to move quickly. Same as CHROs. And I can say, "Hey, look, I've been working through my champion. Here's the deal. You can even say, look, we subscribe to Medpick. You can do all that. Hey, this is what we do. We subscribe to Medpick. Here's the call transcript. I'm worried that the CTO in this call might block me here, here, and here. Give me the three reasons why I'm right and the two or three reasons why I'm wrong. And then give me two possible scenarios that I can do to mitigate risk between now and our next meeting." I can actually have it gut-check. Did I hear right? Did I feel it right? And again, I want you to hear what I said. I didn't say, "Tell me if I'm right or wrong." Because again, ChatGPT is okay. But tell me the two or three reasons why I might be right. And then let me, as the human, two or three reasons. Now let me look at that and be like, "It's right. I messed that up." And so that's why again, a lot of these prompts, you'll hear me do that because I don't say, "Give me the answer." So give me two that are good, two that are not. And then that usually helps you to triangulate as the human. And then we're the top two solutions moving forward, how we could navigate this. Think of like deal inspection. Think of where I might lose this deal. Those to me are two really good ones, especially in those complex deals where you've got different personality types. And again, you can upload like just like you did with the newsletter, you can upload the call transcript and say, "What did you hear? Or what did you see, you know, as a part of it or listen to?" Those are the two for me around that, you know, again, to kind of getting in those later stages of deals. All those questions that you stay up night answering or thinking through in your head, you go to your boss, you guys sit down and you do your deal review. Imagine a world where you do that, you spend two or three minutes before you meet with your boss, and your boss is like, "Oh, this is a pretty damn good plan, Jake." I'm like, "Yeah, I know. I spent a lot of time on this." I think those are, those are to me, some of the best ways to use it around closing is just the things that you used to do manually. I used to do this in my head and I used to do this with other people. I now have kind of my permanent kind of thought partner buddy that can over time, again, with memory, it continues to update, continue to get more effective. I can say, "Hey, that's not a good answer." I think a lot of this stuff is really, really powerful in the middle of the cycle.
You mentioned something earlier that actually happens before and in between each of your calls, which is the prep doc process was literally the bane of my existence. Every single night, I would look at my six calls that I need to prepare for, and one by one, I would literally be populating all of this stuff, information manually, one by one. And then I would have to build a POV on these accounts. In other words, like, understand, like, "Okay, I'm meeting with a Chief People Officer in this one, a Director of People Ops in this one, like what are the types of problems that they're going to have based on their employee growth, their industry, and all that stuff?" So I know you have a custom GPT for building your own prep doc. If I had to build this on my own, what's like my 80/20 prep doc that I can use going into any discovery call?
I would think of it around the things that we talked about. One is industry trends. I need to know what's happening in their space. And a lot of the things that we talk about in particular are sub-industry trends. The thing where a lot of people, as they say, "It's in manufacturing," right? And it's like, "No, that's not going to resonate because there's 50 different flavors." It's in finance. No, there's 50. So one is sub-industry trends. What are two to three trends affecting this persona in there? Whatever it is. So that would be like one thing that I would do. That's next is then again, POV for those personas based on these trends and things that are happening. What's the POV for this persona? And then based on that, what are probably some talking points for each persona? And we've got a program on the backend where it's like, "How do they make money?" That's the very first thing it does. What are the sub-industry trends? You know, what's the job title and the trends for that job title in that sub-industry? Because that's what you really have to know. It's the persona within the ICP that is different from client to client to client. So those are the things.
I like all the things that you just said. Trends with the ICP, trends for the persona related to that sub-industry, and then what could our POV be around how we can solve for some of that. To what extent are you using it to help you prepare for the questions that you will be asking the prospect?
I'm not using it for that because I already know the right and wrong way to run a discovery. A job title is a job title. Yep. A director of operations at one company with a similar employee size has 15 direct reports, and in another company, everybody's a director. It's job is not to do your job for you. It is going to get you there, and then you have to turn on your brain. Yes. My friends, if it can do your job for you, like, you know, you talk about a sales AI going to replace people. Answers: Yes. If you are copying and pasting AI-generated emails, you are putting yourself out of a job. If you're copying and pasting that back and forth discovery call, and that's how you're running it, you are putting yourself out of a job. You are basically saying, "I cannot do better than this." That's the goal of the discovery call prep is to get you there, except my talking points. But I, no, you should know how to run a proper discovery. That is the human nature side of it, which is the techniques on what actually goes into the right and the wrong way to ask people questions in the order of operations, which is, you know, it would be a whole other episode on that.
So let's talk about this last piece of using the human brain side. I remember on your first episode ever of 30 Minutes to Presidents Club, you have a very, very, very specific way that you start all of your calls, where you first ask them about their top three business priorities, independent of the product you solve. And then if they're not specific enough, you will prime them where you're like, "Hey, based on this type of area, what are the top three things that you care about?" So I'm curious if you had a more specific ChatGPT prompt that was telling you the top three trends in the medical device manufacturing space, how would you take those things and present them as a POV?
Yeah, that's a great one. How I'd use some of these is I'd weave them into the quality of the questions. You know, the first set of questions usually around role and responsibility. "Hey, John, I work with lots of VPs of Operations in manufacturing. Sometimes they're a little bit more on the logistic side. Sometimes they're a little bit more internal ops. And I know a really cool thing as well to tell me about the top two things that you work on in your day-to-day." A lot of this is again, we're talking to people here. People want to talk to experts. Do your questions make you sound like an expert? Yes. So that is where it's where I could steal what you do. I'm gonna steal that, by the way, which is you using my advice, taking it to the next level. Now I'm gonna steal it back, which is like, I think you could definitely use it to ask better questions now. I retract what I said by pulling some of those snippets as credibility building.
Yeah, you have, it's like a credibility-embedded question because on its own to plop down with someone and be like, "What are your business priorities?" If they, you're like, "Alright, you're a 24-year-old AE." They're like, "Okay." But when you lead with showing, "Hey, I've got a deep amount of expertise in your sub-industry," all of a sudden they're like, "Great, I don't have to, like, lay the groundwork for this person." The quality of the response that you get, and frankly, the respect that they have for you, is far, far deeper because you showed how much you know their space.
That's right. And you don't have to have all the answers on the backend. Like, that's the key to good discovery. Good discovery is about, you know, think about a therapist. Yeah. A therapist doesn't have all the answers, but you pay them a lot of money to, you know, help you through work through problems. Yeah. You know, asking better questions, you know, doesn't, doesn't always mean, I think a lot of reps might get scared, right, to ask some of these questions. And it's like, you don't need to know more about the whole space than they do. But look, if you care about the space that you sell into and you want to, you need to be an expert to some extent. And I just feel like ChatGPT can help you to get there faster.
All right, Jake, the clock is ticking. We're running out of time. I got to move to the last question. We've talked about some good things salespeople should be doing. What is one bad habit that you see a lot of sellers exhibiting that you think they need to break because it hurts them more than it helps?
Over-reliance on generative AI. It's the copying and pasting and not realizing the key is to understand its job. It can get me to 80% awesome, if, you know, X percent depending on the project, faster. And now it's my job to click on my brain and be creative. Like, you finally have time to turn on your brain and like, I have extra time. I budgeted 30 minutes to build this thing or to do whatever for prospecting. Well, now I can spend 30 minutes making it awesome. That would be it. I want you guys to just think logically about the future of sales. If you are copying and pasting AI, how soon till it just sends or does the interaction or is a chatbot that a buyer wants to interact with? The answer is pretty black and white. So, don't copy and paste. Turn on your brains. That would be my last point.
Jake, thank you for 60-second recap. Coming up soon.
Alrighty folks, it is time for a 2x2 recap from this episode with Jake Dunlap. Nick, what do you have for yours?
So, so the first one is the triangle of what Jake asks ChatGPT before he reaches out to a company. He asks about their industry, not just industry, but the sub-industry. So not just manufacturing, but industrial manufacturing. So you need to understand that. You need to layer that with the job title of the person that you're reaching out to. A CIO will be very different than a CFO, than a CRO. And then you need to tell it, "Here's what we do." And when you tell it those things, it should be able to juxtapose trends against industry, against title, against how your product helps to help inform the messaging when you reach out to them. Number two, to make your messaging get through the typical sales rep noise, you can prompt ChatGPT to help tell you the industry insider jargon that you should include in your message to make it seem like, "Oh, this person talks to a lot of VPs of Operations in light industrial manufacturing." So ask it for industry jargon to make your messaging that much crisper. Number three, when you have the three ways they make money, the three ways you can help, the last piece is feed it three email templates and say, "This is usually what I send for emails one, four, and seven. Would you be able to adjust these and weave in those drivers for this company that's in biomedical manufacturing?" And then lastly, number four, you can use those same trend outputs to help add a point of view and help you ask smarter discovery questions. So for example, instead of asking, "Hey, how are you driving manufacturing efficiency?" You might say, "Hey, I know for operations people in this industry it means this. I know for operations people in biomedical it means this. I'm curious, how are you going about driving manufacturing efficiency as it pertains to this part of your business?" And you can ask a more smart question that shows that you're informed in their space.
Alrighty Nick, how can people help us out here?
So we put together a 30MPC newsletter recently that talks about step-by-step how you can use AI to inform your prospecting workflow. And so I recommend you take some of what Jake talked about, you combine it with that newsletter edition, and you are going to book so many more meetings than the folks who are doing things manually. So there's a link in the show notes to check it out. Go give it a look, and we'll see you next week on 3DMPC.