Agentic Marketing Rewrites GTM

Machine-generated transcript from the episode video. Speaker labels are not included and some names and product terms may be transcribed phonetically.

[0:00] that really aligns with how people are thinking, how people are feeling. Add on complimentary hardware to software. So I map where clarity exists and where it doesn't. Also wanting to take advantage of this massive opportunity GTM vault where enterprise GTM actually gets built. The content shipped, the campaigns ran, the dashboard looked fine, and revenue still didn't move. Not because the team wasn't working hard, but because no one actually knew what buyers were responding to. In this episode of GTM Vault, Annette Sun explains why most GTM teams don't feel why most GTM teams don't fail from lack of effort or creativity. They fail because they never turn customer truth into execution.

[0:48] Welcome to GTM Vault. Trusted by over 25,000 founders and operators building the future of revenue. My guest is Annette Sun, co-founder and CEO of ALA. Annette has built multiple marketing engines from zero as an early marketing hire, a fractional head of growth, and now a founder building AI native GTM infrastructure. Anal built on one belief. If what buyers actually say at each stage of the deal, GTM stops being guesswork and becomes execution. Today's core question, what happens when GTM teams stop guessing? When does scrappy growth turn into a system problem? So, typically you start getting signal that you're acquiring the right ICP and once you've acquired them that they retain. Retention is usually the earliest signal that you have that you're you're getting close to PMF. And once you see this life cycle take place uh you can design a system that is sustainable and repeatable for products where you know maybe the business life cycle or these are life cycles long for you to jumpst start that one way is to start seeing whether or not people are reaching that first aha moment quickly

[2:01] and able to recommend that product uh maybe even before fully activating. So once you see any signal around retention or interest in recommending the product that means you're on to something. Sometimes it is finding that the messaging market fit is there before PMF. Uh it just means that you're on to the right direction. What separated the engines that compounded from the ones that quietly failed? Engine design. It means that when you're designing the engine, it's really just an operation and it needs to be designed in a way that is sustainable from a resource perspective. For example, some people say, "Hey, like paying this one influencer x amount of dollars really converted this much growth. How repeatable is that for the stage of your business, the type of business you're in uh that you're the type of business that you're running, the market that you're in?" And so building that engine in a way that is repeatable is really important. And also oftentimes you'll find that people build the engine a little too late. Once you have your first opportunity to turn something into a repeatable system and you're trying to scale this out incrementally, you actually need to be thinking about that

[3:15] second repeatable play right away. A growth really needs help to continue compounding. And to get that hockey stick growth, it is the effect of compounding. And so often times, you know, you see one thing that's working, you start investing into that play over and over again. Without layering on that second play fairly quickly, you'll find that the momentum will die off. It's cuz the customers that are excited from that play are probably talking about you. They're probably looking for more information about you. They're probably researching in different areas of the internet or in the world really. Um, and if you're not showing up in that second or third place pretty quickly, you'll see that that excitement might die down or they're not really getting what they're really looking for. And often times these people are just looking for validation. They already know they want to buy your product. The desire is there and they're looking for reasons to validate that desire.

[4:07] Why do most teams still get messaging wrong even with tons of data? A lot of people think of messaging as just copy, but really what messaging is is intent engineering. What I mean by that, and I I brought this up in one of my points earlier, which is really understanding the intent that your the level of intent that your buyer has already because often times you'll find that you're selling a product where people don't even know they need it yet. they might not even know they have this problem or they might know they have this problem but they don't think about the business consequence of not acting on it or that there's actually a solution out there for them. So being very honest in how ripe your market is and how mature their intent is and developing messaging that meets them where they are is really important. When people get this wrong, and they usually do get it wrong by overestimating how much intent there already is, but when people usually get this wrong, you mess up the sequence of information people need to hear before solidifying that purchase intent. And when you jump the gun of telling them, "Hey, come buy this because you have this problem." They're like, "Well, I don't. Why do I care?" you know, how do you nurture that desire, that care, that intent to to get

[5:21] them to finally take that action? It's equivalent to like, you know, you don't marry the first person you see. Uh, right? You have to like realize you want to be with someone first and then you go and date them and then you get to know them and then they're like, "Let's commit." It's the same thing in a buyer's journey. Um, and when you jump the gun and start the wrong place, you just become that annoying marketer who's spamming my inbox and I don't really want to hear from you. Where does intuition help and where does it mislead? Usually with intuition is really helpful when you have experience with an audience and you'll find that a lot of marketers build their careers selling things, tools, software, products within a certain industry. Um, and they become experts in that industry. They understand that buyer psychology really, really well. that does sometimes lend itself to a trap where you walk in with biases um and thinking that you already know that audience and actually in every set of product market fit there are different inputs at play and so that bias can be a trap but a good marketer usually should be able to try to avoid that trap. The second I would say is embodying that buyer psychology. So, usually the marketing person is not the demographic or personality or

[6:34] psychographic of that buyer. So, listening in on calls and being able to embody your buyer's lived experience. So, you can start from first principles and ask yourself, if I was this buyer, what is that exact journey I will most likely go on if I'm experiencing pain point X? Where will I look? Who would I talk to? How will I go and solve this problem? How big of a problem is this amongst the five billion things on my plate? And when you're able to be very honest and objective with that journey mapping, then it becomes a faster ramp into finding that potential repeatable play that will actually move the revenue pip uh revenue needle. And the third one from an where intuition helps is really chasing that signal. Usually what performs is not what you expect. I've done this seven times and every time I have a hypothesis, it usually the thing that works is usually not where I was looking the most. But really the hypothesis is a forcing function for you to get into kinetic rhythm. So you can start generating signals from a cold start and start seeing, hey, where is the market reacting? Because actually when the market's not reacting, that's its own data point as well. And that gives you a

[7:47] signal into like, do I continue pursuing? Maybe I did I I have to try this from a different angle or maybe this is the wrong place to be looking and trying things and and showing up or I'm trying what separate what separates interesting messaging from revenue creating messaging you have to really define what interesting messaging means in the first place. So what is the job to be done uh of every messaging right? So if I were to think of interesting messaging, my guess would probably be maybe brand play. And for a zero to one business, you probably shouldn't start here. You should understand what are the inputs that drives revenue first because without that building a brand is pretty much a moot point. But for businesses that already have a clear market position, brand marketing is actually a very good method in solidifying that positioning and helping you reach a broader market. And in that case, interesting messaging really should be like creative storytelling. How can you show up in ways that your audience did not expect, but it's still on brand and and within the realms of of your brand guidelines and your your business objectives.

[8:53] Where does the breakdown happen between customer insight and GTM execution? We often see that there is an interpretation gap between customer insight and GTM execution. And this is an org design issue that is age-old at this point. Usually they're in a marketing organization where there's more than two people. Um you'll find that there is a dedicated person that goes and collects the insights. They probably put those insights in a form of a deck or document um and have multiple meetings with the team that's actually going to be executing on that and turning that into marketing content, marketing campaigns and different types of assets. and and bridging that gap is huge because it's what I call um the biological human context window. Whenever you have a very lean team of maybe two to three marketers that are getting those insights and executing in a single human body, you'll find that all of the memory and learnings and and audience insights compounds, right? And they're they're executing on it. feedback loop is a lot faster and they're able to optimize that go to market execution in a much more efficient way. Once you start scaling a marketing team and you have PMM feeding

[10:09] insights to content marketers or or or social media uh folks or dedicated sales enablement teams that game of telephone is where this information breaks down and and that source of truth breaks down and it becomes a game of interpretation. Why do insights die in decks instead of shaping decisions? That it's an org design issue that we have designed marketing teams in a way where we delegate these skills very separately. We deconstructed the organization to have one person own insights, one person manage a certain channel um and within that you have multiple people executing within that channel. And decks just happen to be that form in how we share information.

[10:58] But what's problematic with that is that the decks are a summarization of that raw data. And a lot of nuances can get can get lost. A really strong insights team will surface those nuances, but there's something around letting the specialists or the executors directly get access to customer ground truths that might actually change the way they create content and change the way they think about messaging. What does a healthy insight to asset workflow actually look like? the tighter and the more compressed you can get that workflow, the the higher the ROI, you'll probably see it's just a more efficient way of of um iterating on the feedback. Um everything in marketing is a feedback loop. So the more efficient you can create a system that enables this feedback loop with shared context, the better. So in this case, you know, the person or even the tool who's executing on that asset should deeply understand that driving insight and all the nuances that come with it.

[12:00] What's the difference between using AI to produce more versus produce better? Yeah, it's an interesting one because AI to produce more is definitely a big theme of like 2022 to maybe early 2024. And we very quickly learned that actually doing more causes more damage than than driving ROI because marketing is it marketing is not about contributing to the noise that's already in the market. It's about how do you create high high value signal to get someone's attention and guide them to that next action. Um and a lot of companies very quickly learn that by just putting a bunch of AI slop into the market. It's it just hurt their brand and lowered the quality of their buyers perception. Absolutely. A lot of AI slop out there nowadays. Um, outside of AI slop though, where else do you see GTN teams misusing AI today?

[12:51] You saw a lot of teams creating generic content and we very quickly learned and diverted from that. We see a lot of teams running generic outbound and I'm also starting to see that go down. So, the nature of the AI game right now is very experimental. People don't really know what works until they try it. And the issue with a lot of these AI tools that were built in the first era and the early second era of you know AI native tools is they really focus on automation rather than being thoughtful around how does the tool and the human work together and having a lot of intention around where the human intervenes or how does that human in the loop look. the difference in AI taking away our jobs in a very bad like a poor lowquality way versus AI will actually make me so much better at my job and actually possibly put me in the spotlight because we were able to do more in a in a very high quality, very thoughtful, very effective way depending on whatever your marketing play is. So not a lot of it is actually fundamentally on a tool design level. We saw a lot of AI tools designed for the last mile is what I call it in the past.

[14:01] Uh so for example tools that the human comes to the platform and I have I give it instructions and I tell it what I'm seeing from the real world and the AI goes and interprets that and creates the final piece of content or the final messaging or the final outbound play. And we saw that end up in a very sloppy way because there's a very big delta on how prompt engineering works and and prompt engineering skills. And then also humans actually have a very short context window on on all the signals that we need in order for the LLM to to generate a high quality piece of asset. So I think that more and more we'll see this flip. And what I mean is that using AI tools to do that first mile, all that research, insight, insight generation, helping humans crawl through large data sets to understand what the truth actually is and and then having the human enter that process and and evaluate the results from a lot of, you know, this research and and and analysis and then working with the AI to generate the assets and and move down that move

[15:12] down. How should founders rethink org design as AI absorbs research and execution? Letting humans do high value work again. Marketing has become such a routine job and the number of channels we have to keep up with is ever increasing. The pace in which these channels are evolving is ever increasing and a lot of most of marketing has become the hamster in the wheel designed uh the hamster in the wheel where our job is to maintain these things and keep it going because the algorithms move too fast. Where AI can be exceptionally helpful is helping us do a lot of this routine work in a very high quality intelligent way and freeing up humans to do that higher value work that we pretty much never have bandwidth for anymore. Investing in big marketing bets, building relationships, being creative, adding that human touch, that craft at the end of maybe an AI generated asset that makes it unmistakably your brand. um and and and honestly building brand values that are fun and feel human or innovative or or unique in some way that differentiates your marketing. Um and the our plates have become so full that

[16:24] there's very little bandwidth to you know even invest in any of this anymore. We're going to see an inflection point where brand marketing is going to come back and it's going to be a very fun time. which roles become more important in an AI native GTM or definitely think that organizations will collapse and uh and and be more streamlined. So traditionally as businesses grow today, marketing organizations subfunctions each become many teams, right? Maybe they're organized by channel or organized by a stage of the funnel and and there are often times silos between each of these sub functions and we have multiple meetings and manager layers to ensure that there is consistency and alignment across all of the different work streams. we will see more single owners with different tools that will help them own each of these sub functions end to end. I don't really know if these owners will lean towards being specialists or generalists just yet. There's an argument for both sides. So on the specialist front, you know, perhaps they are uniquely talented creatives who are proficient in data and at data orchestration and systems thinking or on the generalist front maybe they it's more of a marketing ops role that happens to have a talent with mentoring junior members. This sounds

[17:39] like a very unique skill, mentoring, mentorship, but where this comes in with AI tools is that if you have the ability to give very good feedback and and and mentor junior members, you can probably train or set up LL LLM driven systems to create good outputs. Yeah, that's a good point. What are your thoughts on the convergence of the revenue generating departments um and postAI world as efficiency and productivity increases and headcount reduction becomes more of the norm across the board especially in enterprise GTM will there be more overlap amongst the revenue generating teams 100%. I know it's a hot take, but I actually think that all teams within go to market should be revenue generating.

[18:31] It's just that our our team sizes have grown so much that we each own different metrics, MQLs, SQLs, it's all become a math game. But really, you'll find that when you organize your teams that way, the engines that you build become a little less efficient. And when you can have multiple functions contributing to a single north star, that is probably when you will see the highest ROI. But you know that becomes a slightly different problem. We were double click on it. One is um attribution tooling um or the efficacy of attribution models. And then the second is is managing humaning teams. How do you divide and conquer uh while also collaborating on certain work streams or functions? But yeah, it's a complicated thing to deconstruct.

[19:22] What separates GTM systems that compound from those that plateau? In my experience, it is usually systems design and the timing in which you scale up these systems. So the job of a first marketer is two things. One is like finding that scrap being scrappy and finding that growth hack that actually generated some return, right? like growth has been is growing. And the second is how smart and well how smart can you get about designing that engine with very little resources in a way that's repeatable. And so for early stage companies often times that's figuring out, okay, something's working. We need to scale this. How can we hire or allocate resources to that some that thing that's working and and have that resource be fit into this system that we're we're designing this this repeatable system that we're designing so that the person who the growth marketer who figured this out can go on and what we call crack that second channel or crack that second play that that continues layering on this engine. The second thing if we were to narrow in on early stage companies

[20:33] and and perhaps this is pertinent with later stage companies as well but it's how the founder collaborates with marketing. A lot of founders you know can cause directional whiplash. Uh it's very easy for a founder to see another cool company doing a cool thing and have FOMO but every company has its own set of inputs that is what drives the PMF. Um, and so it's it's every decision you make has to be contextualized into your specific scenario. Or sometimes the founder maybe doesn't trust the marketing person or team enough to just give them that space and safe environment to to cook as they say. Giving them the space to experiment, giving that space to fail because actually failure is a data point. Uh, you need to know what doesn't work and the more the faster you fail, the faster you'll succeed. But rather than, you know, pointing fingers at the founder, I would argue that it's also marketing's job to educate the founder and and manage that relationship, manage those expectations and and actually create the environment that you need, get the resources that you need to drive that result.

[21:40] Where do founders usually double down too late? I touched on this earlier, but I really think it is around resource allocation and specifically for zero companies, that usually means hiring. At a later stage company, when you're launching multiple product lines, it's easier to at least allocate existing resources and and restaff them uh to to a new type of work or a new function. But this is a bigger issue with 0 to one stage companies because for 01 stage companies, hiring takes a very very long time. Um and and the tricky thing is you don't really know who to hire for until what your engine design looks like. Um once you see growth takes off, you need to help it compound. At the same time, with that really lean strap team that you have, you have to now dedicate someone to full-time, find this unique, you know, unicorn to come and join, convince them to join your very early stage company that is just finding PMF um and and works in a very specific way.

[22:35] And so finding that entrepreneurial talent who has the skill for that channel is usually a journey of its own. And and yeah, I I a lot of founders don't that I've come across like they recognize that this is a long journey in recruiting and and they're they're usually having these conversations early on or preempting this need very early on just to keep that candidate top of funnel going. What what is the hardest GTM lesson you've had to relearn? That you never really know until you do it. Every set of product market fit has its own dynamics and even postproduct market fit this this journey continues because the markets the market evolves, humans evolve, the set of components at play are always changing and so never have assumptions. And I've built growth engines for seven companies now. And every time when growth starts taking off, it's never really what I expected.

[23:34] Um, the hypothesis is just one way to force you to get started. What does AI native GTM actually mean in practice? This is an interesting question. I think this is TBD because of how early we actually are in adopting AI tools that actually work. We have seen a lot of singleplayer tools. Um so we're seeing really good adoption there obviously uh you know chachi bt cloud notebook lm whisper flow very very popular tools that are enabling individual level productivity what we're going to see very soon are multiplayer tools and and I imagine this there were going to be horizontal multiplayer tools very verticalized multiplayer tools and that will fundamentally change how organizations operate and and the results that they can expect based on this type of resource architecture.

[24:25] How should founders think about differentiation when everyone has AI? Focusing on the problem you're solving. I think that the core marketing fundamentals don't go [clears throat] away just cuz you're an AI native company. The same types of practices and playbooks are still at play. I don't want to say playbooks because I just said earlier that you should never walk in with a playbook. Your job is to always rewrite the playbook. But but yeah, you know, uh focusing on the core fundamentals. uh the different I will say the one difference that we have with AI now is that we have new problems to solve and and new solutions that we can we can architect. So that's that's one thing but from a marketing perspective things are the same and the things that influence how we market are external to AI is just what are the channels that we can play on who is your audience and how do you show up there those are the three core fundamentals what says human no matter how good the tools get this is a fun question for marketing because it comes down to craft taste creativity and and building relationships with your audience brand marketing was the prim primary marketing play. I want to say around the 2010 2015 era, especially as we started getting very visually driven platforms like Instagram and Pinterest come onto the

[25:39] market. Um, and as programmatic advertising started taking over, we became more and more data driven. Um, and and marketers, you know, started having to hone in on those skills. I think that this new era will be around doing both. So how can we leverage humans in areas where we are unmistakably talented in things that make us human ideas, creativity, taste and then also how can the human be smart in how we orchestrate data the data layer that the agentic tools operate on. When does better tooling stop helping and better systems become mandatory? Better systems have always been mandatory. A lot of the first wave of AI native solutions where they focus on last mile execution didn't live up to their expectations because the system was not designed to enable a really clean curated set of data as the primary input and and you'll find that in leveraging AI tools all it is is the better the system and the the more the user understands the system the

[26:51] better the output. put uh because the user understands how to complement the system. I'd like to move on now to the rapid fire section of the pod. In one sentence, first instinct and that one GTM myth, teams must delete. Um maybe not a GTM myth, but generally like FOMO, forget about what other people are doing. Focus on what you have going on. One buyer signal that matters more than metrics. Word of mouth. One AI habit GTM teams should adopt now. data skills. One GTM metric that predicts revenue retention. One expensive founder mistake you see too often. Copying another company's playbook. One word for the future of GTM. Ooh, there's so many words and I I I'm trying to think of one that encapsulates it all and feel like my maybe need to be more literate in a very nuanced word, but the immediate thing that comes to me is like fun. And if you're going to be entrep or maybe entrepreneurial, if you're going to be entrepreneurial, it's going to be really, really fun.

[27:50] For the last decade, GTM rewarded volume, more content, more channels, more spend, the next decade rewards signal. Teams that win won't guess louder, they'll listen better, and they'll build systems that turn customer truth into execution. And that, thanks for showing us what GTM looks like when insight becomes infrastructure. This is GTM vault. Build systems, not noise. Thank you so much, Rick. The GTM operating system for teams building repeatable revenue.