LaunchPod - Karen Chao
===

Karen: ~I'm seeing there's a trend of a convergence of different roles. EPTO is a more common one. It's very typical as you get more senior in your career that there's opportunities to learn different areas and marketing just seemed that there was an opportunity to bring product and marketing closer. The product person knows a customer, just like the marketing person needs to know the customer.~

~For me, I always thinking about how I can be a. Better product leader and taking over marketing and learning the marketing function is only gonna make me a better product leader. It's not so different than my consumer experience as a product leader because you're very much thinking about the deep data and the conversion funnels and really understand what's gonna drive, what the conversion of where you want to go.~

~Welcome to Launch Pod, the show from Log Rocket, where we sit down with top product and digital leaders. Today we're joined by Karen Chow. Chief Product Officer at Flow Space, an e-commerce logistics platform where she's also taken on the role as head of marketing. Previously, Karen held product leadership roles at Apple Replicon in IT and more.~

~In this episode, Karen shares how she ended up running marketing on top of product. And how bringing these two functions under one leader has improved go-to market for flow space Overall, the biggest surprises she's uncovered running marketing as a product leader from Chaotic tool stacks to the next wave of AI powered go-to market automation, and the obligatory information about how Flow space's product team uses AI tools like Cursor and Claude to accelerate discovery and prototyping and even ship small bug fixes straight to production without engineers involved.~

~So here's our episode with Karen Chow.~[00:00:00] 

Jeff: Alright. Hey Karen. Thanks for coming to the show. How you doing?

Karen: I'm doing great. Thanks for having me.

Jeff: you've been a bit all over the place, right? , So your CPO and running marketing at flow space. Right now, you run product across multiple startups, multiple verticals. You worked in product at Apple Engineering degrees from Stanford and uc, Berkeley we're gonna talk about today.

How. Flow space. Your team is, you know, adopting AI at a just rapid clip in, in all sorts of neat things they're doing, but as well as what it looks like when A CPO takes on marketing. So that's be a fun conversation as someone whose background is in marketing. But maybe you wanna just give us the, you know, 92nd TLDR on, who is Karen?

How'd you get here? And, and, you know, hit the high notes for us.

Karen: Yeah, . Thank you. So I've been a product leader for over 20 years now. I kind of fell into product. Back in the day there wasn't. Such thing as product management. So when I actually got outta college, my first job was in the supply chain industry and I was a business analyst and slowly started looking around and thinking, oh my gosh, this isn't the right role for me.

a quarter life crisis here. And designed and, you know, started looking at it to [00:01:00] see what interesting roles there were. And, product management kind of like stood out to me. And so I transitioned to product management and I've doing it ever since mainly at smaller companies. I started off in like kind of data, then moved into the mobile consumer space then transitioned into more B2B SaaS and more recently have got really fell in.

To more of an operational businesses. So flow space, we're doing fulfillment. We have a very different model, so, we manage it all through our technology as well as our people. , My role is really instrumental in terms of making sure our performance is up and we're able to manage outcomes for our customers.

Jeff: sense. So we are, I assume, catching you right before a very busy season hits the

Karen: Oh yes. Hundred percent. Peak seasons coming up. So a of the, recent conversations and plans within the organization is all about getting ready for peak and really make sure we can hit those volumes.

Jeff: Well, I'm glad we were able to sneak in before you know, before what Black Friday , and Cyber Monday hit, and Yeah, you become completely unavailable for, for a couple months there. , I guess let's just jump right into it, right? One of the things that led to us talking [00:02:00] originally was, it seems like across the board as I talked to product leaders, everyone is at some level either having their board or their leadership team saying, what's your ai.

Philosophy or what's your AI plan? Or they are, diving headlong into it and they have the thing that they're proselytizing around ai and I'd say flow space is kind of in the latter camp where you all are, are quite advanced. So I'd love to just jump into, maybe give people like the set of like, what, what is product.

Using AI force. This is not in the product but this is how you folks are using it. And we can get into like, how'd y'all get there?

Karen: Yeah. Yeah. Great question because I go to a lot of these product events, I'm talking to other product leaders and everyone's always talking about, okay, how are you using ai? And there's like a dearth of tools and it's hard to, through the noise of what really works and what doesn't work. And at the end of the day, I think it's just like kind of tinkering with things.

Because until you try it, you won't know if it really performs, like they sell you this dream, but in reality it may or may not work for your use case. So. I'll say building our AI into our tool did give us a really great background on thinking about how AI [00:03:00] works. Just, you know, writing prompts in more detail to like, versus using it as a, a lay person, you know, asking ad hoc questions about how to write, write this email better.

If you're writing true prompts to create an outcome for your product, you just get really good at knowing the nuance of how to use ai and specifically tools that we're using. And there's probably a ton of tools that are still an opportunity for us to use. But more recently we started using tools that allow us to just improve our product discovery and development process.

And a few things that are interesting is, obviously we talked to chat GPT about. This PRD, we might do some deep research on the industry or how other people are doing, you know, just more competitive analysis and adjacent industries of like how we might design a feature. So we'll kind of do that on the business side, but then on more of the technical side, we're using Cursor and Claude and it hooked up to our code base.

So we have a GitHub integration that's native, and we're starting to use it to start asking questions that we might ask our tech [00:04:00] lead. Right? questions like, Hey, I'm thinking about this functionality. Here's the context. Here's like the features, here's how we're thinking about how it would work.

Kind of poke holes through this. What scenarios or corner cases am I of thinking about? It's just really giving us a good first pass on how to figure out and really bulletproof our product requirements documents so that we can have an even more effective conversation with our engineering lead later.

Right? We kind of get through kind of the, the, except the early stage and we kind of go deeper with the, with the tech lead. But you know, it's not about replacing engineers, it's more like being well, better prepared.

Jeff: Yeah. It's interesting because kinda talk about that, right? We, similarly, we use GitHub and, I think a lot of our engineers use Cursor, but we're at the point now where we're launching a new product. By the time this one publishes, we probably already launched it.

Where we have a, you know, new free kinda AI feedback analytics tool that we're offering. And in the past we had to, you know, work with an engineer to, to design the page and, you know, build it and all that kinda stuff. And we were able to, I was just able to just go in myself and, and use Claude and there's animation and all this of stuff that we didn't even have to do that we need, you know, still [00:05:00] have to have a code review and the engineer do the final steps.

But like you said, it kinda gets you a lot farther, a lot faster. That's clearly a lot easier than building the product itself. But the conversation happens a lot faster when you have that further along starting point to talk to engineers from.

Karen: Yeah. Yeah, a hundred percent. And even things that we might ask engineers, we don't need to, because we'll, a ask Claude 

or cursor. Oh, when did this feature launch? Or what are all the conditions? Or, you know, like little nuanced questions. And the other thing that's really exciting that, you know, when I talk to other product leaders, this isn't something that their team is doing yet necessarily, but.

Some folks, my team have gotten really into these coding agents and are actually starting to like, push out codes, code fixes. So small bug fixes, for example, you know, the data exists, but it might be missing in a certain page. We're able to make those changes using these agents and pushing those out.

And obviously it goes through the typical development cycle where it has to pass the automate test and someone has to review it, put and, you know, merge it to master. But those are things that we can do. instead of sending it to the engineer. 'cause obviously, as you know, like everyone's got a [00:06:00] backlog full of stuff and engineers are working on projects and so sometimes these small fixes just are really, really hard to sneak in.

And so my team can start picking those up as well.

Jeff: it seems like it's, it's just ticky tack little stuff. It's like if you could assign half a point, it, it'd be half a point or less often, but you still wanna get to someone good who's gonna do it quickly and, and make sure it's done right and go pay attention to it.

But they don't wanna do it. Like it's boring, it's not interesting. , But also . For more pragmatic state. If they're working, like you said on a big project, it's gonna take them out of flow state of that thing or the focus, like everything has switching costs and to switch this little half point thing, now they have to like read it back in the flow and it's just I love that, that you all are, are kinda doing that where as you find stuff, if it's a little things you can just ship to prod and like you said, yeah, it's gonna go through code review as it as it should and then, you know, merge the main, but It's not about saving the half point, 'cause the half point is probably a little meaningless. , It's all the friction around it. And also knowing that you're not waiting like two sprints to fix this thing that, you know, probably isn't incredibly impactful to your users. But it's definitely, if you look at a lot of those over time, that's a [00:07:00] lot of, impact to user experience.

Karen: I agree with you. Those little friction points are where it's like. How can you solve for those quicker is a true opportunity. I'll also say something that's interesting is there was another project and see this is where like everyone's using it slightly differently.

And I think one of the things that I think is good for, or people to do from an organizational perspective is just really knowledge share and how people are using it so that they can kind of adopt the best practice. But on another project, the way they started writing the stories were very, very different.

Almost thinking in context of. Working with the agent to write the stories, to like have the agent understand the stories and so , the way we wrote the stories was different and that was actually helping speed up the project as well.

Jeff: So you, you were working with the agent to write , the story for the actual kind of like original dev ticket or dev feature or PRD and then, but using the same agent to then build so it had context 

Karen: Yeah, 

so it, so if you think about well one, it's like knowing what's gonna work better in AI and what's not gonna work better in ai. So there's some thought around, okay, this kind of [00:08:00] story might work better in ai. And then really starting off writing the story in a way where you're just gonna give it to the agent to do.

Jeff: Interesting. So is that your product team using that now to, to build, I guess, prototype or is that like finished

Karen: that's, that's working with the engineers. Like the engineers are, you know, that, that's just in our normal development cycle where, the engineers are starting to think about how to use the agents as well. And then obviously there's some more complex, deep in the code types of functionality that.

We don't feel comfortable using the agent, but then there's some other types of stories that you're like, okay, let's, let's give this a go. And you know, it's like about learning about how to write the story and how to like, prompt it in a way where you get a good output. obviously one shot is not, doesn't typically work.

You might have to have multiple iterations.

Jeff: Right. But it is interesting, even like if you look at the, , foundational models some of them I feel like do better with. You can give it a multi-step, just one shot, but lay it out, like you said, as long as you lay it out well, like step one, step two, step three, step four, step five, it'll sequentially go through, but then.

Some of the other ones might be [00:09:00] really good at writing or something, but if you give it a five step thing, it'll do step one and then maybe step two and then kinda drift and

Karen: the thing about the attention span and also their over eagerness to write too much code. So actually something I've like, I've tried writing some functionality as well, and what I've learned from my own tinkering is that it's really good to just ask it for steps first. I'll understand, okay, how does it currently work?

And I'll kinda understand the, the code and how it's written. And then I'll ask it to write me a pseudo code based on the context of I'm trying to develop. And then it'll gimme some context on that and I'll kind. Give it direction or not, because it'll make assumptions and you might have missed something.

And so then you have to like go correct some of those assumptions and then you ask it to write the code. But you have to first start with a framework of like, understand what's happening right now and then ask it to create a plan and then you can review that plan and then execute it

Jeff: we use Claude on, you know, our, our application for like a coding agent. But before that I was using lovable to build some products we just use internally on the marketing team for like. Things that no one's gonna make.

'cause use [00:10:00] case is so narrowly defined, but we have three people here who use it all the time now. The first one I made is utter trash. I look at it and it took me so long to make just a piece of hot garbage that barely functions. But as I got better about kind of understanding, like how do you prompt it, working, you know, through chat GPT to build the set of prompts first and, and finding all those corner cases first.

Pres answering questions or, or being clear to like, give it space to ask questions. You know, it takes a quarter of the time to build something five times better now. But it really is like, you know, it's, it's learning almost a slightly, you know, maybe not a different language, but a different vernacular of the same language

Karen: hundred percent. 

Jeff: On the bug fixing and kind of, minor change side, have you found that helps with how you all are working with engineering better? Or like, has that helped kind of team cohesion or, or kind of, focus of the teams across everything?

Are you pleased with how that's going?

Karen: Yeah, I am. I think there's like just we call 'em quality of life issues for our customers. Where you, like you mentioned like you would never prioritize this. It's hard to get in the backlog. Engineers don't wanna focus on that 'cause they're focusing on their core projects. We're holding 'em accountable to these sprint goals [00:11:00] for these projects.

We have deadlines for those. And it's just like, it's just a little bit of a distraction and, and sometimes it's like, yeah, it's just hard to justify and so. Those are the opportunities where my team goes in and tries to tinker and we submit our own PRS for those. And they're a little, you know, I don't wanna like overstate how much we're doing this 'cause it's a little bit, few and far between.

I mean, I think there is a balance because I don't want my team to turn into like, many developers, it's an opportunity for us to learn, but most of their focus should be talking to customers, you know. You know, building out new functionality and really doing that dis upfront discovery versus like doing bug fixes.

But I think it's a, it's still a great learning opportunity and I want people to be confident in using these tools.

Jeff: No, I, I think it's a good area. , We've done a lot of research and validation on this piece specifically because, we started doing session replay for, you know, kinda engineering context. One of the things we found that we have several customers on, like an alpha of now, is. This, like, it's basically we will pull the, you know, we can see the minor issues, find the replay and capture that and kind of understand what's impacting users and see code [00:12:00] complexity. And when it is one of those kinda like minor knits, now we have the agent integrated into, you know, you guys a GitHub right?

Integrate into your GitHub. And it can ship those kind of little ticky tack fixes. But it's the same thing you described. It goes to code review and engineers, they gonna review it, but we've found the people using it, they're, they're happy that that stuff isn't just in the backlog and, and they can get it done.

But it's also like culturally has been more helpful than I, I think I realized it was going to be where they don't have to go ask engineers to, to do these things now. It, it, they can focus on talking to, like you said, customers or they can talk to engineers about bigger things. So it's, it's a cool area.

I think we're gonna see a lot of innovation in, of like this idea of self-healing software. Right. And you're still gonna need people for the big Yeah, right. It it, but that's, I mean, that, that, that was the future promise to us a long time ago with AI was this kind of stuff. But it, it leaves, product people open to do a lot more interesting stuff.

But you don't wanna be coders right, the same way. Engineers don't wanna be cleaning up these little BS things. So like, I think it's an interesting future. It's really cool to see kinda how you, how you all have already picked that up and been running with that. Nice. I did the math last night on this. I, I think [00:13:00] over the past year or so, I've talked about 500 product leaders between dinners we've done and the podcast and everything. Very few people are doing that at this point or are doing it. I think effectively at scale. So that's, that's really cool to see.

One other thing if we, if we can jump real quick. The other thing I wanted to talk to you about that's near and dear to my heart, selfishly, is. Something that I've seen a lot across. You know, we've seen a lot of people where CPOs are jumping in to kinda the CPTO role, right? Being Chief Proc and Technology officer.

You, you kind of ended up going the other way over over at Flow Space and I, I dunno what the technical title is, but you're CPO and running go to market or marketing. , How'd that happen? ' I do think that this is something we're gonna see more and more is, is not just CPTO, but people from product also taking over more go to market pieces

Karen: yeah. For me I've always, you know, I, I'm seeing there's a trend of like a convergence of different roles. Like you mentioned, CPTO is a more common one, but I think it's very typical as you get more senior in your career, that there's opportunities to learn different areas and kind of bring kinda a, you know, a different lens on them.

And marketing just [00:14:00] seemed that there was an opportunity to bring. Product and marketing closer together. So, you know, the product person knows a customer, just like the marketing person needs to know the customer. I obviously know the product really well. I can help represent what the product does.

And also for me I always thinking about how I can be a better product leader and taking over marketing and learning the marketing function. Is only gonna be making me a better product leader even if I don't do it ever again. Right. So,. it's actually kind of funny 'cause as I'm digging in and learning, it's not so different than my consumer experience as a product leader because you're very much thinking about kind of the deep data and the conversion funnels and kind of just like really understand what's gonna .

Drive the conversion of where you wanna go. So, obviously the way you go about it can be very different. The different channels there's obviously a lot of things that I'm not an expert on, but, you know, I think as a leader it's about hiring and having the right team that can go execute on , your overall strategy.

Jeff: I've kinda [00:15:00] always joked to people, like, marketing's not hard , at the really, I mean,

Karen: Me this, they're like, it's not hard. I'm like, oh, well right now I'm learning a bunch of new things, but 

Jeff: But, But, I think, I, I, I think to say it is, is a gross over oversimplification. The same way, like, product isn't hard, right? All you wanna do is build the product that people want, right? Just solve a 

Karen: Just, 

Jeff: hard is 

Karen: just do that.

Jeff: It's the same way. Marketing, you know, just find out who you want to know what you guys do and, 

Karen: Just some 

Jeff: the. 

Karen: on these channels.

Jeff: Well, that's where the Right, I think at Core, right, to almost is probably, I, I think it would be easier to step in and run marketing from kind of a, a product side role or something like that than it would be to come up through it and really, 'cause then you have to like learn all the, you're probably not in Google AdWords or something like that, you know, managing bids.

Karen: That's a 

Jeff: That's a, that's a specialty. Like, that's a skill like engineering though. Like you can't, you know, you can't just like intuitively engineer. But product and marketing have elements where you can, you know, talk to users and, and there is a bit of taste involved more so.

And I, I think that's what I mean, like the goals are pretty straightforward. Whereas engineering is like, you have to know how to write code in Flutter if you wanna build a mobile app in [00:16:00] Flutter. Obviously there's other ways to do it too, but 

like, 

so that's always been the joke, like obviously the, the success failure lies in the details and the execution.

That, that can be, I, I, I think difficult is the wrong word, but like, you have to, you have to have the right strategy, you have to execute it really well 

Karen: I think, 

Jeff: just kinda come in and do it.

Karen: yeah, I don't for sure. And I think I think the reason why they asked me is one, I know the customer I know the space, you

Jeff: Which is a great point, like you have such a deep knowledge of the customer in the space. You're already halfway there. 

Karen: Yeah. And then I, I think marketing is very multifaceted. There's obviously , lots of different channels, lots of different strategies, but at the same time something I feel really , passionate about is like as a product leader and now a marketing leader.

Like there's no one playbook. The playbook's always gonna be changing for any organization, especially now, especially with like ai, and we're starting to see more AI referrals, for example. There can be marketers that are maybe in less technical, less aware of ai. And I think I think.

Every person in their functional and every function needs to be [00:17:00] thinking about how AI can help them be more effective going forward in this new way of working. And I've definitely been thinking about that. Not like ai, not just for a product perspective, but also from a marketing perspective because we're being asked to do, you know, more with less.

So I just think about there's all these AI go to market tools and, you know, some, I'm looking at some of the existing tools and I'm thinking. Why can't they do these things?

And so I'm looking at, at add-on tools to, you know, get, get us better at like some of this automation because , I think there's just gonna be a different way of working on the marketing side especially, that just needs to be thought through. And I, I think as a product leader, that's an area that I can really help chip in for the company.

Jeff: Yeah, I mean, I, I think it makes sense, like, . I've kind of talked about in the show before, I run marketing here at Log Rocket, but I've worked closely with our product team for, you know, the entire seven years I've been here. I'm fairly involved there in, in product discovery, because like you said, it's the same basic research we do on marketing applies to product, you know, what should we build, what products should we solve?

What, what are the problems people have? How do they think about those problems? I would [00:18:00] never. Even think of going and having input on engineering beyond, Hey, do you think you can just do this faster, please. Because I don't have that skillset. So was that kinda the, the thought process behind it was bringing those two groups closer together because naturally it synergy?

Or

Karen: yeah, I think it was a little bit due more with less, but also making sure there was like more alignment between product marketing. I think you know, sometimes you know. Not based on any sort of ill intent, like people end up working silos just because there's like too much to do and too much to focus on.

So, this way we're more connected. I have oversight over both. And can really just help drive the team for just better outcomes. I also say like, my personal working style is especially working in, somewhat smaller organizations. Being more of a startup leader is just getting in the weeds.

Sometimes you haven't figured it all out yet. Not like you're running a massive machine. You're still figuring out what are the right channels. You're lighting up new channels, you're trying to look at the data. You're trying to make investment decisions. In that case, it's very similar to product.

It's [00:19:00] like, what do you focus on now? , What's delivering? And once you launch something, monitoring that and seeing how you can continuously improve or tweak it based on the data

Jeff: outputs tend to mirror your inputs, which in this case is your org structure and your people. And, and so to your point, , there's no ill intent to people being siloed. It's just when you build teams that way, , it just accidentally happens even to people's best ability.

Whereas if you kind of build in this natural overlapping org structure you are forcing. That communication to happen, and it's going to be more aligned by default. You would have to try not to have it aligned versus the other way. You have to try to have it aligned. So I, I think it's a really interesting way of looking at it.

Like I, I think it's a really smart move. Here, like I said I just don't let the product team escape me from giving feedback and, and, you know, our head of product is also our founder, so there, like he has oversight over marketing as well. Via me. So we naturally have that. But I think it's an area where you're gonna see more and more of this kinda overlap.

And then you said like all the other things you said just add a lot to it, right? You're in the weeds on it the same way, you know, how do you know what's gonna work in product when you're doing something new?

You have to get down and dirty. Ai, like right now you're building AI on both sides. Why would you have two teams independently building knowledge there? Are you [00:20:00] seeing the same kind of AI pick up on, on the marketing team side or how separate are those teams now?

Karen: Well, they are separate folks just because of, you know, Yeah, 

and focus,

uh, people want to do.

yeah. Yeah. Exactly. So we're definitely further along adopting AI and the product perspective. We've just started sooner. We developed our AI product now over two years ago. Right when, you know, open AI launched chat, bt Right.

We started tinkering on the marketing side. I'm really excited about lead qualifications scoring that we can do. That isn't like the lead qualification scoring from like five years ago.

It's like really, really, really agentic. So there's like opportunity there. I mentioned how we're lighting up these new channels. We have some really interesting tools we're using that can make our marketing more effective. So, there's a, a lot to dig

Jeff: Yeah. 

Yeah. I, I, I gotta ask, 'cause you know, I'm always curious when people are kind of getting into marketing on this, especially when they're coming with, with just so much, you know, experience in, in business context. What, I guess as you've dug into things, , what is the thing you've found the kinda most unexpected that you've found?

Or like what's the most surprising thing you've run [00:21:00] into as you've kind of taken on this new set of roles?

Karen: Oh, oh, most surprising. Let's see, that's interesting. I think I, I dunno if this is surprising, but this is definitely something where I think it's an opportunity , I feel like. organization now has like a really deep tool stack, and I think the marketing tool stack is very complex and there's so many different tools and none of them.

I mean, they're connected, but they're not connected in the best way. It just, there's nothing perfect. And I get why there's these agencies that kind of help you do all these things and they even build their own bespoke tools for you because it's really hard. The reporting side to have everything holistic and really join up and all that stuff is, can be very difficult.

So. always heard about that, but I'm feeling it firsthand as I'm digging into the data of like understanding what's performing, what isn't performing and, and why and the details on that.

Jeff: it's funny 'cause I just kinda take it for a fact that there's just, there's just a lot of like SaaS tools you use in, in marketing and talking to product people, one of the big [00:22:00] feedbacks we've seen and heard is, there's just so many different places to go for user feedback, data and this, that, like, that.

There's just so many, like you said, disparate, tweaking, unconnected, on the pro side. And then I think marketing is like the one org where , we put product to shame on the number of tools sometimes. Like there's just so many things you can have and like you said, they don't talk. Even the ones they're supposed to talk to each other, they don't talk to each other that well.

And this is one of the areas I'm really curious to see over the next, you know, five years, what happens here because. Like, we've seen this push of every company in the world is pushing to have, you know, all the adjacent areas of, of product around their main product as well as, the speed of building has sped up.

I feel like this has just gotta collapse a little bit on product and the marketing side of just less tools and, and everything. But separately, I've always been like. I guess as someone new to marketing, I'm, I'm curious to get your thought on this. For a long time I was the marketer who, who had every tool in the book and it was like, oh, hey, this, if we need to solve this, let's go buy the platform that does it.

And I've grown a little wary of that approach and more, almost probably shied away and gone extreme the other way where I pushed the team to like. What can we do where we have to roll our own? What can [00:23:00] we do where there isn't a tool yet? Where, what are those things that we can do? Because one, my hypothesis there is like once a tool's built, right, like marketing channels work when they're new because no one else is doing it.

So it cuts through noise. As they become more popular, they become less effective because more people are doing it. Like email back in 2005 worked really well. Email now is just spam. And so, but S one builds a tool and the tool gets more popular, like as that company grows. Unfortunately, that means like.

The thing that it built is probably less effective as it grows because it, it becomes a more commonly dumb thing. Yeah, I don't really have a good question there. I'm just more curious, curious your thought on that topic and, and does that mirror like things in product maybe at all, or

Karen: yeah, I mean, I think one thing, some, someone said this to me once, she's like, welcome to marketing where you have all these tools but no engineering help.

Jeff: yeah.

Karen: And I was like, yeah, you're right. So.

Jeff: So we build more tools to connect the tools without having to code, and they don't work very well.

Karen: Yeah, I mean, silly things like I was trying to connect something connect an, an integration. I'm following exactly their stuff, but it's not working. Something to ask them for support and they're like, oh, it's a bug in our system. You know? [00:24:00] Or, you know, things that I haven't had to do in the past, 'cause I've always had them been set up for me is, you know, like I'm trying to get you know, emails set up I have to know like how you set up the domain, like things like that. I'm getting people to help me, but I also had to like ask Chad GBT how are there different ways to set these things up?

Jeff: Yeah.

Karen: So, I think with any role where it's a little bit open-ended, especially at a company where not everything is established and perfectly, you know, in the right place necessarily, it's like you have to just figure out and get your hands dirty and be a little creative and , resourceful.

Jeff: Yeah, I feel like I've blacked out a lot of that trouble where, here I've been lucky where we very early on resourced for engineers that are dedicated to marketing and go to market. Yeah. And, but it, but it's been incredible because a lot of cases, you know, but it, it. It is at some, in some ways incredibly eye-opening and almost you, you kinda look back and go like, what was I thinking before?

Because we'll say, oh hey, I want to do this, so let's take this tool and do this. And, and you get an engineer who goes, what? You don't need that tool. We can just do that. I just build that thing and [00:25:00] like an hour for you and you, you realize like how many times we've bought things or, or done workarounds, where had we just had. Half hour engineering time, an hour of engineering time we would've had a way more elegant solution. And so going back to like, you know, to take it full circle, now that, you know, you have the product team doing some of these, you know, the minor bug fixes and stuff and shipping that , to prod, I feel like we should all start going back to CTOs or heads of engineering, VPs of engineering going, Hey, we just saved you this time.

I want that allocated to these other things so we can have better solutions for it. I'm sure that'll go over Great.

Karen: Yeah. I, I'm definitely thinking about what I could use engineering's help on. I have some ideas, but I, it just has, you know, matter priority in terms of things that we can fix ourselves versus things that we might need some eng help. But I, I'm really thinking about like those integration points.

Those are where the pain is. Versus using like, you know, standalone tools is, is fine for the most part of.

Jeff: I feel like APIs have to be something that are, are, are just not gonna be. Well, they'll probably exist, but , the complexity and the kind of prohibitiveness around them cannot be that long for [00:26:00] this world.

Karen: I've been trying to use some of these off the shelf automation tools where they have out of the box integrations with these tools. You know, there's a bunch of them that have popped up and I've tried a few. Unfortunately, none of them are quite where they're at, but they sell a dream, 

Jeff: That's good to know. I I won't go seeking it out yet. I'll give it a couple months then.

Karen: Well, there's one I've been thinking about trying, I just haven't gotten there. It's on my to-do list.

Jeff: Well, that's, that's been the fun of ai, right? Is like, it doesn't work until it does work. And it oftentimes, it could be almost magic seeming step function. A big thing that, that when we were building our product was we have, you know, if we do social replay and analytics and, and the point of our software is like surfacing what is important to the digital experience.

We were building what we call Ask Galileo, which is like the chat layer. And I had a very strong opinion here on this, which is not surprising if you know me. But I was like, we shouldn't just do a chat layer to do like. Build a chart, it, it, it still just means easier to use the actual UI to do that than a chat layer.

And the chat layer wasn't that great. We were using internally, it was kinda like and I left, I took like four day vacation to go to Niagara Falls. And I was staying on Slack most and still [00:27:00] communicating, communicating with people.

Came back a week later and. It wasn't like, oh, hey, it's gotten a little bit better. It was like, it magically works in ways we never even thought it was going to. And we had SDRs and, and customer success, people using it to do things that we had never thought it was going to do around understanding user behavior because there was like a, some very important but simple seeming unlock that we figured out.

And suddenly it just like had all these kind of, You know, functionalities that worked all of a sudden it was so, I feel like there's a lot of that in AI right now.

Karen: Well, if I put my product hat on, I wonder if they created better prompts or tooling in the background and.

Jeff: Yeah. Whatever it was though. But it was, it was like a four day change where it was just, it was so enormously different that and I, I've seen that across other companies and tools as well, where just all of a sudden one day it's like, oh, it works now. Cool. Well, speaking of, you know, timing and things moving quickly and, and the fact that you have both marketing and product on your plate, I do recognize that , while I could talk about marketing and, I mean, this is like the perfect interview. You're marketing and product, and we're talking about ai. I could do this for hours, but you, you know, you have [00:28:00] real responsibilities that you have to get back to. So, I appreciate you coming on this. This was, this was really, you know, interesting to learn about. Cool. To see how you think about. Product and marketing as a, as a function where a joint, joint leadership and then just, you know, functionally how the team over there is using AI and how you kind of push that forward.

I appreciate you coming on. If, if people have questions or wanna ask more can they reach out to you on LinkedIn or is there a better way

Karen: Yeah, LinkedIn's great. I'm pretty active on LinkedIn, so people can feel free to reach out to me.

Jeff: Awesome. Well Karen, thank you so much for coming on. I really appreciate it. This is fantastic. And I have to follow up a little bit and, and see how it's all going across some of these things. 'cause I'm always really curious when people get into marketing how, how it goes. So hopefully we can talk again soon.

Karen: Yeah, I'm looking forward to that. Thanks for having me.

Jeff: thank you so much for coming on. Have a good one.

Karen: Thanks.