Sales call teardown (#3): Using "leading negatives" in discovery

Okay, so Harold, after I tore down two of his calls, changed some things, and then reached out with a sales call that he had like a day later or right after the teardown and said, hey, take a look at this. I think I did much better. Let's see how we did, why don't we? So this is the context here. This is for a company that reached out to use Slick for their recruiters. So Slick, as you know, is an outbound tool. It's used by kind of like salespeople and founders, not necessarily by recruiters. And so, right, like flag this in brain for future product market fit question. Let's just focus on the sales call and how Harold does and see what we learn from it.

Take your time with us here. Yeah, Michael and Summer had sent over, you know, just the, the Slick tool in general. And mentioned that it's something that the sales team utilizes. Was curious, you know, Summer had the idea to see if we could try to use it as a sourcing tool, recruiting tool, because that's my responsibility here at CEC is overseeing the hiring. So just seeing what other ways we could potentially use it.

Sure. Yeah. Anything on your end to add, Summer?

No, I think really we're just trying to get a feel for what this could look like, what the tool does, and how we could gain access and start utilizing it.

Okay, cool. So here we have a kind of, why'd you take the call? Classic. Could we use this as a sourcing tool generally? Want to see more about the tool? Okay. Hopefully Harold has learned the lesson that we don't just jump into a demo from here and we ask a specific question about like, what's— what are they actually trying to do? What— in what way are they stuck? Okay.

What I'd love to do is just understand exactly what you're trying to get done. I'll be completely transparent with you. We don't have any recruiters on the platform. It's primarily a sales tool right now. So I just love to understand What are you trying to get done? How are you doing that today? What tools are you using? What's breaking down with those tools? And then if it seems like Slick can solve those problems, then I'm happy to show you how. Uh, but yeah, I'm not sure if we can. So I'd love to understand all of that, uh, context first.

Yeah. So this, you might look at it and be like, hold on. Like, don't give away the farm. We don't work with any recruiters. I don't know if this is a fit for you. You tell me. I think that, and at this stage, what that does is that opens them up to tell you the truth. I really like the, what are you trying to get done? What are you using and what's breaking down? Right? That is basically asking for the POL framework as an example, right? Like that's, that's essentially asking what's the project on your to-do list that you're trying to do right now? And what are the list of options you've tried or used and why are those options limited or why, why are they lacking?

Yeah, so I, I'll start Josh and then you jump in just because I've seen this like tool a little bit. Um, so really what we're looking at is trying to build out and find names of people for us to target. So we know account managers in low voltage integrators within these 4 states that we're interested in finding anybody with that title so that we can try and find them on LinkedIn and then connect with them. So really it's the generation of names that we're looking to do or looking for because LinkedIn has a tool. I will tell you, I think it's fine. I don't think it's great. And so wondering, you know, if this can maybe do that better to give us a better list to contact.

Okay, got it. And by that tool, are you talking about the LinkedIn Recruiter tool? Yeah.

Okay.

And then what exactly is like, why is that only fine for you? Like what's missing in that tool? What's breaking down in that tool?

So this is a really good thing too. Harold gets some kind of, uh, Bob, uh, Musta calls this unpacking. You get some vague language here. So she says trying to build or find names for us to target. So here's a specific example. Um, so this is the project I'm trying to do. Um, LinkedIn has a tool for this and it's fine, not great. So that's too abstract. We want to double-click into this and like, there's two ways we want to double-click. We want to double-click to like understand what exactly does she mean? And like, by the tool and being fine or not great. And then secondly, like, okay, is this actually something that needs to change? Because there's so many things in life that are fine, and not great, that like, I'm not actually going to change. So let's see.

I can speak on that. It's just hit or miss, you know, depending on how you search, when you know where you search, you may not be getting a full set of results. You know, not everybody's on LinkedIn visible on LinkedIn, you know, just wanting to kind of expand the pool that we're, that we're fishing from.

And I'll make it even bigger than that. Like, there's times that we know somebody is on LinkedIn and we search the criteria and they don't show up in those results.

That's true too. That's, that's true too, for sure.

For some reason, the dataset isn't working with the query we're putting in.

Okay. And what's an example query? So you already gave the example, people who work as account managers at Lowell Voltage.

So, okay. So we get something here. Why not good enough? Not a full set of results or just like can't trust the results, it seems like. Which is great. The question we're not asking here is like, okay, cool. Like, but is it okay? Like, it seems like you're doing fine. Does this actually need to change? Harold jumps into an example here. So I'm curious where he goes.

Each integrator is in 4 states. Is that a type of query that you can just type in to LinkedIn?

Yeah, it's a little more, you know, it'll be by zip or specific city. You can do statewide, but You know, if you want a specific example, let's say I was looking for a fire alarm engineer in southern Wisconsin, or if I was looking for a low voltage sales executive in western Iowa, or, you know, Iowa as a whole, you know, those are— we, we kind of do a broader search and then narrow it down. Or I mean, I guess we do it inverse as well. You know, we'll do a very specific search and then broaden, broaden things, you know, if we're not getting a ton of results.

Okay. And sometimes like if we're struggling, we'll say like, okay, we'll go to the hiring manager and say, we need 5 competitors in the area, give us those names so that we can target. So then we'll try to find those companies on LinkedIn, but it's hit or miss if, if they're registered with that company or not. So we just struggle sometimes to actually connect.

Okay.

And you mentioned also some reaching out to these people. So how do you do that today?

LinkedIn and mail. Email.

Okay. So you send them a LinkedIn email.

Okay. Yeah.

Okay. So just to recap my understanding, you're looking for something that can help you build out specific lists of people to target based off of just a description that you give the tool. And LinkedIn Recruiter doesn't exactly do that. The lists that it provides aren't as high quality as you would like. Is that the right understanding? Correct.

Okay. Okay. So here, I think, I think Harold's done a, Harold's done a lot better. Good job, Harold. I'd say he's up to like an 8 or a 9, uh, on, on like looking for poll. What I don't know here is like, okay, yeah, is, but is this actually good enough? Is your current setup good enough? Like, do you actually need to change? What's, what, what bad thing happens if you don't change? Like, does it like, tell me, tell me more there. And, and the way that he does a recap. So I really recommend doing what I call a pull recap here. He does a good job of it, but it's almost a, it feels almost leading, um, here, which is okay. You're looking for something that can help you build out specific lists of people to target based off the, just the description. LinkedIn Recruiter doesn't exactly do that. The list provides aren't as high enough quality as we would like. Is that roughly the right understanding? So that's kind of a leading question. If you're going to do leading question, lead to the negative, which is basically, okay, cool. So I'm hearing that you're building out lists today. You're using LinkedIn for that. You, it's like not ideal, but I guess I'm not hearing you need to actually change. Maybe, maybe I'm missing something here. Right. And, and that's, that's actually going in the opposite direction and just making it be totally okay for them to say, yeah, actually it's, It's mostly fine. We're just kind of curious to see what you have. Okay, great. I'll show you something. Maybe you'll buy it. Maybe you won't, but I'm not feeling that you really have pull there. So, all right. But let's see what Harold does from here. So at this point, we've done a pull recap. We should describe the product conceptually.

LinkedIn and mail.

And mail. Okay. So you send them LinkedIn and mail.

Okay. Yeah.

Correct.

Okay.

What we have is an AI agent. That basically finds people for you based off of very specific prompts that you give it. So you'll just say, hey, find fire alarm engineers in southern Wisconsin. It'll go out, search the web, and find that, return that to you. Is that kind of what you're looking for, or is that, is there something breaking down, like missing from that description?

No, I think that's exactly what we want to try and see, like, is that giving us a better pool than what we're doing today?

Okay. I mean, are we able to, you know, fit, probably not, but, you know, say, oh, if we get lucky and it brings up 500 people, can we then add, okay, add those with, you know, nice attitude, or, you know, with nice attitude, whatever it may be. Can you get very cranial with it if needed.

Okay, cool. So see what happens here. Harold describes it at a conceptual level that's like not super, super, like there's not a ton of description here. He's not doing education. It's the smallest possible description. AI agent that finds people for you based off very specific prompts that you give it. So for an example, you'll say, this, it'll go out and search the web and find that for you. Is that kind of what you're looking for? Or is there something missing there? That's kind of asking the negative. That's good. That's leading to the negative side.

And, and the response here is, no, I think that's exactly what we want to try and see, like, is that giving us a better pool than what we're doing today?

Amazing. Cool. Great. It's not that like anything, if he had gone on about his differentiation, if he had gone on about his theory of the case, They would have zoned out. He gave a very good explanation there. Thank you, Harold. All right. And then we get a question here.

Okay. Are we able to, you know, can you get very granular with it if needed?

Yeah, you can get very granular. So for example, some of the queries that I run are find account executives in the United States who work in B2B software companies where the company does not have an SDR or RevOps team. So like super, super, super specific. Obviously the more specific you're going to get, the less results it will return.

But more often than not, we won't get super, super granular or super specific just because it is such a small group of people that fit that criteria.

So. But at the same time, Josh, that's a really powerful tool because sometimes we have a hiring manager that says they want this like unicorn of a person and we'll search it in LinkedIn and it's like one result in the nation.

Yeah. Yeah.

Like, is that real? Because if it is, then like we need to change our expectations of this role.

Yeah. What's happened here? Summer's all in. Josh seems like, I think that's his name. Josh is like, oh yeah, I don't like, he's not as all in as Summer, but they both seem like they are in. They're kind of just based on the description are like, yeah, that sounds exactly like what we need. And so at this point, Harold should ask a very specific question, which is, okay, what do you need to understand or see in order to kind of like make a decision on this. So let's see what he does here. Okay, great.

What would be most helpful for you to see? For you to see kind of what I described, just like searching stuff.

Okay. Okay. That's not the right question to ask. Um, what do you need to see? Ask it like, Okay, it's not a terrible question, but asking what do you need to see? They're like, okay, I'm thinking about the tool. What do I want? What do I want to see in the tool? And then Harold kind of like leads them in the questionnaire. Again, not the worst thing on earth. I'll also say on the supply description earlier where he did a very specific description here, not unhelpful to like actually show them something. I know it was just 2 sentences, but still, if you can just show them something, that can also help them get it too. But anyway, here, this is like, I'd say this is B class question.

Awesome.

That'd be awesome.

Okay.

And now that he's offered to show this, I don't know if you recall from the last teardown, but what he did is he showed the setup flow and the potential customer said, I don't like that at all. I don't want, like, I didn't enjoy you grilling me on how exactly to, like, I don't know, basically doing setup, right? What's the exact search I am looking for? I just want to I want to see what it looks like when it's done so I can understand how I'd interact with it. I don't want to see how I have to set it up.

Can you see my screen?

Yeah.

There you go.

Yep. Okay. So just to orient you to what you're looking at, I'm just talking with an AI agent on the left. I'm asking it to do things for me and it's showing its work to me on the right. So when Michael and I were on the phone, he was describing this case. I think this is almost like verbatim what he said. He said, I'm looking for salespeople in the Midwest that are systems integrators for commercial buildings, schools, hospitals. They must be in their, at their current company for at least 5 years. Uh, so this is what I asked the agent to find for me. It did this and now showing its work here on the right. If I click in here, I can see exactly why the agent is saying this person meets these criteria. So this has 4 criteria. Employees and account executive, their technology systems integrator, they sell, install AV and has been employed at their current role for at least 5 years. So as I click on these, I can see why it meets these criteria that I laid out. And then the other thing is like. I can just tell it to find more people. So I can say, uh, what was it? Find fire alarm engineers in Wisconsin, I think is what.

Yep.

So it'll go out and it'll do it and find these people that meet these criteria and then share its output here on the right. From there, uh, so this will kick off the search. From there, what?

Okay. So this is good. What you'll notice is that Harold is not doing much. He's just showing something that is already filled out. It's just, it's there. He's not clicking a lot of things. They don't have to understand a lot in order to do this. At this point, Harold has explained this in its entirety. He should say, does that fit what you are looking for? Instead, I have a hunch he's going to keep talking.

Some salespeople are using this to do as well is just automate outreach to these people. So I would say something like, You know, uh, John, Ed, Mike, Mike look like great candidates. I want to reach out to them and here's the automated thing I want to do. So what I have here is, let's see. So I found salespeople and I have them going through a sequence for me on LinkedIn. So it's sending out LinkedIn messages on my behalf.

Does this tie in then to your, obviously, like LinkedIn recruiter's profile?

Yeah.

So this is not LinkedIn recruiter. I don't have LinkedIn recruiter. This is just with my personal LinkedIn account. Yeah. So I connected my LinkedIn and this, this person has accepted my connection request. The AI agent has drafted this message for me. I've given it obviously a template. I come here, I click approve, and now it's sending it for me. Uh, it's also, let's see if there's, yeah, this person said that they'd be willing to chat, so I can click approve. I can view the message history and, uh, you know, approve these messages. So this is how I use the product to not only find people, but also reach out to them on LinkedIn in some sort of automated way.

Okay, so I'm gonna pause it here. And we'll just talk about this with this up on the screen, because I think it's better than me typing. This is okay. I would not do this. What Harold is doing is he's increasing the surface area of this product. Now, they seem to like it. But they also very well could have seen this because it's not something they asked for and said, oh, yeah, that part doesn't seem like it fits. I don't know if the whole thing fits. Maybe this does more than we need. That's totally plausible. So does— what I would ask if I was Harold is before jumping to the outreach piece, saying, does that seem like it fits what you're looking for? Yeah, that seems exactly like what we're looking for.

Cool.

Okay. You also— you didn't mention it. Were you also trying to have some way to kind of like do more automated outreach to these people. I don't know if it takes a lot of effort for you. And if they say, no, our kind of outreach approach is like, we've got it nailed down. Okay, great. Yeah, no worries. Don't, don't even bother showing it necessarily. You might mention, okay, cool. Well, if that ever changes, there's a way to do outreach in here as well. No need to focus on that though. Doesn't seem like that's relevant. Now, this is a case where it worked. There are going to be many cases where this doesn't work. The principle here is, if it's not something related to their poll, Showing it is almost always mostly downside. Show the least number of things that fit their pull, that unblock them so they can do the thing they're trying to do, and then stop talking.

That's incredible.

Yeah, that's really, really cool.

So how can Josh and I get access to this?

Boom. Okay. How do we get access to this? That's what pull feels like. It's like, cool, Harold, stop talking. How do we get started? How do we move forward? Um, what Harold should do is he should stop screen share right now and focus on next steps and not make them super complicated. Let's see how he does.

Yeah. So you just go to, uh, getslicked.com. You can sign up for free. We give you 200 credits and those credits are probably enough for you to play around with it. See if it can meet your recruiting use cases. And if it does, let me know and then we can get you on a paid plan.

Okay. Okay. We've just introduced a new concept of credits. Why? What's that going to do for us? Is that going to help us? No, now they have to understand something new. Harold, Harold, please. Yeah, so you can get started, you can actually get started for free, try it out for free right now. You can also upgrade within there. Or I'm happy to get you on a team plan as well. But yeah, you can just get started right away. Or, yeah, like we can put you on a month-to-month plan. It's $50 a month. And we can get started with that if you'd like that, right? Like you could go either of these directions, but introducing credits Not, not ideal.

Oh, oh, is one credit like a search?

Is it? Yep. Oh gosh. Okay. So what's happened is because we've introduced credits, they're asking questions about credits. What the heck is a credit?

Good question. So credits are used for searches and for AI, like AI usage. So if, so this, uh, this search here, let's go back to the search that I was looking at. It was, I think, CC, Korean use case. Oh, and by the way, so it's already returned. 13 Wisconsin Fire Alarm Engineers, uh, in that search we just ran.

But, Chris, again, we are just convincing them, you know, like, persuading them not to buy.

They're consuming the searches. So for each person that it finds, that costs 2 credits. So this search costs 26 credits. And then there's also, yeah, uh, like AI usage as well. If you talk with the AI, that consumes credits as well.

Okay. So, at least to start, we're better off being pretty specific so it doesn't return 100 individuals and there goes all of our credits.

Yeah, yeah, exactly. Yeah, yeah, I would try it out. I think where we kind of are a little better than like a database of people is you can be very, very specific in who you're trying to reach. So I would try that.

Okay. Awesome.

Okay, well, let us play around with it and then give us, you know, a couple of weeks and then we can, or I don't know how fast we'll get to this, get through the 200 credits as we test it out, but then we'll get back with you and discuss, um, moving to a paid plan.

I think Harold froze.

Okay. It looks like Harold's frozen. Okay. So from here, I think there could be a longer discussion to be had about how Harold should think about the stuff that happens after this. Like, what are the next steps? What's the structure of that? Maybe that's for a future teardown. But, but yeah, I think, I think, oh no, oh no. Harold asks, okay, great.

Last question I have for you is, do you ever send these people you're trying to recruit emails as well? Or is it primarily LinkedIn?

No, I prefer once we connect, I get their contact information so that we're connecting directly rather than through LinkedIn.

Okay, got it.

Sorry, what I meant by that question, I didn't phrase it right, was like, let's say they don't respond to your LinkedIn email, do you then try to find their email address and then send them, you know, what contact info is there.

Yeah, if there's additional info with their profile, we'll definitely try it. You know, if there's other things listed out there.

Okay.

But if this has a way of capturing those people's email addresses, that would be fantastic. I think we probably would message people on both.

Okay, this is called selling past the close. It's fine, but I don't, I don't love it. So anyway, Harold, much improved here. I think, I think you can see the things that he really improved here. Discovery, way better, actually looking for the thing, poking it when he, when he doesn't quite know what it is, um, double-clicking on certain things, uh, you know, acting, uh, what he could have done a little bit better on Discovery, I think, was, um, like, do they actually need to change, kind of actually testing it versus there's, you know, a little bit of, uh will say confirmation in there. Like, I don't think they would say— I don't think that they would have said, yeah, no, not really. You know, he was kind of a leading, leading recap. The supply description was great. Two sentences said, yeah, sounds like exactly what I need. Cool. Didn't ask the perfect question on what do you need to see or understand? In order to make a decision on this. It was just kind of like, what do you want to see in the product? I can click around and show you things. Not ideal. Demo was way better than before in that it just started from a completed search and you could kind of talk them around that. That's way better, but demoed a bit more than need be. In this case, it worked okay. In other cases, that won't work okay. And then there's the selling past the close and the structure of next steps. And how do we get these people onboarded and successful? There is a chance, small in this case, but if you just do a, yeah, it's free, go give it a try. And then we can talk about paid plans, or then you can upgrade yourself. There's a non-zero chance that they just won't do it or, or, you know, they'll drop the ball or something will happen. So that's a design thing for a future Maybe a future teardown if Harold will ever speak to me again. So that said, I hope you enjoyed these and that's that.