Watch These 33 Minutes To Become An AI Expert

On today's show, we're going to make you the AI expert at your business. In one episode you're hearing AI everywhere. There's a bunch of hype, there's a bunch of reality, and AI is really great at some things and really terrible at others. We're going to break all of those down for you and give you the map to AI success in your business. Let's get into today's show. We thought we would do a show where we kind of did a level set on AI and one of the reasons we wanted to do this, Kieran, is because I saw some, some data from our friends over at Carta that I thought was incredibly interesting. I wanna share that with you. So Carta, for folks who don't know, they basically run the cap tables and like the backends of startups and so they have really good startup data and what they shared was the number of employees when a subscription software company raises a series a round. So normally if you're a startup, you have an idea, you maybe raise a little seed money, that idea starts to work, then you raise a series a round. And what was interesting is this is kinda like 18, 2019 and then 2022 went to 23, then 2023 went to 27 and now this year it's down to 17. And I'm not here to say like, oh, AI is taking all the employees out of startups. I'm saying there are market conditions that are forcing companies to be way more efficient and because of that, AI is going to be accelerated by like the business needs of the world. And so if you're looking at this data you're like, oh yeah, I want to do more with less. We wanted to do a show around what can you actually do well with AI in your go to market now and what can you not do well like in your marketing and sales with AI and maybe you should wait on. So I know you got some opinions. I want to, I want to kick it over to you.

Yeah, I think even on this, the way I think through AI is like we are in time where companies want to be much more efficient. And I think that's a good thing because we went through the kind of COVID growth and then we are trying to get back to much more efficiency. I always think about AI as can either give you efficiency gains or productivity upside. Right. But they're actually one and the same thing and it just depends how you want to use that capital. So in that case we could say, hey, like AI can now do all of those things, you don't need those people. But I'll keep the people and just grow faster in some other way because I'm going to like re divert them to other high value tasks. So I think that's one way for us to like level set how we think about AI within companies. What we're going to do in this episode, and you want to stay tuned because I think we're the first real people to do this. We are going to step through like a traditional kind of B2B funnel, the go to market for marketing to sales and to customer success and give you a snapshot of just where AI is. Like where we've seen interest in use cases. We're going to talk to you where we've seen like value being created versus where it's not really there yet. And I think we can start with just some guiding principles about how we, you and I think about AI. The first thing I would start with, and I think it's really important for our listeners and companies is AI for the most part, for the most part is still like an enthusiastic intern.

Now.

It's got better over time, I think.

I like that it's enthusiastic though, because.

It wants to do anything you ask it to do. Now the only thing I will say on occasion, ChatGPT has basically told me to do the work versus it doing the work when it gets a little lazy. But it wants to please you ask it to do something and it tries to do that thing. And it constantly tries to do that thing even if it's wrong a lot of the time. Why am I making this point out of the gate? Because what a lot of people and companies are doing is not treating AI that way, right? So what would you do with a junior employee? You would onboard the employee. You would provide the employee with context about the role. You would give the employee context on what good looks like, like the task you are asking it to do. Here's what good looks like, right? So you wouldn't just go, go do work, you would actually teach the employee, the junior employee how to do that thing and what good looks like. And I think what I see a lot of companies do and a lot of people do with AI is just like, hey, go do the work, right? And then they're really miserable with the returns because the output isn't that good and they're like, ah, this thing doesn't work, I'll just move on. And so that's number one, you agree with my enthusiastic intern. That's a good way to frame It.

I would agree with it and I would add a slight modifier. It's like an enthusiastic intern that you have expectations is going to get it right the first time every time. Like so many people that I know when it comes to AI, it's like, oh yeah, I asked AI to do this thing once and like it wasn't good. And so AI sucks. But like you and I know is you have to iterate and work through just like you would work through with an enthusiastic intern is like you can't expect it to get right the first time, Right.

The second one is context is what creates most amount of value in AI tools. So what do we mean by context? Context is the data that we give AI. Now I think we're going to move to a world and we've had Ran Pishkin on recently where the web becomes much more closed. We made an episode to say data is a new Internet goal because data is how these models are trained. And now you can have propriety data and sell that data to AI tools or AI models or use it yourself. Really quick example here is if I go and tell OpenAI, hey, like coach my sales rep, right? Like here's some things they do well, here's some things they don't do well. Provide a coaching plan, it will be fine, you know, not that great, generic fine. Because it's trained on external data, it's trained on the Internet. Now if I take an AI app and I start to feed it the individual sales reps emails, the kind of funnel for that rep, like how their deals are closing the calls now, it can have better context to provide a better management coaching map. And so the more context you provide these AI tools, the better they are for your business. And I think that's why what I am seeing, and we're going to get into some of this, a lot of companies are actually building these things internally versus using anything off the shelf because they can actually get the context and map it to the way that they market, sell and service that business. So they can make it much more contextual. So that's the next principle is like context is really the differentiator between an average AI tool and a really incredible AI tool.

Yeah, so you're basically saying it's the depth and quality of really focused, structured and unstructured data. Structured data like fields within a CRM or some type of software product versus like calls, emails, very unstructured things. And what I think is interesting about this Kieran, is like anytime you go to a new paradigm, things that were previously impossible become possible. And I think the hard thing that you can have in wrapping your head around it is that like we would have never thought possible to. To say, like, oh, I can take the totality of one human's interactions and make it and have a fast and cheap and easy way to access that and predict next actions off of it. And that is really what we're saying. I think that's really hard for people to understand. It's like, oh no, I want everything across my whole company. And it actually turns out that when that data set becomes larger and less focused, sometimes the results are worse.

Yeah, exactly. Yeah, exactly. And I think we're going to get into some examples of where this shows up in the go to market. All these principles show up in the go to market. We are still in the assist phase, not the autonomous agent phase. So I've talked to a lot of.

Can you break down before you even what you like? This is, this is a little AI nerdery you're about to get into. Can you break it down into the plainest of the English for. For everybody, for us.

So if we think about your experience with AI, you're the person that's in control. We kind of call that human in the loop. But I actually think of it like human controlled and the AI is assist assisting you. Like think of the way that you would use AI for writing. Like I would go to it and I would say like, I want to write this kind of content for this kind of person. Suggest me a first draft. Oh, that's not good. Change it, change it this way. Give me another draft and I go in and I edit and I write it myself. That is actually a really good use of AI.

And to be specific, this is like ChatGPT.

Claude ChatGPT. I'm Claude Opus because we have it in Ireland. I have ChatGPT as well. I got rid of my Gemini Advanced. Sorry Google. Maybe I'll come back to you at some point. But they're my two staples are the two ones that I actually pay for. Cloud Opus is my writing tool. I use that above and beyond anything else. And agent would be what we have talked about before in the show. I would say, hey, go grow traffic for my brand and what the agent will do. And some platforms are starting to go in this direction. It will ingest the keyword data. It will look at the first page of Google for those keywords. It would say what is good about that content but what can be improved upon. It would go create content with those improvements. Baked in it would then go bake that into the post and it would go post them. Now I have a framework and some examples when we get into the marketing, go to market cases that actually you can think about how that's going to going to work. It's a proprietary framework. This is one of the things I haven't posted on LinkedIn yet. So do not post this. I will find you if you do it. But go follow me if you disagree with it. Troll the hell out of me. Why not? That's fun.

Hey guys, real quick. We've gotten a ton of requests in the comments and they're all about prompts. You all want the prompts we're talking about on that show. We see you, we hear you, and now we are ready to share the goods. Click the link in the description below to access our new AI prompt library. This resource features real AI prompts that Kieran and I have covered on this show. We broke them down by category and templatized them so you can fill in the blanks and make them your own. Now you have the frameworks you need to use AI, analyze ads, create brand campaigns, develop product stories, just to name a few of our favorites. Keep telling us what you need in the comments and we'll see what we can do to whip up more for you in the future. Now back to the show. Kieran's got a bunch of LinkedIn trolls everybody. So give him some LinkedIn love. But, but in all seriousness, Kieran is sharing some hot off the presses info today that we've been working on. So you're definitely going to want to send this episode. This is going to be probably one of the most forwarded episodes around your team, your company, your peers. So just we'll give you that heads up in advance that that has happened in a few.

So that's, that's the third, uh, fourth is.

Hold on Kieran. Before, before we move on, can I, can I add one thing? Can I, can I give everybody what I think is like the simplest definition of agents?

Yes.

Agents are really just automation with a goal versus rules.

Multi step automation with a goal versus rules. Yes.

Yeah. So right, because you, right now you have multi step workflows and HubSpot and lots of other tools that are rule based automations and agents are going to be able to solve for a goal and make their own tasks and everything based on what they're trying to do. Get to a goal. That's the difference. Like let's not overcomplicate it. Like if you are Talking to somebody in your company who's like, I don't know what an agent is. That is the simplest way to break it down for them.

One of the ways to think about AI and how you apply to your company, we're going to get into the examples, is what would you do if human costs go to zero? Now you do, there's, there's a caveat where we are right now. So one of the most obvious places you would say, wow, if human costs go to zero, well, then I would just apply human capital across the segment of my database that is lowest intent. I wouldn't do that today because I wouldn't make enough money back because low intent, human capital is high. But now I have AI. Hold on.

I don't think people are going to understand that. Can you, can you give us a refresh of the, the, the, the most basic version of what you mean?

Let's break up your database into like best qualified. Right? They are, they're, they're showing they've actually raised their hand and they want to buy your product. And then we have like medium qualified. These people are showing all of the signals that they want to buy your product, but they have not raised their hand yet. Like, that's a really good place that you can go in and do some prospecting, try to email them, get them on the call with, get them on calls for sales. Then you have like the cold part of your database which is like, these people haven't done anything in 90 days. Like, what the hell are they doing? Come on, like, wake it up, do something for us. And they're just disengaged. Right? When I say the human cost go to zero, that could be an immediate use case. And the reason I'm going to use this.

When you say human costs go to zero, what, like tactically, what do you mean?

I mean, I mean AI is free other than the API costs, right?

You mean the AI is able to do tasks as well as a human, as well as have to pay a human a certain amount of money to do that task. And now I can do that task for essentially nothing.

Okay, so one of the things you can like really think through is what happens when human costs go to zero. So when a AI can do a task in a similar way to a human, how would you apply AI to your business? The most obvious use case is we would try to give everyone a better customer experience, like through chat, through email, whatever it may be. Like that customer support is the most obvious use case. And now you have unlimited human capital through AI, why not try to improve the experience of your customers when they have a support ticket, when they're trying to buy your product? So that's the other mental model I would go through.

Yeah. And Kieran, if you're out there and you're watching this show and you want, you're like, hey, I really believe in the AI revolution and I want to be a winner of that. One of the ways you become a winner is you think about the second order impacts of things. And so you're basically saying, hey, we will get to a point and we're already there where certain tasks AI can do as well as a human. And well, that's interesting. You have a tactical question of like, how do I want to use that AI? But the bigger question is, like, what does that mean? Like, if every company can have an AI that's outreaching to people who are like, kind of barely interested in their product, then like, everybody's just going to get literally a thousand emails a day. And then you're like, oh, actually attention is going to become way more scarce. So I need to actually start building attention today. And like, credit to Gary Vee, who was on the show recently, his book Day Trading Attention. He's right. Like, he put that book out. Like, we'll look back five years from now. You're like, oh, that was like the perfectly prescient time to put that book out. But those are the type of insights that you're going to want to get and we're going to keep trying to give you so that you can plan and evolve your strategically accordingly. And sometimes it's not just how you use AI, it's how AI is going to impact everything you do.

Same thing we said. We talked to Rand Fishkin. What did we kind of talk about? We said that we are transitioning from the traffic acquisition era to the attention acquisition era. And the attention acquisition era looks much more like the marketing of old, where you try to like, do creative messaging and figure out who your customer is and where they're hanging out and try to grab some of that attention. And I think what AI gives us is the leverage to be able to monetize that attention much better. Because to your point, if we all have access to the AI tools that allow us to do incredible customer experience across email, chat, all of these different things, then all those channels just become table stakes. So you get no leverage.

Correct.

All right, so you want to get into some fun stuff. Let's get into like, what is happening across marketing, sales and customer success when we apply these principles. I've got some cool frameworks that I haven't brought you through yet. One of them, one of them I really like here. So what is happening in marketing today?

I love it when you pat your back on yourself for your own frameworks. Oh, this is because half the time it's really good and then half the time you're like going through it and you're like, oh, I thought this was really good but I want to change this thing now. So I'm excited which one we're getting.

I'm going to do these. I, I haven't fully, I haven't fully like articulated them out before. So I'm excited to try them on you. And I've got a little graphics I'm building in the background. But this one is really good, I think. So what's this is actually happening in marketing today? I will tell you that what's happening in marketing is that each individual role is going to have a full end to end tool that could interweave AI into every part of what they do. And I'm going to give you three killer examples and the loop to think about here. The framework to think about what these tools are doing. They are giving you inspiration. So they're building context using external data, right? They give you an inspiration board, just kind of follow along. Then they allow you to see through that inspiration what good is. Right? So the AI can break down why these things are good, templatize it for you and then allow you to scale creative testing. Right? So you can then choose, I'll give example, you can choose a bunch of these things and say, okay, test these things for my brand. And then it can actually re ingest analytical data to see what's working again. Now it's built an internal context to make recommendations better over time. So starts external context, then allows you to use the AI to break down what is good and what is relevant for you and your brand. Then allows you to do that at scale so you can test and build internal context to make recommendations better over time. So let's start with three core places. This is happening in content. There are some incredible content tools that basically ingest keyword data and then all of the content that currently ranks for those keywords, it will ingest that content and it will say, this is the breakdown. This is why this content is doing well. These are the recommendations on how you can do it better. Do you want us to assist you in creating that content? Now you can create that content much more rapidly. So now you could scale what good is through AI and start to publish that across all of the different places you want to publish it. So you have testing at scale within content. Now it also has built in analytics so it will say hey, this is performing well. This isn't performing well. This is performing well based upon that and the external context we have from what others are doing. Here's further recommendations. Do you want us to assist you in creating that? And I'm the writer and I have now I'm the content creator, content marketer and I have this little team that and it's a singular experience that does the end to end version of my role. Right. AI plugged in at every piece. I'll stop there because I have two other really clear examples. And so like that's the first thing that is again these are all today, these are not tomorrow.

Yeah, look, content is a great use of AI. At HubSpot we've got our content remix feature is doing great, doing awesome. People, people are loving it. So like we're seeing the usage and the data of that firsthand. I'll tell you the line around all this Kieran that I used earlier this week to see if you agree, which is it turns out AI is better at guessing at scale than a human.

Yes.

If you want the one line of oh, what are use cases right now that AI can do really well. It's anytime where you have a lot of information.

And you need to make.

A lot of choices and you gotta do that at like scale. So a lot of choices quickly. It's better than a human because a human just can't do it as fast as it have the time to do it. If the human could go through and look at everything perfectly, it would make a better guess. But not anywhere close to the same time horizon. Right.

And the thing we're lucky with digital is that we have a bunch of external data points that the AI can use. So if you're in content it can use the Google search results. Right. I know this is performing well because it's right there in the top three results and same. So the next one really has paid. This is happening in paid. There's killer tools out there. Builds inspiration boards, allows the paid marketer to just quickly scale the versions that are good and relevant for their brand, can create them in tool and then can ingest data to be able to do that for the paid marketer. And again you have external signals to tell you what good looks like. The other one is that is like going to happen but be saturated really really fast on the how do I acquire demand is and I think because Google search is going to diminish is the, the ability to basically have the AI create your, your icp, which is your ideal customer profile and just create a proprietary, like basically build your own little proprietary contact database. So it's all outbound, so it basically goes off. It says this is your icp. Here is a database that I've created just for you. And then it has AI tools built in that allow you just to go and create outreach to that and start to generate demand for your business. And again, each one of these is using that framework I talked about, right? It is building external context. Right. Content is building it from the Google content, from paid, from what's doing well in these channels, from the proprietary database, from what contacts fit your icp. Then it is telling you what it can do to scale good. Like here's how you can do more content like this that's relevant to you, here's paid ads that are relevant to you, here is leads that are relevant to you. And I can create the really great emails, the outreach and then I can do that at scale and then ingest internal data to get better over time. So that's the number one thing that I see happening is that individual roles, whether it's the paid marketer, the content marketer, the marketer who has to generate demand through these outreach tools, they actually have this kind of end to end experience where they can have an AI tool help them at every single point. So they're still driving the experience, but the AI is assistant in every single point. I think that's really interesting. You used to see a lot of disparate tools. I think they're actually going to consolidate into one experience because it's much better, it's much easier to actually use. And AI can then be woven into each and every spot and in a really interesting way.

Yeah, well what happened is you used to need a ton of different tools to do very, very specific little use cases and now AI can do all those little use cases and do them better. And the thing that actually matters is that AI have all the context as you're saying. Right. And you're more likely to have great context in one tool that has all your Data than across 10 tools that has bits and pieces of your data.

Exactly, exactly. Because what did we say in our principle? These all fall back to the principle. It's more context, more value in the AI tool. So that's marketing. The other place in marketing that there's like AI does have some capabilities. We've seen it in brand and creative. Now we started with Midjourney, Dall E all of these different tools, but they did not unlock creativity in the way that I can just go use these tools to create a brand campaign. If it would, it would be a pretty bad brand campaign. And I think what we're seeing now is tools built into platforms that designers use that videographers use, that have kind of AI built in or even run Runway, which is one of the better video tools. It still really needs like a pretty good seasoned videographer, someone who's really good at video editing and things like that to use the tool to have like really great output. So those tools help make creatives better and help scale their work better. But I don't think it's on. We have not, we are not at the point where, you know, all of us who want to doing more kind of video and these other kind of formats have the tools where we don't need someone with experience in those areas to go do the, go do that stuff.

AI creative tools are not good enough that they can be our outsourced creator yet creative yet. They can help you if you know what you're doing and do it really well. But. And they can help you do low fidelity tasks really quickly. But if you're talking about like a high fidelity brand campaign or very high fidelity piece of creative, like you need to have a creative expert and they're going to use those tools and they're going to have a really good output, right?

Exactly. Now there are, there are a bunch of other use cases in market. Like I'm trying to consolidate it into like the non individual use cases. Which means what are marketers doing who are just using ChatGPT, Cloud or Gemini? Like I do see marketers doing fun stuff like creative briefs. That is a really good use of these tools. Like if I'm a marketer, a lot of the reasons that things go awry when I'm working with creatives is because I just can't write good creative briefs. And actually if you go to Claude or any of these tools, it can write really good creative briefs experiment docs like if I'm a marketer I do a bunch of growth experiments. I can use it to create an experiment framework. It's really good for stuff like that. How do I craft an hypothesis like how do I do all of these different things? I'm not going to get into individual use cases. I still think it's hit or miss in terms of how marketers are using it. But in terms of how it's going to reshape how we do marketing, there are some of the best use cases. The next thing, if we get into sales, okay, that's how does sales work? Like we usually have some form of like a BDR and SDR who creates demand for the sales team in the form of meetings booked and then we have the sales team who sell it into that demand to create customers. We are. There is a revolution or an evolution going on in terms of the kind of BDR role. Like we have a lot of AI trying to digitize the BDR role. And if you think about the BDR role and I'm going to oversimplify here for the sake of this episode as like research, email call and kind of like LinkedIn messages and WhatsApp and SMS, I think the place in terms of where we are in practical terms is AI can do research and AI can do incredible email outreach, right?

Yep.

Now to your point, what happens when everyone does that really well, there's going.

To be some problems, but to your point, will take a little while for that to happen.

It's going to take. And I still think there's creat. Like the creativity in the email comes from the external data sources you can use and the fun things you can do. Right. Everyone is just going to have the email and say, hey, I see you like paid market marketing tools. Can you speak to me about paid marketing tools? If you're not the person who owns paid marketing tools, can you direct me to the person who owns paid marketing tools?

Right.

You know what I mean? Everyone's just going to use the lazy templates. The there's still going to be the separation between the people who are truly creative and the people who are not. Let's get into voice. So we are an interesting place with voice. Right. We had GPT 4.0 this year, which was a big leap ahead in terms of voice because of the latency. Right. The ability to have a back and forth because of the ability to do multilingual, because of the ability to actually interrupt the model for the ability for the voice to be changed so we can sound much more human. Like I don't know how far away we are, but we are still not at the point where I think you can actually replicate a kind of BDR type sale in a kind of. If you had a very simplistic sale and you had a very defined playbook because they train off your sales playbook. What I mean by that is most companies will Say, well, this is the way you should sell, this is the way you should handle objections. This is the way you should discover what that person needs and it's really well defined. Then it can kind of follow that and do a pretty good job if you're going. If you have a customer that goes way off and does things that the AI may not have seen before, it can kind of get glitchy. So the voice is, I think we are like maybe 12 months before we can start to really have interesting voice enabled agents that can do all manner of different sales. Now, not like selling snowflake large complex databases, but a lot of B2C sales are not complex. Right. I'm selling a informational course. I'm selling a packaged holiday that is super fascinating to me.

Well what's interesting and what I would tell everybody watching the show is like you have Talked to more AI startups that I would say 99.9999% of the population outside of probably the partners at the top 10 venture firms. You've talked to everybody and you are giving them the cheat sheet to where technology is today. Which you're like, wow. For most things, especially in sales, it is not yet there.

Right.

But I see on the horizon 12ish months from now there's gonna be a lot of stuff that is there.

Yeah, I think Right, exactly like this. Again, if you want to take one thing away from the show, this isn't a case of if it is all when. Now. I am definitely in the AI because I was trying to group AI people in my head this morning before doing this, which is like the AI who overly enthusiastic people and if you listen to them you might be disappointed when you get in to try the tools. I think I'm in that category where if you listen to me and then you go and use a bunch of the AI things you might, ah. Kieran is like not telling the truth or like he is just like too enthusiastic with this. And I agree, I agree. Like I am, I, I live in the future and sometimes I have to like pull myself back. But I still think I'm right, that what I believe will happen will happen. It's just a case of like when it will happen. Right. But the thing we're giving you right now is the things that are happening right now. And I think that's why this is valuable.

This is a dose of reality. And you're an AI optimist. I'm an AI optimist. We know the future is coming. It's just when it's going to be happening. But there are some things that are here, especially in marketing.

Yeah, and sales. Like I think the prospect and the other thing that's in sales is there's a bunch of great tools that basically can help you identify your net best fit company, which I know is machine learning and data science versus like Gen AI. But if you pair that with Genai, it can actually crunch all your internal app signals, it can crunch all your external signals and then it can say like hey, these are all your next best customers to reach out to. And then the AI can be plugged in to actually write the emails and do all the sequencing. The other thing is even a very tactical thing. I talked to a really great analyst who's spoken to a bunch of AI companies and there was just one small tactical thing that stayed with me that he told me about which is even if you're doing rep sequences and you want AI to create the sequences, the problem with a lot of sequences today is they get outdated pretty fast. You create a linear sequence about a company and that company then have a rand they raise or they have a bunch of news that go out about them but the sequence doesn't get updated. But the AI can basically update the sequence and rewrite the email so that it's always filled with like the current information. So like there's some like real use cases you can do in there and we in HubSpot are like working through all of these practical use cases and then the last one I'll leave with you. And in sales, which I think is fascinating because it's happening in engineering as well, is when you're a sales manager you don't actually get a ton of time to coach. Like you do coach your pressure the.

Quarter in the month. Right?

You got the what I mean like shout out to everyone who works in sales. It is tough. People are awesome, awesome people and they have to like multitask. They have to hold a bunch of things in their heads, they have a bunch of time sensitive things they have to do. So they're doing the forecasting calls, the pipeline in, they're digging in to try to figure out where things are working, where things are not working. So it's hard to step back and go like, oh well here's some real coaching information. And so what's happened in engineering, very similar, like they're always kind of debugging and things like that. So there's these kind of new tools that allow you to figure out where you should coach your engineers based upon the data that's crunched like here's a place they can improve and here's some like guiding points that you can tell them. I'm seeing a lot of that in sales. Right. There's a bunch of sales tools that, that can. Basically they've been trained on years and years of sales coaching material. Right. So they have all of the best sales coaching material that's ever been done. They're trained on that and then they go in and they ingest all of your sales team's data and then they look for places where your individual reps may not be performing like the best rep and they can start to provide the manager the cues. Here's a place where this rep is currently underperforming. Here's some coaching advice on how they could potentially improve. So the human is still.

That's hugely valuable.

Yeah, they're still like directing or they're still the one who is actually in control. But they get this guidance given into them by the, by the AI Again, I think we're like not great now, but I think that is going to be a great use case in the future. Another great use case from an awesome guy called Vedant who I was lucky to spend some time with. He put me in touch with this really cool company called Hyperband. They were actually The HubSpot Gen AI Conference in San Francisco. And so shout out to Atul, I'm going to meet up with him soon. And they're doing something that I hadn't thought of, which is the opposite of what I've been thinking about, which is they can actually train to mimic your customer. So basically they can actually take all of your data and then replicate your customer. So why would they do that? Because when I'm a new salesperson, I want to be able to train my sales techniques and the playbook. Right. And you can program the AI agent to act like your customer in a bunch of different scenarios and then have the a sales rep call it and actually grade the sales rep against that call. So that's another use case I hadn't thought about, which is how you train your sales rep with the AI actually mimicking your customer.

The sales trading industry is a massive, massive industry that nobody thinks about and AI has a enormous potential to disrupt it.

I'm going to stay out of customer success and support. I think that is the most well known. Basically we are going to have AI agents do a lot of the customer support role. It makes a ton of sense. We've seen public examples from Klarna. It's able to train your knowledge base. It's able to train on internal data, answer all of the questions in any kind of form. I think when we get voice enabled and when we get video avatars, like, it's going to be incredible because. Because I think everyone. Customer support is one of the places that lets a lot of companies down. I think it's going to be incredible where every company can actually provide really stellar customer support. So that. That is where I think we are. There are some of the more interesting things that where we are today in terms of AI go to market. From all of the research that I've done across different companies and talking to, you know, different people so that I.

Think everybody watching was a masterclass. Kieran has spent. This episode is a culmination of like hundreds of hours worth of conversations and work between startups and doing the work in the AI field. And like, we'll be back. We're working through all of these problems and we're going to tell you what works and what doesn't work. But this should be your map to where you should be focusing now with AI and where you might want to think about focusing on AI in the future. And as always, we're going to be here on the show to help guide you along the way. So hit subscribe, make sure you check out future episodes. We'll be back with you again real soon.

This data is wrong.

Every freaking time.

Have you heard of HubSpot?

HubSpot is a CRM platform where everything is fully integrated.

Whoa. I can see the client's whole history. Calls, support tickets, emails. And here's a task from three days ago. I totally missed HubSpot grow better.