6 AI Marketing Strategies To 3x Conversion in 2025

Summary

Podcast Summary: Marketing Against The Grain – Episode 6: AI Marketing Strategies To 3x Conversion in 2025

Introduction

In Episode 6 of Marketing Against The Grain, hosted by HubSpot Media, Kipp Bodnar (HubSpot’s CMO) and Kieran Flanagan (HubSpot's SVP of Marketing) delve deep into the transformative role of Artificial Intelligence (AI) in modern marketing. Released on August 5, 2025, this episode explores innovative AI-driven strategies aimed at tripling conversion rates by 2025. The hosts provide actionable insights, supported by real-world examples and expert opinions, making it an invaluable resource for marketers navigating the AI revolution.

---

1. The Evolving Landscape of Marketing in the AI Era

Timestamp: [00:00 – 03:10]

Kieran opens the discussion by acknowledging the uncertainty many marketers face in today’s rapidly changing AI-driven environment. He poses the fundamental question: "How do you even do marketing now?" (00:00). Kipp echoes this sentiment, highlighting the frequent queries from founders and CEOs about adapting to AI's impact on marketing strategies, team structures, and rising ad costs.

A notable insight shared by Kieran (03:10) is the reassurance that marketing roles are not disappearing due to AI. Instead, the synergy between human creativity and AI intelligence is set to redefine successful marketing practices. Kieran emphasizes, “The best marketing in the world is going to be done with humans in partnership with AI.”

---

2. AI and Job Security in Marketing

Timestamp: [03:10 – 04:18]

Kieran references a Microsoft job report outlining roles most susceptible to AI disruption, noting with relief that marketing, advertising, and sales agents are not at the top of the list ([01:33] – [02:03]). The discussion pivots to the broader implications of AI on creative professions. Kieran draws a parallel with AI’s dominance in chess, stating, “the fact that AI was better than a human at that thing did not slow that craft down” (03:36), suggesting that human creativity in marketing remains irreplaceable.

---

3. Building an AI-Driven Marketing Playbook

Timestamp: [04:42 – 05:06]

Transitioning to actionable strategies, Kieran introduces their "AI marketing playbook," a curated set of six AI strategies essential for modern marketers. He underscores that this playbook represents the initial steps towards integrating AI into marketing, with more comprehensive tactics to follow.

---

4. AI Search: Maximizing Share of Voice

Timestamp: [05:06 – 09:14]

Kieran delves into AI search optimization, emphasizing the importance of increasing a brand’s share of voice within Large Language Models (LLMs). He explains:

• Share of Voice: “How many times is your product given in an answer to a query related to your product or service?” (06:39)

• Citations: “How many times did our CRM get mentioned alongside great for small businesses...?” (07:30)

Kieran highlights tools like xfunnel and UseBear AI that help marketers monitor and enhance their visibility across various LLMs. A compelling statistic shared is that by 2027, an estimated 95% of all B2B buyer journeys will start on an LLM (07:39).

Kipp adds, “You have no choice but to make a bet to focus on search in an AI first world” (09:14), reinforcing the necessity of adapting to AI-driven search paradigms without abandoning traditional Google search.

---

5. Mastering Short-Form Content

Timestamp: [09:44 – 16:10]

Kieran expresses his enthusiasm for AI video tools, particularly VO3, a Google tool for creating short-form videos. He demonstrates how AI can streamline video production, enabling the creation of consistent characters and engaging narratives. Kieran remarks, “VO3 is so good” (12:08), showcasing its potential in capturing audience attention.

Kipp reinforces the shift towards multi-modal content consumption, stating, “The Internet is going to get fully optimized... there’s going to need to be a lot more video, both short mid and long form” (13:32). They agree that mastering short-form content is essential for grabbing and maintaining audience attention in an AI-saturated digital landscape.

---

6. Code-Powered Experiences: From Content to Interactive Tools

Timestamp: [16:10 – 18:04]

The discussion shifts to the transition from traditional content lead magnets (like blog articles and white papers) to interactive, code-powered applications. Kieran cites a statistic that 70% of B2B buyers prefer self-serve tools and interactive content (16:43), illustrating the demand for more engaging marketing tools.

Kipp shares his experience with creating a fictional landscaping company’s budget estimator tool, highlighting how easy access to coding resources enables marketers to develop valuable applications without extensive technical expertise. Kieran suggests leveraging platforms like Replit to simplify the creation of customizable web apps, enabling marketers to offer tailored solutions to their audiences.

---

7. The Importance of Intent Engines and Signals

Timestamp: [18:04 – 23:33]

Kieran introduces the concept of an intent engine, a critical component for harnessing both internal and external intent signals to refine marketing efforts. He defines:

• External Intent Signals: Indicators such as job postings, hiring activity, and product launches that suggest a target audience’s intent to purchase ([19:26]).

• Internal Intent Signals: Actions taken by users on a website, like visiting specific pages or utilizing freemium product features, indicating a readiness to engage or buy ([21:08]).

Kipp emphasizes the necessity of context in AI marketing, stating, “It’s all about... what can you know about them that really matters to them” (22:05). They discuss leveraging HubSpot’s connectors with ChatGPT and Claude to analyze sales and marketing data, identifying additional data signals to enhance marketing precision.

---

8. Enhancing Sales Productivity Through AI

Timestamp: [23:33 – 26:23]

The hosts explore how AI can alleviate administrative burdens on sales representatives, allowing them to focus more on selling. Kieran points out that currently, sales reps spend only 28% to 33% of their time selling, with the rest consumed by administrative tasks ([24:10]). By automating pipeline-building and prospecting tasks through AI, marketing can significantly boost sales productivity. Kipp highlights the economic benefits, suggesting that freeing up sales reps’ time can effectively “double the size of your sales team” (26:23).

---

9. Micro Audience Campaigns: Hyper-Targeted Marketing

Timestamp: [26:23 – 29:42]

Kieran introduces micro audience campaigns, which involve targeting highly specific segments of an audience based on detailed intent signals. He shares a compelling statistic: targeted campaigns can increase click-through rates by 209% and conversions by 3 to 5 times (27:48). By leveraging AI tools like Perplexity Labs and LLMs, marketers can create personalized campaigns for smaller, highly engaged groups rather than broad, generalized audiences.

Kipp elaborates on the cost-efficiency of AI, explaining that with the same budget, marketers can run 3,000 micro campaigns versus a single large campaign, thereby maximizing ROI and revenue.

---

10. Essential AI Tools for Modern Marketing

Timestamp: [30:31 – 32:28]

Kieran provides a comprehensive toolkit for marketers looking to implement the discussed AI strategies:

• AI Search: xfunnel, UseBear AI
• Short-Form Content: VO3, Flow, Captions Mirage
• Code-Powered Experiences: UseReplit, Agent AI
• Intent Engines: Perplexity Labs, Grok, Claude, ChatGPT, HeyGen

Kieran advises marketers to prioritize these tools to build an effective AI marketing playbook, tailored to their specific needs and audiences.

---

Conclusion

Kipp and Kieran wrap up the episode by highlighting the transformative potential of AI in marketing. They encourage listeners to embrace AI tools to enhance their marketing strategies, automate mundane tasks, and create hyper-targeted campaigns that drive substantial conversions. The hosts promise future episodes to delve deeper into specific tools and strategies, ensuring marketers stay ahead in the evolving AI landscape.

Notable Quotes:

• Kieran (03:10): “The best marketing in the world is going to be done with humans in partnership with AI.”
• Kieran (07:30): “Citations are for us, how many times did our CRM get mentioned alongside great for small businesses...?”
• Kieran (12:08): “VO3 is so good.”
• Kieran (27:48): “Targeted campaigns can increase your click through rate by 209% and conversions by 3 to 5x.”
• Kieran (26:23): “You're literally doubling the size of your sales team because you're getting more than half their time back.”

---

Further Resources

For listeners eager to explore the AI marketing strategies discussed, the hosts have curated a playlist of relevant episodes and promised detailed deep dives into specific tools like VO3 and Perplexity Labs. These resources are accessible through the links provided in the podcast’s description.

---

Final Thoughts

Episode 6 of Marketing Against The Grain offers a comprehensive roadmap for marketers aiming to leverage AI to its fullest potential. By blending human creativity with AI's analytical prowess, Kipp and Kieran present a future where marketing is more efficient, personalized, and impactful. Whether you're just beginning to explore AI in marketing or looking to refine your existing strategies, this episode provides valuable insights to help you thrive in the AI-driven landscape of 2025.

Transcript

[Kieran]

On this episode, we are going to go into something that everyone wants to know how to do and that is marketing. How the do you even do marketing? In the world of AI, everything is changing so much. If we were going to build a marketing plan from scratch, this is exactly what we would do. With all the plays and the actual AI tools we would use, that's what we're going to cover in this show. So, Kip, are you excited to talk a little bit about marketing on one of the shows? It's been a while.

[Kip]

Look, I want everybody to know, first of all, it's been a little while since Kieran and I have even been together on a podcast and there will be a couple of weeks where we're probably got some more solo pods. But you and I talk to founders and CEOs and everything, honestly, most days of our lives and we are getting this question like, how the hell do you even do marketing now?

[Unknown]

Right.

[Kip]

Everybody's like, this has changed. Like, what does my team look like? What strategies do I do? I'm not getting any traffic from Google, My ad prices are going up. What the heck do I do? And I think we just felt compelled, Kieran, to just try to take a few minutes to calm everybody down, give everybody some tactical, practical things that they can go do. Things that we're doing.

[Unknown]

Right.

[Kip]

And that we're finding value from and getting more and more value from as we're learning them.

[Unknown]

Right, Right.

[Kieran]

And I actually did have this. Did you see the Microsoft job report for AI? What's going to be lost and what's not going to be lost?

[Kip]

No, I missed that.

[Kieran]

These are the jobs that Microsoft believe will be most impacted by AI. The reason I'm showing this is for all of the marketeers. You're not on the list. So basically all the high volume, entry level, white collar jobs are on here. Nothing around marketing, advertising, sales agents, web developers are on here.

[Kip]

But they're at the bottom half of the list.

[Kieran]

Yeah, they're in the bottom half. What's really the library science teachers post secondary. Jesus. These are the ones that it believes AI has most applicability to, to be able to do these things. I'm struggling with the hosts and hostesses.

[Kip]

Like they believe that robots are going to take.

[Kieran]

How does AI is this like humanoid robots? Concierge?

[Kip]

AI is not just digital AI, but they're talking about AI robots, the humanoid robots.

[Kieran]

These are the ones Ed released that are kind of safe. So I could be a massage therapist.

[Kip]

There's a new robot massage place in Boston Is there?

[Unknown]

Yeah.

[Kieran]

Would you let a robot massage you?

[Kip]

For sure I would. I think it's better than a person.

[Unknown]

Right.

[Kieran]

That could be the start of the Terminator movie. Masseuse, Robot Goes Wild.

[Kip]

I mean, I'll try anything once. Look, one of the things, Kieran, that I think you and I are coming to right now is that if you're a marketer, your job's not like going away because of AI and so you should, like, take a breath and feel a little bit of peace about that. But I think you and I also are seeing, and we believe that the best marketing in the world is going to be done with humans in partnership with AI.

[Unknown]

Right, Right.

[Kip]

It's not going to be just AI or it's not just going to be humans. The humans who are using AI, they're going to win the marketing results game.

[Kieran]

Yeah. I think the thing I took away from that list is there's not a lot of creative jobs on it, which is great.

[Kip]

Totally.

[Kieran]

I know AI can't do creative tasks, but one of the things to think about is what AI has already surpassed humans on and what has happened to those things. So let me give you a good example. I think it was maybe 20 years ago that humans started to. And people are going to scream at me if I'm wrong in the comments, but I think it was like 15, 20 years ago where AI was first able to beat a human in chess.

[Kip]

Yes.

[Kieran]

And so you would say, okay, well, AI has now surpassed human in chess. And so de facto we probably would not want to do that anymore. And actually, chess has never been more popular. It's the most popular it has ever been. So the fact that AI was better than a human at that thing did not slow that craft down. And I think there's a lot of examples of that where AI maybe surpass certain people in creative tasks, but people want the thing from the human.

[Unknown]

Right.

[Kieran]

I think that really matters for a lot of the content we create. I think we've talked about things moving towards personality, LED content where you really think about not just the content, but who's giving you that content. I think it's why we believe that most brands will have some sort of creator part to their content team doing that content, which we have. But I think in a lot of cases, human creativity plus the AI assistant is where marketing will end up. And that's what we are going to show you here is like the AI marketing playbook. If we were going to build a AI marketing playbook from scratch, we had to go into A company we had to start today, what exactly would we do? And we're going to go through six things that would be part of that playbook.

[Kip]

And I would say this is a very abbreviated playbook. Right. This is like the list of the first new things we would start doing.

[Unknown]

Right? Right.

[Kip]

Like we'll have more comprehensive stuff in the future. But this is like if you're just like, hey, I want to go from the old way of doing marketing to a newer way of doing marketing that's going to be more effective in this new AI world. Here are the right places to start.

[Unknown]

Right.

[Kip]

And the stuff to start learning and the stuff to start having tools to help you around.

[Unknown]

Right.

[Kieran]

So I did the lovely graphic. I was up late last night. I was trying to get ChatGPT ImageGen to do this. I could not do it, so I reverted to canva, which is another good example of where these AI tools fall short.

[Kip]

Have you used canva with Claude yet?

[Kieran]

No, we talked about this. I just think the ability to edit in the LLM for a lot of creative assets is just not there yet.

[Kip]

Correct. I just didn't know if Claude could give you a better first pass in Canva faster.

[Kieran]

I should try that. I was going to try it in ChatGPT and I just reverted to my old ways. Good example of like, you should always try the new thing. I should have started there. All right. The first thing is non controversial. It is AI search. It is something that we have made a big bet on. I think you want to grow share of voice and LLMs. We can talk about the fact that this is share of voice and not conversions, which is a whole thing. People are struggling with a whole bag of worms. We can tell you some things. We had Asia on this show. You should go back and watch that show around how to do AI search. We talked about LLM optimized niche pages. So instead of having one product page that's optimized for like three to five keywords, which is the old Google search way, you conversate with an AI assistant. So you ask lots of questions in different ways. And so instead of having one product page, you actually need to have lots of variations of your product page for all of the ways that people can ask questions about that. And I think that's one of the things that we found works pretty well. That and the citations.

[Kip]

Yeah, when you say citations, Karen, while we're talking about AI search, many people watch this are probably hearing about for the first time, what the heck is share of voice? What is a Citation and what's the real business value of AI search? Maybe just touch on those three things for us.

[Kieran]

This is great. So share of voice would be music to all brand marketers ears that we have a performance marketing channel now being measured against a brand metric. So now performance marketers will have to deal with the CFO asking, hey, is this actually having any impact? But share a voice is basically how many times is your product given in an answer to a query related to your product or service? So basically what you do in these tools is I will go in and I will say, okay, for our CRM, we're going to upload 100 different questions that people may ask that should trigger an answer that includes our product. And then we would say out of those hundred, how many times did we appear versus our competitors? And that's some way to get towards your share of voice. So it really is, how visible am I for the questions that I care about?

[Kip]

Okay, so that's share of voice, what are citations?

[Kieran]

And then citations are, is your product or service mentioned with the key phrases in that question that you want to appear for?

[Unknown]

Right.

[Kieran]

So citations are for us. How many times did our CRM get mentioned alongside great for small businesses? Or how many times did our CRM get mentioned along great enterprise features for large companies? And so you want to really care about citations. Am I getting mentioned and trusted sources for the things that I want to be known for in the LLMs? So when someone's asking questions about these things, there's enough co citations about your product which means that you get mentioned with those words in lots of different sources. And the LLM goes, oh cool, I know HubSpot's a great answer for that. HubSpot CRM is a great answer for that because I've seen it in so many different places mentioned for a great CRM for small businesses. And the thing to think about there is these LLMs like OpenAI, Gemini, Claude actually have certain preferences for which sites they trust more than others. It's not like Google where Google owned all of the kind of search queries and they basically looked at links and domain authority. These LLMs really have partnerships with different sites and relationships with different sites. So they trust different sites more than the others. Right. And I think that's an important thing to know is whatever LLM you are trying to appear in, you should really find out what are their most trusted sources to understand if your product or service should appear in the answer. And I think this is a wild stat here that like by 2027 it is estimated that 95% of all B2B buyer journeys will start on an LLM. So this is why you want to make that bet today.

[Kip]

Yeah. I think that you have no choice but to make a bet to focus on search.

[Unknown]

Right.

[Kip]

In an AI first world. And it doesn't mean you completely abandoned traditional Google search.

[Unknown]

Right.

[Kip]

But you have to start moving some of your time and effort to this new evolving way that search is happening.

[Kieran]

Exactly. So I think this one great tools Xfunnel. There's a tool called UseBear AI. Basically you can do a bunch of things in OpenAI and ChatGPT to help you build all these different optimized pages. All right, next up, this is my favorite pastime.

[Kip]

You love AI video, baby.

[Unknown]

Oh.

[Kieran]

So the tools here you have a tool called xfunnel. And so what xfunnel will allow you to do is show your visibility in these LLMs. So all the different LLMs, you can see how visible you are, you can see your share of voice, you can see how the things that you're doing are increasing your share of voice over time. A really great tool. We were a customer. There's one called UseBear AI that I caught up with the founders very recently. It's a very, very cool tool. Again, very similar to xfunnel. They have a couple of differentiations that they can do. One of the things I find fascinating they do is they take a snapshot of the answer every day to a certain query. And what you can see is how often the LLMs change their answer, like just subtly. But they're never given back the same answer each and every time, which I think is super important. These Query map and LLMs, this is basically you can fine tune ChatGPT and Claude to actually help create these LLM optimize niche pages. And so they are the tools that can really help you do AI search. All right, let us get into master short form content. So I think this stat is wild. 65% of B2B buyers prefer short form content to learn about products or services, which you would really think about as a metric for B2C.

[Unknown]

Right.

[Kieran]

You would normally not associate this with B2B. And so I think a big part of this is how you can grow influence in text and video using AI. And so I want to go through a quick example of what I have been building in VO3 to show you how cool these tools are. And it is a opener to an incredible up and coming movie called a Timeless Marketer. So let me just Bring it up.

[Kip]

Just so everybody knows, Kieran's in love with AI video tools. They're his favorite thing in the whole world right now.

[Kieran]

If I could just go and do this full time, I actually would. Right now, it's like bringing all my wildest dreams to life. All right, so what I will say is VO3 is a Google tool, a video tool that allows you to create eight second clips. And so I'm gonna actually tell you how I managed to create a 24 second clip. I'm actually trying to get a 5 minute mini movie together. There's a lot I've learned here about using VO3 over the last week. The big thing here is, like, figuring out how to keep the character consistent across clips is not easy. That was one of the things I really figured out. But how freaking good is this?

[Kip]

So good.

[Kieran]

VO3. It's so good.

[Kip]

I think people need to understand. It's like you just typed text in and it generated these videos and you edited the videos together.

[Kieran]

We would do a whole episode of VO3 because I'm going to do. I'll do a five minute and I'll tell you everything I learned. But, like, you have to actually build character prompts. That's how I've done that. So I build a detailed character prompt, and then every time you create a new video, you insert the character prompts. So you can build basically a prompt library of characters who appear in the movie.

[Unknown]

And.

[Kieran]

And then anytime you want them to appear in a scene, you would just, like, add that prompt. So there's a bunch of, like, insanely cool you can do. But that's why I think for me, short form content is probably the thing that I'm gonna spend all of my time on.

[Unknown]

Ooh.

[Kip]

Why?

[Kieran]

Because I think it's how the Internet is going to work. Cause do you not think that, like, we're being taught as Internet users to just consume short form content over anything?

[Kip]

Okay, so here I would reframe it just a smidge. I think tomorrow we're recording with Alan Go, who is the founder of Willow, which is like a voice AI app. Kieran and I had a whole conversation with him and he was talking about, you know, how voice is really good for, like, dictation and creating stuff? It's really bad and slow at consuming stuff.

[Kieran]

Yes.

[Kip]

And I think what the Internet's going to move to is you are going to be in the perfect modality to get the information as clearly and as quickly as possible.

[Unknown]

Right.

[Kip]

Whether that's text, whether that's voice, Whether that's video, what have you.

[Unknown]

Right.

[Kip]

And because that before voice and video were so much more expensive than text, we used text for most things. And now we're going to play a big rebalancing and you're going to see a huge growth in voice and a huge growth in video, especially short form video. So when you say that's how the Internet's going to work, I kind of think of it as like, yeah, the Internet's going to get like fully optimized. And for the Internet to be fully optimized, there's going to need to be a lot more video, both short mid and long form video.

[Unknown]

Right?

[Kieran]

Yeah. I do think if you can master shortfront video, you can master like attention grabbing.

[Kip]

I agree.

[Kieran]

Mastering attention grabbing, actually, even if you're going to do long form is a pretty great skill to learn because what's going to happen, right, like every platform is being crushed under the weight of AI generated content. And so the ability to acquire the first second of someone's attention, where they stop to scroll is probably going to be like the number one skill that anyone can actually master. Because if you can't do that, you can't get them to consume anything. And so that's why I had here like master written hooks and short form storytelling to kind of capture that intention. I had LinkedIn X YouTube, but really all platforms can work for short form.

[Unknown]

Yeah.

[Kip]

Your argument is like, look, the don Draper of 2025, it's short format. Fans would be just obsessing about being the best hook writer for short form videos of all time, right?

[Kieran]

Yeah, exactly. They would just obsess over the first 60 seconds at any asset.

[Kip]

I agree with that.

[Kieran]

Whether you're reading that asset, listening to that asset or watching that asset. And I think short form content is a great way to master that skill set. The tools I put down here was VO3, which is awesome. There is a tool called Flow and Flow actually is built on top of the VO models and what it allows you to do is start to create real movies. You can actually start to create scenes. Unfortunately for me, I don't know what it's like in the US that tool is still on VO2 when you want to add multiple scenes. But you could just go in and use VO3 in Gemini if you can't use Flow Captions Mirage, which we had the founder on, that is a really cool tool.

[Kip]

Very cool.

[Kieran]

I thought I was blown away by how great that tool was. So this allows you to create great short form video and then if you're doing text based. We probably will do an episode in this. I started working on a episode for Claude where you can basically create write an app within Claude, fine tuned to your own styles. That's really great at hooks.

[Kip]

I think we should do a show just on like short form hooks and like building a project or an agent to do short form hooks and do just a whole show on that.

[Unknown]

I agree. Yeah.

[Kieran]

So you're doing your AI search now you're mastering short form content. These codepowered experiences I covered in another episode. We have covered a bunch of this in singular episodes.

[Kip]

You and I both have done a codepower experienced show. You can search lead magnet. It was my episode where I go through and show cloud artifacts on how to basically build lead magnets. But really what you're talking about here, Kieran, I think is that people used to give away blog articles and white papers and now you can just have an LLM do all that for you. So you need to give away instead of information action.

[Kieran]

Yes.

[Kip]

And the best way to do that is some type of really lightweight free code driven application. And like the best example that I had, I did like this whole fictional landscaping company. And so it was like a budget estimator, a curbside appeal grader. Things that somebody who was thinking about redoing their landscaping would care about, Right?

[Kieran]

Yeah, exactly. Again, I've pulled in the why here? 70% of B2B buyers prefer self serve tools and interactive content. You did a great episode to show how you could move from like content lead magnets to code lead magnets. But really marketers used to create content because it was way too hard to create code.

[Unknown]

Correct.

[Kieran]

And now it's like as easy to create code. I know developers will hate me saying that, but for the things that marketers want to build that are like lightweight things, it's as easy to create code powered experiences as it is content powered experiences. And I think that is a huge change in how we're going to market to audiences.

[Kip]

I think that's true.

[Kieran]

One quick idea I had in a video that I put out around this for repl.it was that very thing that you built. Shouldn't replit just have a library of drag and drop elements that allow people to create all those kind of lead magnets 100%.

[Kip]

And just all the basic front end and back end components that you can remix into like market specific little web app? For sure.

[Kieran]

Exactly. The next is you have to have a great intent engine that is going to power all of your AI experiences and that's really, really important. So your intent engine is really how you get your internal and external intent signals around your ideal customer profile. So your ideal customer profile is who are the audience that is going to actually buy your product, buy your services. There is more intense signals than ever before now because we have so many great tools that can go and scrape external signals or gather internal signals. The kind of power of everything else you build in AI is going to be powered off this really incredible intent engine. And again, we see this actually working. So 93% of B2B marketers who have tried to create experience around intent signals see a huge increase. Intent signals are things like they would suggest that you are a good person to connect with. So an example of an external intent signal would be you change jobs, right? And when you change jobs, that's a good signal that you can market to that person and market your tool. They may get promoted. External intent signal, maybe a startup gets a funding round. They are very specific to your niche. And then internal signals are that person is showing intent that they are interested in purchasing your product.

[Kip]

So this I think is the most advanced concept that you are sharing with everybody today. And what you're really saying here is that in an AI world, context matters, right? And we say that on no matter what we're doing with AI, it's all about, oh, do we have the context of who we're talking to, how to do this thing? And one of the things you need context on is not just who you're talking to, but what can you know about them and what can you know about them that really matters to them? So that when you engage with them, you can reference that and connect with them on that or even set the timing of your message around that. To really transform and get really different results in your marketing, you need to transform how you time and personalize your messaging. And you can't do that without really good intent data. And Kieran, maybe you could give everybody a couple specific like types of like data fields. That would be really good, emblematic examples of what people would be doing here.

[Kieran]

So some of the best external intent signals are things like job postings and hiring activity. Actually like hiring is one of the examples I showed. I did a quick video for everyone on our channel around how to build this thing which we're going to get into called micro audience marketing. And mine was really based upon small audiences who are hiring. So looking at companies and who are hiring job profiles that would be suitable for HubSpot, that the fact that the executive or team member on LinkedIn could change their job. This is a signal of budget shift. There's a bunch of like thermographic changes, new product launches, investment would be one. So like a common one companies use is that they just raise because in just raised they were looking to upgrade their tech stack.

[Kip]

Exactly.

[Kieran]

You can get a bunch of like data from these third party sites. So are they viewing certain product pages, like all of those kind of good things? Again, on third party sites you can get search signals. Is there like an increase in search on some of the sites, like Stack Overflow, Hacker News. But these external signals that would say, hey, they're in the market and researching something that is relatable to my product. And so they're kind of examples of external signals. Internal signals are very unique to your business. So they may be. Someone visited a certain page on your website, someone is viewing your pricing page. If you have a freemium product, someone has taken certain actions within the freemium product. And the word you use is the right word. Right. The more context I can gather, the more personalized and I can make my AI experiences. And actually the more context the AI experience has, the better job it can do.

[Kip]

Exactly.

[Kieran]

And so that context layer, which is your intent engine, is going to be do or die for having an incredible AI powered marketing machine.

[Kip]

Kieran, I think I have an action item here that would be good. You can tell me if you agree with this. Let me pitch it to you and to the audience. So if you don't have a lot of intent data or you're just kind of getting started here, you're like, okay, cool, well what data do I need and where am I going to get it?

[Unknown]

Right.

[Kip]

And Kieran, you even here, like there's data all over the place. It kind of depends on who you're talking to and your market and your industry and your region, all those things. Well, at HubSpot we have our connector with ChatGPT as well. As we just launched Claude this past week. And I think, Kieran, I would go in and I would do a prompt for ChatGPT or Claude and say, look at my sales and marketing data and tell me what other types of data signals would help improve my results. And then I'd have it return those results and then I'd have it do a deep research project of the best ways and vendors to get that data.

[Kieran]

Yes, right. That's awesome. So you use the HubSpot connectors.

[Kip]

Wouldn't that work?

[Kieran]

So ChatGPT and Cloud, you ask it, what are the improvements I can make to enrich these Personas. Like what data?

[Kip]

Yes.

[Kieran]

And then you take that and you do a deep research project and you say what are the best sources to get this data correct. That is a great way to start. I think people should go do that.

[Kip]

I think 10 to 15 minutes you would have a few vendors that you could go and talk to and probably instantly make your data better. Obviously we have some data in HubSpot. There's a bunch of different amazing data vendors out there, but I think that process would get you what you needed to know.

[Kieran]

Yeah, I would say that a lot of the interest in external data is you can just get through something like perplexity.

[Unknown]

Yes.

[Kieran]

There is a big difference in what perplexity will give you via the API versus what you will get via the chat interface. So just keep that in mind. All right, so this would be music to marketers ears because I think one of the things that will happen is marketing will actually help our friends out in sales and start to eat up more of the customer journey. Not as in take their roles, but actually allow sales to get back to what they probably would love to do, which is selling. And so the average REP today spends 28% to 33% of their time selling.

[Kip]

It's wild. Shouldn't they be spending like 80% of their time selling?

[Kieran]

Yeah, and I'm sure a rep would love to spend 80% because it would make way more money.

[Kip]

Way more money. Literally 3x more money.

[Kieran]

Yeah, 3x more money.

[Unknown]

Right.

[Kieran]

And we all know sales reps, they like to make money. So why is this happening? They have to do a bunch of sales admin. And so one of the things they actually have to do is build pipeline. And to build pipeline they have to do prospect. And prospect is a lot of what a rep has to do, whether you're doing it through these different roles, SDRs, BDRs, or just a sales rep themselves. And I do think there is a world where this becomes part of the marketing engine because marketing are able to automate all of this. Well, the text based portion, they can automate all via AI. We have done that in HubSpot to great results. But they can also then start to use AI to create a call script, again using your context layer. So you have a great call script for the rep and you can start to actually do most of this work for the rep. Now obviously the rep is still going to call the person, but everything else you can kind of automate there.

[Kip]

Yeah, every company I talk to, they want more sales productivity. And one of the ways to get more sales productivity is to give your reps more time selling and less time admin. Writing emails, doing all the stuff that the reps can do, can do very well, but maybe like to do less.

[Unknown]

Right.

[Kip]

And then, Kieran, I think the point we should make here is that the whole buyer's journey is kind of getting shifted, right? The buyers have access to way more information and because of that they need marketers and sellers a little less. And so everything that's happening is kind of getting shifted around. And one of the things that's happening is that marketing is taking more of what may have been thought of the early sales process previously, but we think that's a big economic upside because it's going to mean that your salespeople at the end of the day are going to be much more productive.

[Kieran]

Right, I agree. Like marketing are really the, the automation team, in a lot of cases, a lot of the stuff that they can start to do is automate out all of the things that some of these other teams have got tasks to do. But actually then they could be freed up to get back to selling. And it's great for the company because if you have a team implement this and their reps can sell for 60% of the team, it's like you've doubled the size of your sales team.

[Kip]

Exactly.

[Kieran]

That's incredible, right? Like you're, you're literally doubling the size of your sales team because you're getting more than half their time back. All right, the last one is micro audience campaigns. What is this? All right, so we should start with the stat here that targeted campaigns will increase your click through rate by 209% and conversions by 3 to 5x. So targeted campaigns is a way that you can really increase your results as a marketing team forever. We have done like segment level marketing, right? So we have firmographic and demographic data. Then we build these kind of groups of people that are all like each other, kind of like loosely like each other, and then we market to them because that's just the way the tools work today. What AI is going to allow you to do is market to like small slices of your audience because AI is going to do most of the work. So instead of marketing to 300,000 people, you may market to like 3,000 people who you've condensed down into these small groups based upon all of that context. The more context you have, the smaller you can make those groups and then the more contextual your marketing campaigns can be. And I think this is a big, big, big thing in Terms of the future of marketing, I think this is a big change that will happen. We're doing a bunch of this now in HubSpot, trying to work through the different concepts. And I suspect that when you add up all those 3,000 person campaigns in totality, the conversion rate is going to be way, way greater than just marketing to the 300,000 people from the I set.

[Kip]

Well, I think this is one of the most transformational points in marketing right now, Kieran. And what we've kind of alluded to throughout this show is that the cost of doing marketing is going down. Like the cost to write something, the cost to create a video, the cost to do research, the cost to get data, all of those things are like, kind of flat to Dale. And what's great about that, Kieran, is that you could do more marketing for the same amount of money. And so I think there was this sentiment in the early days of AI that, oh, like we're just going to cut marketing teams and budgets and all of this stuff. It's like, no, we're going to have those people and have that money accomplish much more. Instead of maybe spending $300,000 to do a big brand shoot that gives me a couple of ads, I. I might spend $300,000 to get 3,000 ads that I could do 3,000 micro campaigns with influencers and niche markets and all of those things and get double the revenue out of that.

[Kieran]

Right, Exactly. I think the way this works, again, is you use these intent signals to split that ideal customer profile into micro audiences. So nothing again works unless you can build this intent engine. And then because of that, you have these AI powered agents helping you create these campaigns. If you had a marketing team having to do this and you had to have a marketing team set up all these different campaigns for small groups of people, it doesn't scale because the size of your marketing team would probably outweigh the number of customers you have. But because we're going to infuse AI, so AI is going to be able to craft the email, AI can craft the ad, AI can set up the campaign. And what the human is going to do is orchestrate the AI based around these micro audiences to really provide a ton of value. The tools here are LLMs, Perplexity Labs.

[Kip]

Yes.

[Kieran]

I love this tool. You love this tool. You just did a whole episode on their new browser, Comet, which you have. I don't have.

[Kip]

You're not using Comet yet?

[Kieran]

I can't.

[Kip]

Why?

[Kieran]

Oh, Ken, you told me how to do it. Oh, sorry. I haven't upgraded.

[Kip]

Come on man. I need you to use it real bad. I need your review.

[Kieran]

I use it. I was excited that I did the episode on ChatGPT agent because I got that the same time as everyone else. The first time I felt like a proper human in a civilized world where I got a tool the same time as other people got a tool. I will upgrade to comment actually tonight. I forgot I do have access to that. I need to upgrade. I'm on just the most expensive plan of everything now.

[Unknown]

Yeah. Oh yeah.

[Kip]

My AI cogs are high.

[Unknown]

Yeah.

[Kieran]

I have all of these tools and then you can like agent whatever your agent platform of choice is. We have a platform called Agent AI which is a great platform. You can build some agents.

[Kip]

Yes. Here. Let's go top to bottom the tool. Quick hit for everybody. I know people are going to be like well cool. I'm buying what you're telling me to do. Where do I go to get started doing this? So like give everybody a quick tool summary.

[Kieran]

Here's your tool set that you can go start to use for some of these. So for AI search you want to grab yourself one of these tools that allow you to see visibility within the LLMs. You can get an X funnel, you can get a use bear. There's some others you can go use. You want to get really good at using ChatGPT and Claude and these things to actually fine tune on your product pages and create niche product pages for you around relevant questions. If you don't know how to do that, go look at the episode we give with we did with Asia Master short form content. You want to really get great at VO3. We'll have a whole episode coming at you. When I do my little five minute movie I will demo it on the show and then I would tell you everything I've learned about promptin for VO3. That's exciting. You can use captions mirage and then you can fine tune these LLMs ChatGPT Claude to be a great Ryan partner. We're going to show you how to do that in a future episode. Code powered experiences there's lots of great tools. I love cloud code Lovable replit. We have Agent AI to build the agents and so you can build agents and give them away to acquire demand. Build an intent agent Kip Give the thing you should do. You should go connect your HubSpot connectors, pull the data and then run deep research to figure out what are the best data sources for your specific intent signals. Because there's a wide variety of things you can get AI sales prospect and we are doing this right. So we are doing it via HubSpot. We have our intent data built up here. If you want to run some personalized video as part of your prospecting and try that out, the new hey gen model is incredible. Go do that and then your micro audience campaigns again, you have your intent engine. You may want to do some Perplexity Lab stuff, which I did to like categorize my audiences in cool novel ways. Generally just being able to do great stuff with LLMs like Grok, Claude, ChatGPT, Perplexity, they can actually help you segment and cohort that audience in interesting ways. So these are some of the tools that will allow you to run your AI marketing playbook. And if we were starting from scratch, going into a startup today, going into a company today, these are the things that I would prioritize.

[Kip]

All right, well, I know we covered a lot of ground and we've done a lot of shows on some of these topics. Kieran gave us a really good overview. We've created a playlist with all of the shows relevant to today's episode. We're going to link it down in the comments below if you and so if you just want to go on a full deep AI marketing deep dive, you're just going to want to watch every video on that playlist and we should answer all of your questions that you would have from today's show. All right, everybody, that was a deep dive in some of the latest and greatest in AI marketing tactics. If you're starting to do those, then you are going to be at the leading edge of the AI marketing curve. We will be back with way more shows on these topics and other new ones over the rest of the year. We'll see you real soon. Soon. I'm marking it's the grain.