Yeah, yeah. I mean, we should get another month hopefully of semi good weather. So yeah, the barbecue is certainly getting moved closer and closer. Well, I really do you have any Irish heritage, by the way,
yeah, I mean, a bit,
okay, just just Pollock, because in my elementary school there was a family of Pollocks, and they were in classic Irish style. It was like two parents are like, eight kids. They're all so I was just wondering, when I saw your name, I was curious.
Yeah, I think we claim it's Scottish. But I mean, you know, Scotland and Ireland are close, yeah,
origin, yeah,
I don't know. I mean, they say, Yeah, our family's Irish ancestors came over from debtors prison, you know, like five generations ago. So I'm not especially close to it. But like, you know, a lot of Americans are basically Scots Irish,
yeah, they, I think we count 70 million Americans, yeah, which is amazing, considering that there's only four and a half million of us here. Well, listen, I really, I really appreciate you jumping on and speaking to speaking around random stranger. Great to connect. Let me give you a little bit of background on me and what I've been doing, what I'm up to. So I live back in Dublin now. I've lived in a whole bunch of places, in San Francisco for a while, across the US and various places in Europe. I started in tech right out of university, went into a sales role in an Irish tech company called intercom. They do customer support software. And, you know, following a few different things there and a few different other jobs elsewhere, I started my own business back in 2017 2018 very different to what I'm working on now. The premise at that point was we were building an AI audio co pilot for blue collar workers. So folks that worked in hotels and healthcare manufacturing, the premise being, they work on their feet, their eyes and hands are super busy. They get of all the money invested in technology, they get 1% of it. And so there's just this huge kind of inadequacy or lack of lack of investment in technology for those folks back then, voice was a disaster. It's still a little bit of a disaster in terms of, you know, accuracy, but particularly when you're working kind of high noise environments, you've got very diverse teams, language, dialects, accent, cell, things is just all over the place. What we actually ended up doing. We pivoted the solution in order to deliver the voice. We had to understand what every single person on a large frontline team was doing every day. And we were able to just kind of convert that to a work management solution, replace all the paper checklists, Paper Assignment boards that they were using, and just digitize that world that went well. We we kind of grew that business through through covid Because, you know, handing stuff out. Managers of teams were, were were at home. We saw that business in 2021 completed, I don't know if you've been through that sort of a process, but completed all the lock up and the non competes, and all of that beautiful legal crap, which expired a couple of months ago. And so now working on my second, second business, and this one, is really born out of a frustration and a pain that we had when we were growing our first business, which is, okay, you know, more and more of a data inside of businesses unstructured. More and more of that unstructured data is conversational. How do you actually just stay on top of everything that your customers and your prospects are telling you, whether it's about your product, about your pricing, but service that you provide so so often kind of referred to as like customer feedback. But actually it's kind of a broader, a broader thing than that. So I have some like, really concrete ideas on what I want us to build and how I want us to kind of do that. I'll share those in detail with you at the end. I don't normally speak to VPS at companies that are quite as Big a scale as you Guys. You still there. I
Hey,
system crash.
The computer is restarting. It does not do this because it's highly disruptive, but, but here we are. We'll do it on the phone.
No worries.
Are you sure you're all good? That's yeah, no, it's no worries. Just cool, yeah, yeah. Sorry, sorry, yeah, you were gonna talk,
yeah. So, so, that's kinda thing. So, so we sell that business. 21 we completed all the other stuff and started the next business, born out of that frustration and in kind of the unstructured data world, conversational data, how to map, how to handle everything your customers and your prospects are telling you across a variety of sources, we got some really concrete ideas of what what we want to build. We'll go into that in a little bit. I don't want to color your color reviews too much, but, but can you kind of think, if we, if we go into that domain for a second, can you kind of share a little bit like, what do you see as as some of the kind of biggest hair on fire macro challenges that you've experienced over the course of your career in dealing with that sort of data and dealing with that sort of feedback and input from customers and prospects? How do you think about that world and what comes to your mind as I start to describe that? Yeah, I
mean, I guess so I've been at upguard for 10 years, and so it's been a while. So I was in before that. I was in B to C companies. But, I mean, it was a long time ago, so my context is basically upguard, which is B to B security, and it software. And we have, we have some systems in place for this that I think are reasonably good. They could, they could be better, even with just what we have. But so we have our tier one support people are in house, and they're, like, reasonably well trained, so they can have pretty decent conversations. We also have always had a pretty extensive CSM team, and, yeah, a lot of that stuff gets recorded with Gong. And so we have, we do have a shitload of data. We have, yeah, text conversations through intercom with tier one support, and then recorded video with CSMs as well as our AES. They most, most sales demos and sales interactions. So pre sales and post sales get recorded. And as I'm sure, you know, Gong does a decent job of summarizing. It's also it, um, it's, yeah, the
summaries are way better than the transcripts. Yeah.
I mean, what I'd say is, like, this is what I say generally about Gen AI is that it's, it's not wrong, but it may, it may not be helpful, because so it'll be like, Oh, these people, like, here at, you know, minute 20, they talked about pricing. But What the Did they talk about? Like, what, like, what? How did that guy feel, and when he said that, you know, um, so, so we do have a lot of the raw data, and then we have the AI transcription, etc. It's okay. It's the, the way that we operationalize it. That's a little bit better. Is this something I do like about Gong? We use its integration to product board, which is, in theory, our intake for customer feedback. And I like that a person can tag some part of a gong call, and then send that as an insight to product board with whatever other color they may have on it. That is, that is useful feedback. The problem is that it gets done extremely inconsistently and infrequently. So, yeah, I mean, and another, I guess, problem that I see is that the you know, the they're the motivations of the people involved. So AES are motivated to make their deal. CSMs are motivated to retain their customers and and so they are listening for, like, what are the problems I need to solve to to get that milepost. Whereas, as for me as a product manager, I'm listening for like, what's this person trying to do? How do I how do I build a product that satisfies, like, what they need and and so the feedback that we get are usually these, like, fairly specific things, where it's like, well, I want this CSV to be formatted this way. It's like, it's like, okay, cool. Like, I gave your customer watch, that that is not a game change. That does not change our business. What I want is the feedback across, yeah, across, like lots of users. And then also sometimes this is maybe, well, this is specific to us, but I think it probably is generally true we we sell to there are different parts of an information security organization, and like very different users and possibly teams within that. And so we may be talking to the the vendor risk management team, and they're sort of like run, laying or describing, oh, my colleague and incident response would do this, but I don't actually do it. So it's like something I spend a lot of time not doing is being like, can you give me that guy's email, and can I, can I get that guy to talk to me, to tell me what he wants? Because it seems like there's, there's a customer for me over there, and it's not you, and I need to get to that person, and I need to, and I can't, I don't want to talk to you, the vendor risk person, about your speculation on how Incident Response uses it. Because, yeah, yeah, yeah. So those are the, yeah. Those are, like, the sort of the technical systems are gong and intercom, and same thing with intercom, then those can be sent over to our sort of feature tracking intake. Our feature tracking is, in theory, well, we're redoing this in a stupid way, but like the way it works pretty well is things go to product board. Cool, at least those are those linked back to the original source, then PMS intake, those, you know, link them up to features. We've defined them, and we can use that to, then, you know, measure volume of demand, and then, and then, also, like specific churn risks. And then, from then, from there, it goes into, you know, like a product development, like the software development, is not really interesting. That's, that's the point at which cool. It's a feature. It goes in, we goes in JIRA, etc. It goes
into JIRA, yeah, understood. Okay, interesting. So, so tech stack is, is kind of as expected. So, really interesting. Way talking about in the gap, in motivations, right? Because this is something that we're trying to, trying to think through as well. What I found from being one and have from running, running those teams is like AES will feedback from AES, and often CSMs is governed by, like the loudest voice in the room, and the loudest voice in the room is also most often governed by who has the biggest renewal, who's got the biggest net new deal coming through the pipeline? Is like, hey, fucking product just built this shit. I'd close this deal already and like and like, way more often than not, the deal still doesn't close. And then there's a new request, and it's like, you know, because, because sales people aren't in their nature product people, and they're not PMS, they're not good at ideation or kind of solution planning, or really problem discovery from a product perspective. And so I think that's really interesting. So what do you guys, if anything, have you tried to do on the compliance stuff? Right? Because it seems like it there's a failure point pretty early on. Like, if CSM or an AE just doesn't bother to tag it in Gong and send it award. Like, if they just turn around to a customer, like, hey, yeah, that's on our roadmap. We're looking to build that in the future. And they bullshit, which happens all the time. Like, you just missed that bit. So, like, have you tried to do anything on that. Is that an active thing? Do you?
I'd say that's not really a problem. I mean, it's there the, there's enough, like, visibility of what they've done that, that CSMs, yep. Like pretty they reliably pass on their customers requests, I would not say that they bullshit about what's on the roadmap. If I mean, the more the more common problem would not originate from the CSMs, but from leadership, where we put something on the roadmap and then they say, actually want you to change it. And so now we go back and do that. But um, no. I mean, like, yeah, the the compliance is, I mean, it's sort of, it's in their interest to capture any customer requests so that if a customer churns, then it's like, Did you well, did you do your job? And they're like, Yeah, you can see, I launched 20 feature requests and they were never delivered. And, and I would say the business, like, reasonably good at understanding, like, okay, cool. We we have those feature requests. We do have them in product boards so we know the level of demand from other customers. We're not going to build things for one one user. Or if we are, then it's going to be because they have, you know, a half a million dollar deal or whatever. Yeah. So yeah, the, I would say, like, the compliance is not the issue so much as, like, the there's just a blind spot of, of listening to, also listening to things that the product does well, that we could use, that I would use for, like, go to market. So if you know, CSM gets on a call and they're like, oh, yeah, we like all these things. But here's the problem. The thing I'm going to hear about is, here's the one thing I need, as opposed to here are the 10 things I could use to market this product better or to better define our ICP,
interesting. Interesting. Do you feel like that is like, full capture of the negative stuff and full capture of the positive stuff. Like, you know, the really high level way of explaining the business impact of that is, like, yeah, we lose revenue, or we have a higher churn rate, or, like, we win fewer deals. But like, it's kind of easy to always trace anything back to those points. Do you feel any like, real kind of impacts of those as they happen in your business? Yeah. I
mean, so we've been going through a project to better define our ICP and then better align everything around that, yeah. I mean, just in part because, you know, like when you're early on, then, like, maybe you you intuitively know your user, and then you don't need to do a huge volume of deals for the business to grow at a good enough rate. But now we have enough different, you know, we've done enough different, like diverse deals, that there's some noise in the system, and we really need to be like, Well, are we actually successful with FinServ, or do we just have four of those in a row? And that's like, you know, misleading, yeah. And so, yeah. And so there really is interest in, I mean, I guess churn is really the problem we're trying to solve. It's been, been growing. And so there are all these theories about why. And I think one of the theories, which are not just like multiple theories, I think there are multiple contributors. I think one of the contributors are deals we've done in the past with, like, non ICPs. And so it's like they come in, like, for whatever reason, they buy our product, and then the year later, maybe they realize, like, their business can't really operationalize it. They could just use some free stuff. It's like, for whatever reason they turn out. And so, yeah, and so there is a real desire throughout the business, high level goal, being reduced churn. But then also, like each of me, like the product team, the product management team, like all of our kind of, like, OKRs are around, like, are you defining ICP and are you, like, you know, working against that?
Yeah, interesting. I mean that that's often the case, right? Like, when I was early intercom, we had the same we had three different products at the time, three different ICPs. And sales always wanted to do the deal today, and the rationale is like, but isn't it better to get the revenue for one year than none, and then products will say no, because trends going to go up, and there's a cost of delivering that customer and all that that brings in the wrong direction, so that that dichotomy is always there. Um, okay, interesting. Yeah,
that exact thing of like, yeah, like, like, we're trying to make this quarter's number. It's, it's tight. So like, just take the, just take these five shitty deals, and then we'll make the quarters number, and then, like, we'll deal with this later. And then after a while, then it's like, well, yeah, it adds up, and it makes the Yeah, the CAC goes up. Like, all the calculations start to get shitty. The business looks bad and and ultimately, then it affects, you know, like, your multiple in a way that is more important than, like, the $50,000 of revenue when we're sure, yeah,
sure, like, if you're talking about shareholder value, like the it definitely destroys that more over time. Interesting. So if you could think of a way with all the gone calls that are happening, like, you guys are a big scale, right? So you're having 10s of hours of gun calls every day. I don't know what volume of intercom conversations you're having, but probably hundreds a week, or something like that, right across your Yeah,
I don't know that. I'm not as I don't look
as much, yeah, but, but, like, some some level, like, what would be the ideal solution to try get the value of all of that stuff. Like, is it that? PMS continue to screen. Like, how would you think of an ideal way if you could just click your fingers and do that better? Yeah. I
mean, obviously I'm in product management, and I only care about my interests. But yeah, I mean, the thing that I think that I try to do, which is it's time costly, is go to Gong, and then there'll be different topics that I want to learn about, and then assuming the transcriptions are correct, then I can search, and I can find those conversations again, I'm just like, assuming I know there are errors in it, and then there are also, you know, specific terms of art or company names that may or may not transcribe well. But yeah, I mean, so that's so it's like, I have, I have, I guess, like, strategic topics that I'm interested in, and I want to know what's being said about those, and really the only way for me to do it is to go search across the corpus of recordings for those and then go listen to those calls and form an opinion based on that which is, which is quite time intensive. And I also, I don't even really trust the CS teams to like, understand those topics deeply enough. Also, it's 10 different CSMs. You know, we're tagging each of their own calls. They're just not going to do it the same way. I one person would do it, listening to those calls and having a consistent view across them with the context
of the problem or the challenge that you're trying to solve. I'll give you one
specific case that, um, is, yeah, it's been quite time intensive to, like, get to the bottom of so in in our industry, there's something called Cyber Risk quantification, and what that means is, when there's some some risk associated with, like, the potential for data breach, you quantify it in in dollar values. And so you say, this, this company, if this company had a breach, it would cost you one to $5 million if this dollars. If this company had a breach, it cost you 100k to 500k and that is a way to that's a derivative of like measures of security, and it's and it's an attractive feature because then it allows teacher to communicate to the business about in dollar values. And so our product does not have that feature other competitors do. So it comes up as a comparison point. However, there's another related concept called and so that, instead of that will sometimes be called the like cyber risk quantification, or like the financial measure of risk, there's a separate topic called financial risk, which you would know about, which is like you're done in Bradstreet credit rating. And both of these are relevant for cyber risk measurements and like cyber risk products, but they're technically like totally different things. One of them is the projected cost of a data breach, and the other is like the projected, you know, likelihood of this company becoming insolvent. And when prospects are talking about that, the prospects know what they're talking about, because they use these products. And our AES may or may not understand the different ways in which risk is financially quantified. Okay? And so we get and so they they get tuned to one of these, they think, oh, fuck, like we lose on cyber risk quantification. I'm going to ask about that. And then if I hear the word financial, that's what they're talking about. So I went through and listened to all these calls, and that's not what they're always talking about. Sometimes they're talking about these totally different things for totally different use cases and totally different implementations. That's a case where, with pm knowledge. Then you say, okay, here are the topics I'm interested in, and I would like to, actually, it's really important to me to be able to get measurements of demand for both of these separate features, right?
That's really interesting, and then super helpful for your product marketing people to then go and actually, you know, build content and educate salespeople on, yeah, and I'm
sure, like, I'm sure it's the same for your business, or any business where there are two or more things that sound similar, and you, you know, you make an effort to explain how you're different from x, or how two things are slightly different,
yeah. And more so when you're in kind of an IT, buyer, security, buyer, technical world like you guys are in, it's definitely the case interesting. So let me, let me paint a picture of what I where I think we're going to go and get your feedback on it. I'll be conscious of the time. We've only got a couple of minutes. So as I said, all these kind of conversational data comes in in so many different ways, so many different platforms. And ultimately what we find is that people are spending a ton of time going through Gong calls, ton of time going through on conversations, or relying on the compliance of other teams to try surface that stuff that works sometimes, sometimes doesn't. And so actually, a lot of the gap that has been created in this is because business intelligence tools can't handle unstructured data, right? And so if you want to know anything about your sales numbers, your you know PNL lines, your calls, you know your usage metrics, it's in some form of BI tool. But actually this level of intelligence is not it's scattered all over the place, and everybody has an individual interest and perspective when they want to look at that data, right? And so we try to capture it all with with capture all tagging for CSMs, but, but again, as you said, it's not localized to you and to what you care about. So trying to do two things. Firstly, just connect all the places that that these play, that these conversations, interactions with customers and prospects are happening. And then firstly, what we're thinking of doing is just building out really simplified reports, but that those reports could be curated to each person across an organization. So questions like, hey, what's all the feedback on this specific feature? Or like, Hey, how are salespeople positioning? You know, our cyber risk qualification feature that we don't have versus the financial risk that people are talking about, right and and so, you know, the very kind of bespoke interest areas and contexts that people have. How do you actually just mine all of the existing conversational data that you have and generating a simple report for people based on that? We've got a couple of, like, really, really early design partners on the system now, and they're using it to, like, just summarize all the gone calls from the last week, because we've had 200 and we can't possibly listen to all those and, you know, do it under these specific headings, and then those headings go to different people. So that's kind of version one of what we're thinking like. And it like, and it was a while to get here. We tried a whole bunch of more complex stuff. It's just like a really, really, really simplified reporting system for conversational data that can be curated to each individual. We could be really sophisticated about that in terms of connecting to JIRA, connecting to product board, understanding what you yourself are working on, and tailor it to you know, maybe we'll do all that shit, I don't know, but actually, you can just tell us by email, and we can just build it for you, and 72 every week. So that's where we're starting. I think beyond that, there's a whole bunch of possibilities in terms of, okay, do you actually start to open up some of these databases to people, to search, to build views, to build reports themselves? And do you actually platform it? I'm thinking, as a first instance, I would love people not have to ever use our system, right? They just tell us what they care about over email or slack. We build the reports, and they get them every Friday morning. Like, just start really, really simple, yeah. What are your What are your thoughts on that? And I asked that particularly, like, are there specific things that you wish you could see or know every week that you don't or that take a huge amount of effort to do.
I don't know that. Um, the I don't, yeah, so I don't know that, like the topics I'm interested in, I expect real change on sort of a weekly basis. That is, yeah, move, you know, a fairly large sales team. And so if, like, you're saying like, I would do an investigation into how they're positioning some feature whatever to to modify sales behavior, and then have that reflected, like, at a level where I see changes I don't expect, like weekly variation in that that's, you know, quite slow things. Yeah, interesting. I mean, overall, like, what you're saying makes sense. So yeah, like, yeah, I guess, like weekly sort of a weekly cadence, then, yeah, demand for things that is useful. The cyber risk quantification feature is one where it jumps up and down, where there's a huge deal depends on this. And so it's what everyone's talking about. And then it'll be silent for three weeks. And so, yeah, being able to kind of measure, yes, sustained I guess, yeah, I guess, in terms of, like, how our team is behaving, I don't expect weekly changes there, but in terms of measuring demand, or like trends in the market, on the other side that that's that would be useful, yeah, yes,
interesting, interesting. We hadn't thought of that angle on it. Yeah, because I think, I think for like, VPs and above, there's a whole bunch of pretty simple questions that are really, really hard to get the answers to. And typically they take days or week long investigations, be it by the person themselves, or they ask somebody else to do. Yeah. I
mean, this is exactly one of those things where one week, the sales VP will be screaming, we need this thing. We're dying without it. And then two weeks go by and then it doesn't come up again. And so you had this question of, is this really important to the business that is? Yeah, that's where, where longitudinal data would be useful.
Okay, okay, awesome. That was really helpful. I'm conscious of time you need to reboot your your computer, if possible. Cool, as I said, we're super we're super early. We're just figuring our way through the space and trying to speak to some folks to build up. I really appreciate you taking the time, whichever way we end up going. I'll let you know, but I don't know how it is, but I really, really appreciate your time and sharing those there's some really awesome examples there.
Yeah, no worries. You know, I used to, um, when people would, like, cold, reach out, and I'd be like, I'm not talking to you. Like, this is totally optional. I was like, You know what? It's so hard to start a business, like, like, and it's so hard to get anyone to answer. And I'm like, if anyone reaches out and wants 30 minutes to talk about what I my experience? Like, yeah, cool. Like,
yeah. Would you, would you ever do it yourself. Yeah, definitely, yeah, something, yeah. I
started thinking like, oh yeah. Like, it's, you know, you know, there's so much failure involved in in trying, in sales and, like, supposed to try to get something new, then, like, everyone in the world can ignore you. Yeah. So I was like, You know what I gotta, I gotta do the the right thing here and try and try and be the personnel on the other end.
Yeah, that's awesome. What I really appreciate it, it's, it's, I tried to do the same, but it's hard because time is precious. Yeah, Greg, thank you so much. It was really great to meet you. I'll keep you posted, but, yeah, awesome to Connect.
