Latest / The Chief Customer Officer Podcast / EP017 Uncommon Circles, Moderation Agents & The Rise of the Forward Deployed Engineer

Transcript

EP017 Uncommon Circles, Moderation Agents & The Rise of the Forward Deployed Engineer · The Chief Customer Officer Podcast

• Jeff Breunsbach: All right, welcome back to another episode of ChiefCustomerOfficer.io. This will be coming out to you on Thursday, May 14th. Jay, how we doing? Jay Nathan: Doing great in Kansas City. It's beautiful here. This is a really pretty place, actually. How are you doing? Jeff Breunsbach: I mean, yeah, I mean the Midwest is a great place to be, you know, I have fond memories of my time in the Midwest. There's great, you know, rolling planes and you know, you get a different beauty than the low country. Yeah, yeah. Jay Nathan: Yeah. Was it Omaha? Is that where you were as a kid? How long were you there? Jeff Breunsbach: gosh, 96 through 2002, maybe six or seven years. My brother went, you know, my, have, old ties to the Midwest of Nebraska. My, my, ancestors owned a farm out West Omaha called the suitor farm. ⁓ yeah. ⁓ So, so yeah, we. Jay Nathan: Okay. All right. didn't know that. All these years I didn't know that. Jeff Breunsbach: So yeah, I spent a bunch of time there and my brother still lives there so we go back and know hang out every so often and it's a good place. Jay Nathan: Visit with Warren Buffett. Jeff Breunsbach: Yeah, the Oracle of Nebraska or Omaha Oracle of Omaha, I think is what they called him. But yeah, we used to drive by his house, you know, I think was relatively modest for, you know, being a multi billionaire. If you if you looked from the outside, you didn't think, wow, that guy, you know, has hundreds of billions of dollars. Jay Nathan: Yeah. Yep. I think he still lives in the same house that he when he got married, I think. Very modest. ⁓ sure it's got a fortress around it, something like a digital fortress of security, but ⁓ cool. All right, ⁓ are we talking about? Jeff Breunsbach: I'm pretty sure, yeah, yeah. Yes. Yes. ⁓ Well, let's dig in. mean, I think we're, you know, let's dig into uncommon for a minute and we'll jump through and give people a little update. So uncommon, we announced it week, but we really think that there's a wave of, you know, customer use cases around AI and Jay Nathan: Okay, yep. Jeff Breunsbach: We don't think that there's a community out there right now that's curating those conversations and making sure that people are staying on the cutting edge of what you can do really talking about the things that they're building. So Uncommon is a community that we're launching. And great part is we're actually building it ourselves. We are building it and trying to, I guess, dog food using AI. So we've know, Cloud Code helping us essentially build a community experience that we're excited to bring to the market because it's essentially going to be crafted for what we think the the great user experiences at the end of the day. So we've been working on that. So think there's a couple of things maybe to chat through that I think are pretty interesting. So one, if you haven't yet signed up, sign up for our show and tell that's going to be coming up here on, it's about to butcher the date, May 27th ⁓ at 2 So it's a Wednesday. And so the whole idea there is that we bring three to five people who are actually showing something that they've built. It's live in the wild. It's something that they're using with their team. It's really like coming a product or software company ⁓ in terms of what their use case is for AI. It's literally a leader who's built something. I think last time we had a VP of CS who built her own basically CS platform for her CSMs to use all cloud code. She had it up and running in two weeks. Now it's been two months, she's been curating it, making it better. And so that's the type of stuff that we're doing. So that's going to be May 27th at 2 p.m. as a reference point, Eastern time. Yes, good shout out. ⁓ we'll put the link in the show notes here, but what else is going on with Uncommon Jay that you think is interesting right now? Jay Nathan: Eastern time. 2 p.m. Eastern time? Okay. Well, think so, you I think the thing that we've been thinking through on that is how do we really drive true engagement and from our GGR days, live like we're doing the AI show until in a couple of weeks, those really ⁓ become sort of the nucleus of it. Our goal is always meet one person, take one new idea that you can, that you can. run with as a leader or as a practitioner. And so I think you'll get that out of those events. But, you know, when we think about the community part of this, we're actually thinking about the user experience of an, of an online community for this. And we don't want to just create the same thing. Like community has been created and created and created again. So like, what are we actually trying to do here and how are we going to engage people? That's still one of the questions that I have. And so maybe we brainstorm that a little bit here. I'm worried that we're going to build just another forum. went out and looked at our old community the other day and it was sad, Jeff. Like a bunch of spam, a bunch of people who are, you know, trying to get the attention of people to test their product and, you know, to, get consulting work. It's like, that's not what a true community actually is. This is hard, right? It's hard to actually drive an engaging community experience. It's not a Slack channel. Jeff Breunsbach: Yeah. Jay Nathan: right, or it's not a slack, you know, whatever, it's not a slack work. But ⁓ should drive connectivity between individuals and people around specific ideas, especially right now, because things just moving so quickly. So I guess we've envisioned so far is a feed that is really curated toward customer success and customer experience use cases for And, you know, I think Jeff, should just, we should really, be hard nosed about the type of content that we allow in there. Like one of the, one of the little pet peeves that I had in our old community was somebody a question and somebody responding saying, Hey, I, we've done that at my company. I'd be happy to tell you about it. Let's schedule a call, especially if it was a consultant. Right. That is the antithesis of what we're trying. Jeff Breunsbach: Yes. Jay Nathan: to achieve, which is great. If you did that at your company and you have an idea of how it works, post it there, share it there, because like it's one to many, right? It's not one to one. We're not trying to create one to one connections for consultants and for individuals. We're actually trying to create content and information that broad swaths of people can learn and get information from. So anyway, that's my big fear in this. Like we got to solve for that out of the gates. Jeff Breunsbach: Yeah. Well, I think there's two things that are going to help us do that. One, think is we've made a deliberate choice with this community to charge for it. And I think that that's actually going to create one way for us to, I think, the experience, right? Is that you're going to have to essentially pony up. And again, I think we did that for a deliberate reason. We saw when we had a free community, the... Jay Nathan: That's right. abuse. Jeff Breunsbach: I'll take the challenges that come with that. Yeah. And I think people were trying to abuse basically network that we were building, right? Like they were trying to effectively use it for themselves versus it for themselves in terms of trying to bettering their own personal businesses versus like trying to better themselves and move helping others the So I think that's one thing that we're already doing. The second is I we've thought through this a little bit, but just talking it out loud is I wonder if we can have a a loop or a protocol that's running with Claude and MCPs and, or Claude Howard thinking about this, that when somebody posts a response or when somebody posts a question that like, you know, it used to be that you, if you were going to put kind of guidelines in place about a community, you'd have to have some human moderator go verify those or you'd have even a queue, right? Maybe it, maybe something triggered into a queue and then you had a human moderator go into that queue and look it up. And I wonder now if we essentially are able to. almost removed the human from that loop entirely. Maybe there's some edge cases that we got to cover for it, but I think, like you said, I wonder if it's for us coming up with a couple of different categories of posts that we could essentially make that we're checking for on the most basis. Cause I think that was also right. Like we had, we volunteers, we were hoping that those volunteers were in specific channels at specific times to be able to like verify this stuff. And you know, again, like that just created so many challenges that I think here, like we can probably create a more systematic protocol. So I think like, To me, those are two things I think we're already trying to do off the bat that to me would hopefully alleviate some of these challenges around building a community. Jay Nathan: Yeah, we need, we need a moderation agent. That's for sure. Built into our platform. ⁓ we probably need a, like you're just sort of mentioning the looping piece part, part of what ⁓ is, Gary from ⁓ Y Combinator, the Y Combinator posted a really interesting article the other day. And if you listened to like Alex Carpathy, is that right? Andre Carpathy. He's like the AI. Jeff Breunsbach: Yeah, ⁓ Jay Nathan: He's like the AI czar, like he's the guy, right? He's the one who, ⁓ implemented self-driving at This guy's like top of the top when it comes to AI and everybody listens to him. he has whole, and the way explains AI is very simple too, which is cool. But he has this whole concept of like looping and creating continuous learning and refinement systems. Jeff Breunsbach: I'm Drake Arpathy. Jay Nathan: Tan actually posted about this, his personal knowledge graph. So Gary Tan is the CEO of Y Combinator, right? And he's like, hey, ⁓ publishing and open-sourcing his personal graph ⁓ how he builds this big knowledge base of the people he's come in contact with, the companies, the conversations he's had, all that kind of stuff. And actually to do something similar now. I've talked about our... what we call Balboa big brain, but like our content repository at Balboa and our knowledge base, actually starting to build that looping knowledge base creation tool that essentially takes every call transcript and all of canonical bits of information about a person that we've interacted with, a company, an opportunity that we have, an ⁓ engagement that have, delivery, an employee that we have on our team. Any one of those core entities has its own knowledge repository. It's basically a markdown file for that particular element or item. And we're tracking history. So if I have a call, you and I knew there's going to be a transcript that comes out of this podcast. If we talk about, you know, one of those things, then we go update the knowledge base afterward. And then that call that corpus of data and knowledge just grows and grows and grows. And over time, it's exponentially valuable, helps people on board faster. helps us understand the situation that we're in with the client. There's just a lot. bringing this back to uncommon, I want do the same thing with the content out there and just be continuously harvesting the content and the best out there that you and I can then look at and sort of share with the community as a knowledge base the best and brightest of what's going on and customer success for deployed engineering. You all the things that we talk about on this podcast. Jeff Breunsbach: Yeah, no, think that makes a ton of sense and that should be. mean, that's a benefit of the community, right? Like if you are an active and engaged member of the community and you're in there, then like the more active and engaged you are. That means that the community knowledge graph is only going to get better and improve and we're going to be then sharing that back with you, right? It's a it's a loop of value in and of itself that like. the more active and engaged the community is, that means the better our knowledge graph is, which means the better the output is for you as a leader and understanding what's possible, what you can be building, how you can be improving your teams. And so think that that is a great loop for us to build around. that kind of leads me to you asked me a question over the weekend. I think you texted me yesterday about like, yeah, you know, this idea of just like, how are we going to keep people engaged in the community and like, what can we be building? And I don't have a great answer of like what the features are. Jay Nathan: Yes. Jeff Breunsbach: But I think what I came to was, I guess, more of the ephemeral of like, what are we actually, like, why do people want to join communities and what are we really solving for? I think it's this, to me, I think it's this concept of like, getting someone to be a better version of where they are today is like the reason that you join a community. Like that is the, like if I'm joining a community, obviously it's not to be a worse version of myself, I don't think. ⁓ And it's not to be, it's not to be. Jay Nathan: Yeah, that's right. Jeff Breunsbach: It's not to stay steady, right? It's not to be at the same place that I am. Like I'm making a bet that I'm going to basically be surrounding myself with people that are doing better things than I am that are pushing the needle that are, you know, was equating it to like, I remember when I was in middle school, I used to play basketball and I wanted to be better. And so my coach was like, great, you know, in your free time, you should be playing with kids that are better than you. And the whole concept was that like, If you're playing with higher talent or people that are more talented than you, essentially at some point those skills are going to like almost not through osmosis, but essentially like through playing the game with those people, you're going to notice things. You're going to see things that they do. You're going to able to pick their brain, right? There's all this kind of, and I think that's the same reason that people join a community. And so I think of like, at the end of the day, we're trying to help somebody be a better version of themselves. And we need to make sure that like the features of our community, we can all tie back to that, right? Like, okay, I'm today you're at point A. and I want you to get a point B. And so everything that we build as a feature should be geared towards getting you to point B effectively and efficiently as we can. Jay Nathan: All right, you just triggered something for me that'll connect to something you and I have been talking about for a long time. ever heard the that you basically become the product of your five closest friends as you grow up and get older? What if we, well, for a long time, you and I talked about small groups ⁓ in as well. What if we had uncommon circles where once you're the community for, ⁓ you know, 60 days, you get matched up into a circle of five people in the community that are of your circle going forward. then you, you know, you engage on your own and ⁓ we pair you based on the kind of company that you work with. Well, actually the more the better actually. So anyway, maybe we do uncommon circles as part of this. Jeff Breunsbach: love that. No, I think that's true. I think what you know, like maybe like the route you're going down is you try to like, are some attributes of those people that you could essentially like bring together like a recipe, you know, like you said, you maybe want somebody who's further ahead. You maybe want somebody that's a little behind that's learning stuff. Maybe you want somebody that's in the middle, different industries, different sizing, right? Because again, I think like, I think too often we try to just replicate a company that looks like us and you're kind of like, okay, Jay Nathan: Yeah. Jeff Breunsbach: But that's not really where innovation comes from. Like innovation, like I have a team of four CSMs right now. I'm probably doing stuff that a large company might not be able to do, but still is valuable to them to see, right? ⁓ that's possible. How would I go do that in my kind of situation? So I love this. The, ⁓ and the thing I think you could be doing around these circles, right? It's like, could be building a set of simple features that help that group, right? So if we wanted, if we wanted these uncommon circles to become valuable, right? Can we help them schedule meetings? Jay Nathan: Yep. Especially right now. Jeff Breunsbach: Is there a way that we could essentially help them schedule their monthly meeting? Is there a way that we could essentially try and spur them to ⁓ discussions amongst each other? Is there spaces that are closed off just to them? I think kind of a feature set that comes around that that is interesting. I think to your point, I think that should be the... mean, there's a concept that I think is probably more doable in today's age with the tools that we have. I mean, if you remember back to our game grow retained days, like I used to be throwing events for those 150 people, you know, a hundred people would show up and we'd be building tables. And I was doing that by hand of like, okay, which industry does need people in? Let me get them to answer some questions. And so now I think if we are able to come up with those attributes and criteria, it's a little easier for us to spin out, but I think that's a great. A great concept we should be building into this. Jay Nathan: Yeah, I mean, that's the beauty of our own community platform is that we can be as specific and strongly typed as we want to be around some of those concepts. So, ⁓ right, I just got even more excited. So we need to net it out. need, we've got to have moderation agents. We probably need content curation agents so that, you know, we can lighten the load, but... Jeff Breunsbach: Yeah, that's cool. Yeah. Jay Nathan: I don't want just like automated crap being posted obviously, right? ⁓ And then. Jeff Breunsbach: No, it's almost like the way I thought about the content curation piece, because I was looking at our community this morning, starting to dog food it and starting to post is, yeah, I think there can be a series of prompts. And I think that the, in the past, again, I think that those are almost like hard coded where it'd be like, what's a recent win that you had or like what's a loss that you've had. And like that was just almost like hard coded. It's like every time I would go, just be that same one. And I wonder if you can. And now if we can like open up the aperture through AI to basically have prompts that are different every time you go to like the community to look to posts, like is there, you know, do the prompts change? Hey, this week it's, you know, what, what's a recent one that you had this week that you're able to deploy for your team. Maybe next week it's that one doesn't exist in your little view. Maybe it's something else. And so I think like being able to curate like different questions and concepts as you go. Jay Nathan: ⁓ yeah, yeah, that's interesting. prompt agent? Well, I think also could be like an engagement agent, right? Like what are the types of, based on what's happening in the community, what are the types of like digital outreaches that we'll send from the community to the members to get them engaged? Like, so one of the things that you've done a really good job of with our newsletter is every once in a while you'll send a note out to our newsletter subscribers that says, hey, Jeff Breunsbach: Yeah. Jay Nathan: just checking in on you here. This is not a newsletter, it's just a quick check in. What's one thing you're working on? How are you liking the newsletter? Like just a couple simple questions. Those are really nice nudges. They're all pre-programmed and deterministic, right, which is cool. But now we have the ability to sort of send nudges on a very bespoke basis, depending on who you are part of our community. So that could help drive engagement. Jeff Breunsbach: Yeah. Yeah. And that to me, like, yeah. I think, think like that could also, like you said, right? Like, Hey, seems like you've been most interested in these types of topics before we just had two new questions that, you know, were asked from the community. Like you'd be a perfect person to answer those. and, I think like that being probabilistic, right? Hey, you've answered these recently. It's been, you know, almost like in our algorithm, right? It's been something you've done recently. That's helped us like to trigger those things. so yeah, I think that's a good idea. That's cool. So yeah, so we need a couple of these agents and then it seems like, I guess some of the, we'll call it like the foundation of the community experience around like kind of building around this idea of uncommon circles, which I like the naming of too. I think the branding, I think the branding we kind of threw out there, but it's sticking for me. And I feel like we're, can kind of build stuff around it that to me is like starting to feel like it folds in, which is cool. Jay Nathan: Yeah, me too. Yeah. Yeah, I think naming things is the hardest thing you can ever do. And so just like having that taken care of is just makes life so much easier. So thank you. Good job. Jeff Breunsbach: Yeah, you got it. right. Well, that's enough of our uncommon corner, but hopefully, you know, some people get some value out of some of those ideas. And, you know, it to me actually screams like, could you be doing communities for your customers so much better now? Like, like, could you actually be having a more thoughtful, curated experience for your customers ⁓ terms of like community building? And it kind of gets me to a question right now, which is interesting to me, like, I we've talked about this before, but at Jay Nathan: I was just thinking this... Yeah. Jeff Breunsbach: Junction part of our, you know, we're an API and developer first platform, I would say. so like part of our shtick is that when you, when you pay for like a platform fee, a ⁓ fee for our products, essentially get access to us through Slack. It's, know, we see it as a major benefit because if somebody is developing something, we want to be there to answer questions quickly, to be able to move them along. Like we want them to be essentially like rapidly. We don't want them to feel stuck at any point. And. Like that to us is like a big differentiator. right now though, that's a very human driven experience. Like we have support agent, you know, we have support people, we use a tool called Pylon, which I think we probably need to be leveraging some more AI from. And I've got CSMs. I mean, we basically have all of our team in there, right? And so, but I wonder to your point, like right now it's just very much like we wait for them to post. Once they post, then it triggers workflows, but that's really just. support tickets that go into a queue that a human's looking at. And I know we've developed some technology to basically go start triaging tickets with like some but I wonder if there's other, I don't know. It just makes me think that there's other things that we probably should be doing in ⁓ that that are less triggered just by the customer solely and ⁓ more triggered us or other questions or things that we're seeing. Like how can we be more proactive in those channels for our customers might be a better. Jay Nathan: So, would say my first reaction is like, who's doing the coding now? It's not against Junction, against your APIs. The agents are doing the coding. It's agentic engineering now. So I think we talked about this before, but like, how do you take all of that knowledge that you have about how to answer those questions and make it available to Claude through a CLI, through the API docs, like what's your self, you know, referencing knowledge graph, you know, builder. Jeff Breunsbach: for, ⁓ like. Yeah. Jay Nathan: system that you're putting in place based on all those questions. I think ultimately you got to do if you're selling a developer product is ⁓ you to really cater to who's writing the code to make that thing work. And that is the agent. The engineer is just putting the system, they're putting systems together. They're putting the pieces together. And of course you need an outlet, right? But like the more you can actually do in the agent, the better. So to me, that's a bunch of CLI. CLI means command line interface. And so when you have a command line interface to your products, I'm not telling you this, Jeff, because I think you know this, but Cloud Code ⁓ run commands and call commands and interact with your product through the command line, because that's pretty much how Cloud Code, you know, that's how it works. So a lot of coding agents work. is through the APIs, the MCPs, and the CLIs. A lot of acronyms there. But that's how I would solve it, I think, as opposed to trying to make the human process better, because the human's not your user. Jeff Breunsbach: Yeah. Yeah. No, think it's definitely, it's The other thing, the other challenge that I'm working on trying to solve right now too is. slightly similar vein, but I guess different end user where, you know, most of, think who are core buyers are, are like actually product managers because like the product manager is responsible for, you know, the product manager is essentially responsible for that product line. so they're the ones who are looking for solutions and for, okay, what's the product experience look like? What's the user experience? Like they're the ones thinking through all those things. And think we've We've got a lot of API docs and developer docs built, but we essentially haven't translated those, think, to almost like to the business side of like what we're trying to help them accomplish. And so it feels like there should be a agent that we're essentially like, how can I translate these API docs into an experience for these business users essentially to help them understand like, okay, why, you know, why are you doing certain things in certain ways? You know, why does it make sense to use Junction? Like some of these questions that I think like, we need to be helping them answer it is I think like there's a almost like this. We don't need to create. I would say like something that's totally net new. It's basically we just need to translate these engineering and documentation that essentially helps you understand how to do the coding into, okay, what does that actually translate to in terms of the business value? And then like the user experience or the, the, end user experience about how this is deployed in their systems, almost like their patient experience. Jay Nathan: Yeah, absolutely. It's like just a different view on the same information, right? Jeff Breunsbach: Yeah. But it's, but to your point, think that, and I like what you're alluding to is essentially like the core information shouldn't really change. It's just the, essentially your end user of who you're getting to look at it change. And so that, that, that view essentially is just a different lens on the same information. So if you have the underlying information, you should be able to be able to essentially architect experiences for whoever that end user is. And agent is one of the end user experience, like is one of our end users. Developers is one of our end users and I'll call them product managers. or the, our, like our third end users. And so like, we should essentially have the same view of information or we should have different views of the same information, sorry, for each of those end users in order for them to essentially have what they need to be able to do their jobs. Jay Nathan: Yeah, that's super cool. what we wrote about in the news, you just triggered something for me because you talking about the CSMs and some these technical tasks. Are your CSMs technical in nature? Do they know how to code against the APIs sort of show things, or are they more relationship folks? Jeff Breunsbach: Uh, there are more relationship folks right now, but I would say they have, um, I would say so far we've been putting them in situations where they've had to at least understand the technical components of it, understand our endpoints, you know, what's possible, where we should be doing certain things, like certain actions that are essentially are, are the endpoints that they're calling. Like, uh, so I wouldn't call it that they've had to, you know, that they're super technical and could, um, go into a, um, go into like an engineering call and be able to like hold themselves. But I think what they can do is translate. think they can, right now they're almost like the translation between some of our API docs and then like the end user experience. They can help our product managers understand, okay, you know, which end points are important for the user experience you're trying to create or the patient experience you're trying to create. How, you know, at what point would you call them and like. That's, I would say like where their technical aptitude is right now is like this lens of, let's help you translate where you should be calling our documentation for like your patient experience then like relationship focus. But I think, I guess like I'm looking at as a, we should probably get that like technical aptitude into the CSM skillsets as we go. I think it actually starts to merge. Like we should, we should basically be, Jay Nathan: Yeah. Jeff Breunsbach: merging that into one role versus like having a multitude of roles. Jay Nathan: That translation layer to me feels like it's ripe for AI agents to play a strong role in just because I mean, that's actually what it's really good at. But I was thinking back to the article we shared this past weekend about forward deployed engineers and ⁓ how, know, ⁓ like and more we see companies that are moving toward one role for deployed engineer or a role called forward deployed engineer. Jeff Breunsbach: Yeah. Jay Nathan: It's not forward deployed CSM though. Like I think some people have tried to just commandeer the forward deployed language and I don't necessarily think that's correct. the original definition of a forward deployed engineer as defined by the people at Palantir who really this concept years and years ago, even before the Chad GPT moment was 20 % sales. So that's SE portion, right? They support sales 30%. Jeff Breunsbach: Yeah. Jay Nathan: product, meaning they're not just engineering custom solutions, they're engineering in the field to identify the product use cases that need to be productized. And then 50 % engineering, which is you got to have the technical chops to do it. Now, I think the thing that's missing here, like, so we're starting to do some of this work, you know, I've talked about it, just AI. data and AI projects that are agnostic to any platform per se. But I think what's missing in this is some level of engagement management, meaning like for that customer, they still are trying to get a specific outcome and we need to be delivering on that outcome. Yes, the forward deployed engineer is like helping to prove technology and identify product, productizable. solutions, but you still have to deliver for the customer, right? So it's just hybrid of AI delivery productization. And I think whenever you have delivery, beyond the proof of technology, you have to have some level of project or engagement management and some level of solution consulting to where you're actually re-engineering business processes. You're putting the technology in to support those business processes. Jeff Breunsbach: Yeah. Jay Nathan: And I don't know that that's a one job, right? All those things, change management alone and some of these things is gonna be a massive, massive undertaking. When you think about, was on the phone, I'll leave them anonymous at least, but I was the phone with somebody last week who they have a team of people in offshore, hundreds people. that handle their invoicing because it's all done manually for a long tail customer base, thousands of customers. They sell devices ⁓ monitor and track those devices. like customers will call and say, hey, you're billing me for six devices and I only have five. And a human has to go deal with that, like $200 difference. and it gets one ⁓ them, example the person was giving me is that It got escalated all way up to a VP level in this company, thousands of employees, right? Big company. it was over $200. And this is happening across hundreds of people who are handling invoicing. think about the change management with like evolving that kind of process. And that's a customer facing process, right? So huge, huge value there, but I don't think an FDE alone goes and does that. Jeff Breunsbach: That's pretty funny. Jay Nathan: It's my only point. Jeff Breunsbach: Yeah. Yeah. Like I, I think that's where, we're, I'm hiring for an FDE right now on my team. ⁓ decided this last week. and I'm thinking of like the FDE and the CSM should basically be, we should carve. Yes. Like we should carve that out. Like you said, like we should basically carve discrete responsibilities. And I'm like, I think to your point, like we're landing on the CSM. Jay Nathan: ⁓ cool. Okay. Tune a box. Jeff Breunsbach: being the relationship management, the change management, the layer lens to our business that should their sole responsibility along with the commercials. So my team is handling renewals, upsells, cross-sells, they're handling all the commercials on those things. And then you should essentially have access to an FTE who is working with you essentially make sure that the solutions that we're giving to our customer are in place and operational, that they're optimized. that we're making sure that they can achieve what they can. so like, like you said, like putting them together on a customer to me is like the recipe for how we do that. And like you said, there should be a discrete level of responsibilities of like, where's the CSM going to engage and what are they looking to do versus, cause it's, I don't want the same, ⁓ Like you don't like the fear of everything. that like, cool. Now I've just got two people that are effectively just like sharing work or passing work between each other when like, it should really be that there is like a discrete number of things that this person, like the CSM can do and should be responsible for and a discrete number of things that like the forward deployed engineer should do. And like you said, I think like the FDE to me is the like technical, like we need to technically get on level with our technical counterparts at these customers in order for them to basically make sure the solutions they're oriented are in and operational. And like the last thing I'll mention too is I think like this to me allows us to also have a stronger point of view about how our, how our customers are deploying our solutions. Like I think up until now we've kind of let them build a little bit of how they wanted to. because, know, it's flexible. It's an API. you can kind of use it, you know, where you need to. ⁓ And think now we can become a little bit more specific around like, this is actually how we've designed like these experiences for these situations. And this is how you should be deploying it. And I think that's actually going to less impact for us in terms of like support tickets and custom situations, you know, that we find customers in today. Jay Nathan: Yeah, I guess the key question for you is, and if you don't want to talk about this on the podcast in public, that's fine too. But like, where is it going to report? Is it going to report into product and engineering or is it going to report into customer experience? Cause like, Jeff Breunsbach: Forts into me. The FDE will, yeah, the FDE, it's gonna report into me. Jay Nathan: So do you think that that dilute, like if you think about the 20 % sales, 30 % product, 50 % engineering, do you think that dilutes the product element of it to have it report into the customer experience side of the business? Jeff Breunsbach: Um, I don't think so because I think, um, one, we're a small enough company now that like effectively who you report into, like, don't think, I think we're in enough common channels and spaces and we're doing enough common work that like, I don't have that fear right now at the size that we are. Um, and then two is I think, um, the more that I've thought about this is I think we need to, um, Jay Nathan: Yeah. Jeff Breunsbach: I guess we need to make sure that we've got the right mechanisms or loops internally to ensure that we're handling those situations in the right way, to your point. Are we productizing the right things? Are we thinking about solutions that could not only impact this one but multiple customers? How do we get this level of feedback back to the product team in a way that they could actually engineer a future solution or something that's bigger? So I think to me, it's... It's more just about creating the rituals, I think, between the other teams on what those teams need to make sure that this, ⁓ this deployed engineer is like going into those situations with like, what are some of the core concepts? Like you said, like, think outlining some of those things upfront is probably going to be a big part for what I should do is like you said, like we need repeatability. We need products that can essentially scale across multiple customers, not just single customer solutions. Like, I think that productization piece will be a big aspect for us to outline. And then. Jay Nathan: Yeah. Jeff Breunsbach: Once they uncover those situations where they feel like they can productize something that they can build something. think like, to me, does that mean that they're spending time in linear building like similar tickets and similar they're in similar spaces that our product and engineering teams are? Do they attend? You know, in my mind, like I think they attend probably sprint planning and some other stuff, right? Like I think there's probably some rituals that it just makes sense for them to be in. even though they kind of report into me and you know, effectively, like we're trying to handle customer situations. Like it seems like there's just rituals that we should probably get them involved in that keep them close to those teams. Jay Nathan: Yeah, it's like it, it almost doesn't matter. where people report these days, if you're working in an agile, agile is so out of vogue, but if you're working in like these flattened organizational kind of structures, which we're seeing more and more. And I I think what you're saying is exactly right. What was the other thing I was thinking there? Like, I mean, at a minimum though, does your engineering team get to interview this person? so that they're brought into who it is that you bring on, right? Because that's half the battle, right? It's like, hey, we are the engineering team, we're the product team, do we trust the people that are out there in the field giving technical advice to our customers? do we trust what they're bringing back to us? I mean, that's what it really comes down to, right? Jeff Breunsbach: Yeah, yes, product and engineering are involved in this, as well as our CEO who is engineer by trade as well. So like we've got, I would say that we've got the technical side certainly nailed down for sure. ⁓ But I think like to your point, like the, you know, the, thing we're just trying to look out for is the, like, can we put you in front of customers, right? Do you have like the, do you have the skills to be able to, communicate effectively with customers, to be able to be presentable, to make sure that like, you know, we've got kind of the right, Jay Nathan: Cool. Yeah. Yeah, that's cool. Jeff Breunsbach: external facing view of company. Jay Nathan: Yeah, that's the interesting thing here is like, how do you scale this role? Right? It's not, that you need to scale it today necessarily, but like how did I need to go learn more about this? Like how did Palantir scale it? I mean, I'm sure they pay very high prices and I'm sure if they have to give anything on these roles, it's probably the soft touch of ⁓ a, of a customer facing person, right? They want an engineer first and foremost that understands the tech through and through that can help make Jeff Breunsbach: Yeah. Yeah. Jay Nathan: hard decisions that the customer has to make can bring insights back. And sometimes it's know, timeline, and ⁓ scope, you two. And I'll tell you the third one is gonna be. So I think that's like ⁓ the same way to about these kinds of people. Jeff Breunsbach: Yeah, agree too, by the way, just in terms of like, I think like the, facing stuff is probably last on the list of things that would essentially like need out of role, right? Like you should really optimize for the product. Like I think you should, it should optimize for like the engineering background. think for like a systems thinker, like there's other kind of like attributes that you'd want to, I think about something like, you know, like what's their ability to, be proactive and like basically, you know, Jay Nathan: Yeah. Jeff Breunsbach: do stuff before they're asked. Like those types of skill sets are the ones you want to optimize for. ⁓ Jay Nathan: All right. So, just a quick plug, you know, we do these digital experience roundtable sessions for Balboa or with Balboa. in June, we are, our, session going to be on four deployed engineer and like the collapsing of the roles, like we wrote about this weekend, into one or, you know, five down to two or three or one role that we're seeing across the industry. So. We can share more about that. think that'll be an interesting... We may want to do the same kind of session with the uncommon community as well because I think there's a lot of people learning a lot of things right now that... It's interesting. What you typing there? Sounds like you're clicking, clicking. Jeff Breunsbach: I was just sending you an article. I was sending an article so didn't forget before we jumped off. It was about stuff for Uncommon. So I wanted to make sure I didn't forget before I did that. But yeah, that would be an awesome session. So looking forward to it as well. Cool. I've got to run, but this is a good one. I appreciate you talking through this. To me, it's always fun to get together and do it. Jay Nathan: ⁓ okay, cool. Okay, cool. Yeah, man. All right, have a great week. All right, bye. Jeff Breunsbach: All right, we'll see you soon.