Drive Impact Through Systems Like an AI-First PMM

0:02Steven Schuler:Hi friends, welcome to Product Marketing Adventures, the only show that takes you beyond theory and into the real execution of product marketing. My name is Elle Grossenbacher, your trusted guide and host of this show. In each episode, I bring in expert product marketers to co-host two segments of this show. First, a case study example of their work, followed by a messaging critique for companies we admire. So if you're a PMM looking for practical guidance and inspiration, you're in the right place. Are you ready? Let's go.

0:42Steven Schuler:Product marketers everywhere are feeling this. Product teams are shipping faster than ever. It has dramatically accelerated how quickly features get built, tested, and released. But while product velocity has fundamentally changed, many PMMs are still operating on the same quarterly rhythms we've relied on for years. And many PMMs feel like they're constantly reacting instead of actually driving strategy. Don't worry, we have a fix for you. With that, today, it is my pleasure to have Steve Shuler on the show. Steve has spent his career at the leading edge of a number of technology trends like mobile marketing, customer data infrastructure, and now how AI powered products and systems affect our everyday use.

1:29Steven Schuler:Today, he helps product marketers rethink how they operate in an AI first world with a focus on building systems that scale knowledge instead of simply producing more content. If you've come across his writing on LinkedIn, you'll know he has a knack for turning complicated AI conversations into practical frameworks that PMMs can actually use. Steve, it's amazing to have you on the show. El, thanks so much. It's wonderful to be here. All right. So before we actually dive into this big topic today, let's bring everyone up to speed. You were recently at a company and you noticed this trend, as many of us have experienced it firsthand, this trend where product marketing teams are struggling to keep up with product because AI lets us move at the speed of light.

2:14Steven Schuler:So tell me more about that moment. Yeah, I think there were really three things that I picked up on. Number one is that product and engineering teams can ship today in a way that they've never been able to before. They're talking about things like weekly release cadences and six-month roadmaps instead of year roadmaps or two-year roadmaps. But PMM teams are still often working on a quarterly rhythm because that's what we've been used to. And that's sort of what we've been drilled to deliver to sales. I think the second symptom of this challenge that I've seen is that the best customer intel is now just trapped in all these disconnected tools.

2:54You might have support tickets from customers in one place. Products got a notion that they maintain that has all their interview transcripts and product information. Sales is in gong. And then there's this like loose information in Slack that's just all over the place. And it's hard to pull it together. And then really the last thing is, you know, marketing teams themselves. We've got this incredible power now to create more content so quickly. But what I've seen is some teams creating sloppy content faster. And that's not the answer. We've all seen these posts on LinkedIn that you know are written by AI.

3:32And it's just such a turnoff. And I think that we can all do a much better job. Yeah.

3:38Steven Schuler:Yeah. Oh, my gosh. So much of what you're saying resonates so much with me because I feel like for product marketers, I see two really big, I don't know, maybe symptoms as you were saying is what I'm trying to say here. But first and foremost, it's the stress of keeping up with the cadence of release. And then second, now we have this large volume of releases, whether they're small feature releases or larger actual product releases. And we are the ones who are responsible for communicating that to our customers. And while AI has accelerated our ability to do it, that doesn't mean that the humans who are receiving the information, our customers can keep up with the volume of content that's coming at them.

4:26Steven Schuler:So the answer isn't produce more content because I don't think our customers can, they're already maxed out. We've been maxed out for years. So we can't just do more. So our jobs as product marketers has actually become a bit more interesting because now, as you pointed out, we have to be a little bit more thoughtful with how we're packaging up all this information. And yes, using AI not to produce the sloppiness, but I think that's kind of... I don't think it's because PMMs want to. I think we've just been in reaction mode. and I'm trying to keep my job. I'm trying to keep product happy. My stakeholders happy.

5:05Steven Schuler:I just want, I just got to put something out there. So I'm going to use AI and just push it. And so I think that's how we got here. So, um, but the good news is, especially with all the chatter that we've seen across various marketing communities, we know it's a problem. We want to get out of it. So how do we do it? Yeah. Uh, I, I totally feel what you're describing. It was, you know, It was like being on your back foot. It was like feeling like you have to be in reaction mode, all of that, right? It's just like, oh my gosh, I have to keep up with all this stuff. So I just like caused one day and I took a breath and I was like, all right, I got to go back to basics here.

5:43Like how do you take on any product marketing challenge? And so I had given some thought to this like a few years ago, actually, even before this AI stuff really exploded. And there was a group called Product School. They're in San Francisco and they do all different kinds of classes for people that want to learn about project management. And they reached out to me and they said, hey, come teach these product managers and emerging entrepreneurs about product marketing. What do they need to know? So I came up with a framework, right? PMMs love a good framework. I was like, I need my own framework.

6:18So I came up with this framework, 3D product marketing, right? Amazing. Brilliant, right? Right. Well, the three D's are data, decision and delivery. And data is all about collecting market information. Right. What are your customers saying? What are your competitors doing? The decisioning is all about, hey, what's the messaging, positioning, go to market strategy? And then the delivery is how does this come to life in a one sheet, on a website, in an email, at an event? But I think the challenge with that for a long time was that we spent so much time on the data collection and the data and the just project delivery that we didn't spend as much time on the decision making.

7:00And that's how I think we go from being on our back foot to being on our front foot and offering more to sales and product. We use the AI for the data collection and the delivery, and we use what we're great at, the decision making, and it can actually spend more time there because we have these tools that could do a lot of the rote work much more quickly.

7:23Steven Schuler:So in that moment when you realized that you have your framework and you're in that moment yourself as a product marketing leader, you've got your stakeholders who are like, move faster, move faster, move faster. How did you actually... I guess I can walk us through your thinking around what you decided to build that's incorporating this 3D framework. That sounds like it's kind of your fundamental philosophy as a product marketer. So talk through what that looks like. Yeah, that's the guiding philosophy. But where AI comes in here and where I think we can really accelerate work is building like an always-on intelligent system.

8:05And the idea is that this is collecting that information. It's synthesizing it. It's alerting you when there's something that you need to know about. And then it's improving too when it gets a signal that there's a trend that's popped up in gone calls and all of the people in your key persona have started saying this one thing that needs to be incorporated back into the persona docs that you have so that the next time you write an email, that new sort of important thing for that audience is incorporated. So it's about wiring all these different things up and creating this loop, right? Still employing the 3Ds, but doing it in a really modern way with AI tools at the center of it.

8:49Steven Schuler:Yeah. Can I just say that I love that you didn't start with writing prompts? Like, I feel like AI has moved so quickly that now, I hate to make this analogy, but it's almost like it's almost like the old school version of like you know how younger generations will poke fun at their elders a little bit for you know not knowing how to do things I feel like now if you're a marketer who's like still just writing prompts it's kind of like where even are you get out of here you can't even save a pdf you know what I mean it's like one of those type of, we're not going to, okay. Boomer this, we're not doing, I didn't want to say it, but yes, exactly.

9:40Steven Schuler:But it's moving so quickly that now it's, um, doesn't matter. Just take your age out of it. It doesn't matter if you're just, if you're behind, you're behind. So, okay. You're starting with the knowledge, which I think is so smart. And actually let me take that back. you're starting with your framework, your philosophy as a marketer. And then with that comes the knowledge component. So once you build the system, like what actually changed for you as you put the system in place and started putting into action? Yeah. Well, I mean, we can talk about how to build it in a second, but the results that we were really looking for was like faster execution, like for PMMs.

10:20So like make our lives easier, better consistency, right? That messaging doc that you spent all that time slaving over, like, does that message come through in the fourth email in the sequence that the BDR sends out? You know, you can help, you can help do that. You get better alignment across teams, right? Yes, maybe you talk to sales, but are the CSNs, are the SEs, you know, all these other stakeholders, are they saying the same thing? You know, do you build stakeholder trust? You know, is product inviting you to more meetings because you have a unique view on the market? And are you doing like less redundant work?

10:57You know, you're not writing the same email over and over again. You're customizing it now, right? You're writing it for this vertical, then the next one, then the next one. And you're doing things that, you know, really add value that we never had the power to do before because you can deliver so much faster.

11:13Steven Schuler:Yes. And I just want to reiterate something that I realized as you were talking, And this relates back to your framework, right? Like you said at the start here that the whole purpose of this was not to do more work, not to produce a larger volume of work, but to spend more time in that decision part of the job. And now you can do that. So it's trading off where you're putting your energy. So I just wanted to call that out. That was a big takeaway for me as you were talking. Yeah, absolutely. And like, I mean, just for me, like, that's the part of the job I love. That's why I started doing this.

11:52That was what people promised me, oh, PMM, it's so strategic. And then I'm on like, you know, you know, after work writing the one sheet, and you're just like, this does not feel strategic. But, but, but this is how I think you get some of that, you get some of that leverage back is you build the system. The system helps you abstract away some of the things that used to take a long time. And now you could be more thoughtful and you're being more well-informed because you have a broader view of the market.

12:24Steven Schuler:Yes, exactly. Okay. So I'm sure there are PMMs listening right now who are thinking, okay, I really want this. So let's turn this into a playbook. We will role play for a second and assume that you are my PMM coach and I want to become an AI first PMM and I'm trying to build up that system. What's step one? Where do I get started on this? Yeah. Again, it doesn't start with writing a prompt. What it actually starts with is what I would call mapping your knowledge foundation. So just like think about your organization and all the places that information about your customers, your competitors and market trends actually lives.

13:05It's gone. It's Slack. It might be Zendesk. Use them for support. I mentioned, you know, product teams working in Notion. It's customer interviews that are hidden in Zoom recordings on some G drive. All of this stuff is out there. It's in your organization. And the first thing I would recommend anyone to do is to actually, you know, think about what they have access to and kind of map it out before taking any next steps. I think that's a really good step to do, even if you're not building an AI model, because it's a step, especially if you're like starting somewhere new.

13:42Steven Schuler:It's a step that I think a lot of marketers forget to take. And then you have resources that you just forget about or don't know that you have access to. So great idea. Got it. Okay. Step one, map all of the potential data sources or places where knowledge could be hiding inside your organization. What's next? Yeah, still not writing prompts. Next thing to do is talk to stakeholders. And, right, I mean, it's business. You got to have the relationship from the folks. And you need to understand what do the product managers need? Where do they feel like they, you know, don't have a view into what competitor X is doing?

14:23What does sales need? Which emerging trend are they following but don't know enough about? The demand gen team and your own marketing organization, right? Do they need to know what's working? What topics are trending right now? You can help them with that. Brand might be thinking about what the persona cares about. And even customer success orgs, they want to know how does the product that we're trying to upsell relate to things that my customer cares about. And I think all those things are things that you can pull from that knowledge base that we were just talking about, but you need to ask and find out from all these different stakeholders what they actually need in that moment.

15:06Steven Schuler:I love this. And I'm going to add up a little personal experience here. That's not necessarily related to AI, but I think it could be. So when I was at Cisco, a ton of stakeholders, so many stakeholders. I couldn't possibly remember all of them. So I had to make, there's so many, I'm sure it's like that at many companies, but I was very overwhelmed with how many stakeholders I had at Cisco. So I literally made a table, like a spreadsheet that had everything that you're saying. It had all my stakeholders. It had the things that they cared about, the things that they wanted to learn, the things that they wanted to know, obviously who they were.

15:44Steven Schuler:It had so much information on it, but I didn't keep it to myself. So for example, like when I met with my product stakeholder or my demand gen stakeholder, I showed them the whole table so that they knew what I was doing with all of my stakeholders. And I felt like that level of transparency, it earned trust. Like they knew what I was, it let each individual stakeholder know, first of all, that I was managing so many stakeholders. So I was given some grace when I needed it. But two, they knew that I was solving for something pretty complex. And then what I could imagine, how I could imagine this being adapted for something like what you're doing is adding in some column related to what we hope to get out of this with AI or whatever, whatever we're building with AI so that we can be a little bit more proactive and feel more like we're doing it as a team, not just I'm delivering something to them.

16:47Steven Schuler:So just a little tip. That's something what I would do. Would you suggest that I do that? Or is there anything that you would add to that type of table? I think that's a wonderful idea, especially in large organizations, but also in smaller ones, because even at a company that It might be, you know, 300 people, 200 people. The number of different departments and teams and sub teams that are asking for PMMs time in one way or another gets big pretty quick. I think we've all experienced that. So, yeah, map it out and then add that column, right? Hey, what do they need as an output of this system?

17:23What do they want to know more about? And then those become your first early wins that build that credibility. So then you get that seat at the table and help put strategic decision making.

17:33Steven Schuler:Yes, I love that. So step one is map the knowledge sources. Step two is map the people, the stakeholders. What is step three? So next step, I would say, is to start building the system. But I started in a very specific spot, right? If we go back to that 3D product marketing, the second thing we're talking about here is like these decisions. And I don't think we always talk about some of the decisions that we make using that term. But it's like, hey, what's our brand voice? Who are our personas and what do they care about? What's our ICP? Which verticals are we playing in and what matters to them?

18:14What are the key trends in their industries? And so all of these become what I call golden docs. And that ends up being the knowledge base that all of your AI answers are pulled from. And if that knowledge base is complete and informed from all this live information that's coming in and from your team and your peers and partners, putting thought into, for example, what that brand voice is and making it really specific to you and your customers. Now you've got the foundation for AI outputs that are highly relevant in your voice and that you can produce quickly and easily.

18:55Steven Schuler:I love that you said that. And then I'm going to share an insight that I have. And I hope this is validating for you too. You are maybe the fourth or fifth of my podcast guest who has talked about building some kind of AI system. And when we start talking about how we build it, this is the first thing they say. They say to have, I've had another guest call it a manifesto. I had another guest call it, she called it her messaging house, but it was basically this, it's everything that you just described. It's the brand guidelines. It's the persona documents. It's the market research. It's the source that you want your AI system to pull from in order to reduce hallucinations and all that good stuff.

19:46Steven Schuler:So I'm seeing that as a trend for product marketers. And I really, I just, it's, and especially noticing it from all-star product marketers like yourself, who you understand how product marketing works. You're an expert in your field. So you're still doing the product marketing work. You're just using AI to do it sharper and not necessarily do it faster, but to trade off where you're spending your efforts. Yeah. And I'm encouraging people that spend time on this step, right? This is where you want to put that thought in. And if you do want to like, you know, look good in front of your technical stakeholders, the term that I've heard used for this is a RAG database.

20:29RAG stands for Retrieval Augmented Generation, which in plain English basically just means when I tell Claude to write a, you know, one sheet, it retrieves the information from this set of knowledge dots that I've created.

20:48Steven Schuler:From the RAG database. From the RAG database. So if you want to tell your stakeholders that you're creating a RAG-powered knowledge base, this is what it is. There you go, PMMs. There you go. That's a good tip. Okay, give me now some examples. Like once you actually, you do the PMM work of building out the persona documents, and I actually have a question on that real quick. when you're building out those core documents, I'm just going to reiterate what I think you said. You talked about persona docs, ICP docs. Are there specific things that you would always include in those golden docs or does it change depending on the company?

21:31I think those are relevant to almost every PMM job I've had. The other couple of things that I would add is certainly the product documentation and messaging. So like your PMM messaging that that you create, the PRD that the product team creates, those should be in there as well. And then I've had a lot of success putting into that knowledge base what good looks like. So like, what's an example of your best performing content? Which was the email that got the most opens? What's the social post that got the most clicks? Put that in there. And then put in a couple examples of what bad looks like too.

22:05It could be your own stuff, could be stuff that you found online, but I find some of the AI tools I use in particular are great at, you know, me giving it something and saying, hey, this sounds terrible. It sounds like it was written by AI. And then it'll be like, you're absolutely right. This was written by AI. And here are the things I'm not going to do, you know, fewer M dashes. And I'm not going to phrase things that way. And it'll come up with rules on its own about how it's supposed to write. And then that, you know, stays in the system forever. And your outputs just get that much better.

22:35Steven Schuler:It's a really good idea. Okay. So quick question. When you're developing out some of those, we're going to come back to the AI workload in a minute, but just speaking to some of those foundational, the golden documents as you were describing, are there any tools that you use to create those golden documents? Are there databases, customer serving tools, market research tools that are kind of like your go-to preferred tools to try to create those? Or are you starting very much from scratch where you're using firsthand data, like customer interviews and that kind of stuff? Yeah. I mean, I always love a good customer interview, but that's where I think that things like the Gong transcripts and the product team, you know, Zoom calls that are on the G drive and the persona doc can start to get fed by those sources.

23:34And what you want to do is get information. And in the most basic form, it could just be like, hey, let's comb through all these things. Let's find every time we have a conversation with a CTO. And what are they talking about? What do they care about? Hey, help me synthesize this. And then that's something that at some point you can wire in and it can update automatically, but you don't have to do that. You can just copy and paste. And it's great of looking across a thousand gone calls and a bunch of disparate things on G drives and then pulling out trends. Read them yourself. Think through whether they make sense.

24:12Pressure test them with your peers, but then put them in that persona doc. And now every time you write an email to a CTO, it's going to draw off that. And it'll just make it resonate with your audience that much more. And hopefully I'll lead to things like, you know, more click throughs and more pipeline eventually. Yeah.

24:31Steven Schuler:I'm going to repeat something you said, and that's to pressure test it yourself, read it for yourself. I feel like that's the decision D part of your framework. Yes. Yes. Like apply your judgment. This is like, don't let the AI take away like the fun part of the job, you know, get this information, apply your own critical thinking, right? Everyone listening to this podcast is super smart, I guarantee you. Listen to your gut, share it with your peers, and then codify it, put it in that persona doc, and then use that as the foundation for all of your launch materials when the time's right. Right, exactly.