Speaker 1: 0:23

Hello everyone, welcome to another episode of OpsCast brought to you by MarketingOpscom, powered by all the small pros out there. I'm your host, michael Hartman, joined today by nobody, at least not my co host. I do have a guest, so joining me today is Chris Golick, a serial entrepreneur who is currently founder and CEO of Channel 99, a company that leverages AI to help companies improve B2B marketing programs and invest in growth. Prior to Channel 99, chris has been active advising CEOs and marketing executives, investing in startups and supporting various philanthropic organizations. Previously, he founded Demandbase to transform B2B advertising, marketing and sales through innovations in digital technology and pioneered the ABM technology category. Prior to starting Demandbase, chris founded Supplybase to help global manufacturers collaborate with their supply chain to bring new products to market. And before becoming a software entrepreneur, chris held multiple sales marketing engineering positions with GE, dupont and GM. So, chris, thanks for joining us today.

Speaker 2: 1:25

Happy to be here, Michael.

Speaker 1: 1:27

All right, well, for our listeners, I nearly screwed it up too, so I'm glad we were able to make this happen. Get you on before Mopspalooza. We're recording this in early October 2024. So we're ready to go. All right, so lots of places we could start. You have a really interesting background, so why don't we start with this, though? When you and I spoke, you told me a little bit about the kind of origin story and history of Demandbase, and something that was interesting to me that I hadn't heard was that you started the company. When you started, it was initially not focused on ABM. Your initial focus was going to be on something else, and you ended up having to, not having to, but you ended up pivoting to ABM for B2B businesses. Why don't you? Maybe?

Speaker 2: 2:17

tell a little bit about that story. Sure, so you know, I've been a B2B guy for 30 years, since the late 90s, and back then people thought B2B was bed and breakfast. So you really had to educate people. And we started a software company me and a couple of guys from GE called Supply Base. It was all supply chain software and as a marketer at that company, I was always very frustrated because all the technologies were focused on consumer marketing. And so we ended up selling SupplyBase to a public company, which is now Blue Yonder, and after leaving that company I really wanted to look at how do we help marketers manage their supply chain.

Speaker 2: 3:03

So it was kind of like this supply chain idea for marketers. That's the initial concept of Demandbase, and it was just too early for that type of solution. So we kind of shifted Demandbase to focus on how do we fix B2B marketing, how do we build technologies to help B2B marketers reach the right accounts. That's what sales wants and that's what they should get right. And so that was launched in 2006, 2007. And it was a journey right of building ABM, creating the category, lots of different products and, you know, fast forward to today. It's, you know, it's, you know, a big business, lots of players, but the opportunity to kind of revisit that platform to really help marketers understands what's working, what's not working and how do I fix it. It still existed, and so that's why I started channel 99.

Speaker 1: 4:03

Oh, it's really interesting because I think my guess is, when you say it was before it's time is maybe like it was pre-COVID, right? I don't think anybody knew the term, other than people who are in supply chain roles. Right, knew that term until COVID. Then the whole world knew about it Exactly. My oldest son is in college and he's studying finance and supply chain. That's right. That's a good combination these days and hopefully we're not all going to be hearing that again. I haven't heard.

Speaker 2: 4:36

Is it the?

Speaker 1: 4:36

whole dock workers thing in the East Coast and Gulf Coast. Let's fingers crossed. We don't need that again. Run on toilet paper, okay, so one of the other things you and I talked about and I think is surprisingly, we haven't had a ton of conversation about it here on our podcast. I'm sure it's coming, though is AI and its potential impact on marketing teams and marketing ops teams is AI and its potential impact on marketing teams and marketing ops teams. So how do you expect things?

Speaker 2: 5:18

to be impacted within the marketing ops realm by AI technologies? That's a good question. You know, as marketers, we're all experiencing the opportunity to leverage AI to write blogs. Some are using it to create brand and creatives. It's pretty good at journalism not as good as humans. It's pretty good at some graphics, but not as good as humans. But the reality is, as you jump over into marketing ops, where you have vast amounts of data and interpretation, that's where I think the biggest opportunity for AI is to look at all that data and interpret questions of you know, where should I spend more or less, or how do I improve this campaign. I think that's a huge opportunity and that's where we're focused, frankly. But you know there's a lot of things that happen have to happen first. You just don't plop AI on top of a data set and hope you're going to get all the answers.

Speaker 1: 6:19

Right. Yeah, it's funny because I would put myself squarely in the camp of skeptic maybe six to nine months ago. Squarely in the camp of skeptic maybe six to nine months ago, um, and I think it was getting so much hype about and at the end of the day it was chat gpt was was the driver for the majority of what was coming out um, that it was going to be a boon for content, or maybe it may be, you know, a negative for those people who are in content, writing content. My sense is that that was an initial wave that sounded really good, but then when people really understood under the hood like what were the potential pitfalls, like risk of exposing IP and things like that that that would sort of slow down. But I've thought, similar to you, that for a little while that the the amount of data, like one of the one of the limit.

Speaker 1: 7:15

There's at least two limitations I've seen with the data that marketing ops and marketing teams have generated over the years.

Speaker 1: 7:21

Which is significant is that I think there's a gap in the skillset right. I don't know that there's a lot of people within the profession that really have deep or not even deep, but enough understanding of analysis, statistics, things like that, to be able to really do the kind of deep dive analysis that would be, I think, useful for business. And then, second, even if they did, there's limits on the time they could apply to it, so they would have to go down the path of choosing scenarios or hypotheses and evaluating those, as opposed to spending time digging through the data to see if there's additional hypotheses or patterns that could be exposed. Do you have a similar take on that? Is that why that the data is there? The ability to really find, sort of discover patterns and things like that is really, I think, what I'm expecting to get out the benefit I expect to get out of AI-driven technologies. Think what I'm expecting to get out the benefit I expect to get out of ai driven kind of technologies yeah, I think.

Speaker 2: 8:32

Well, there's. There's data challenges that I think have to be solved first for ai to be efficient and effective. Um, some of those relate to you. Know what are all the account signals that are happening both on my website and off my website across the web? Um, you know that direct channel that's coming to your website? What is the true source that's coming in? Because, to me, any attribution system or any ai technology, if you're ignoring the largest source of traffic, you don't have the right, you're not going to get the right answer. And so, to me, solving for those things first is imperative to having a quality output from the AI. And I think, to your, to your question, marketers want the answer, they want to be told what to do, but they also want to understand why. Why is the AI making this recommendation? And is it just based on my data? Is it based on industry data? What are some of the inputs? So that's really important to integrate into any AI-based solution.

Speaker 1: 9:44

So, exposing what's happening in the black box based solution, right.

Speaker 2: 9:46

So exposing what's happening in the black box, yeah, and you know the term co-pilot is probably the most overused term in AI land, but it is right. I mean people want, you know, this tool to give them insights and information at scale that's very insightful that they maybe couldn't do on their own. But at the end of the day, they want to shape that response and so making that all possible is really key.

Speaker 1: 10:13

Yeah, I don't know. I think I know some marketers who don't want to be given an answer because they don't like what the answer is, telling them that they should be doing right.

Speaker 2: 10:25

They'd rather go with instinct a little bit. Yeah, one of the things I've learned, michael, just early in this journey, is that, um, a lot of marketers don't want to be, don't want to hear if something's not working, but they're happy to learn how to make improvements right, like yeah, it's almost the same thing, but you're kind of inverting it to make improvements Right, it's almost the same thing but you're kind of inverting it to make it a positive improvement, versus you're wasting 80% of your money. Who wants to hear that?

Speaker 1: 10:59

Right, nobody wants to hear their baby's ugly.

Speaker 2: 11:04

Exactly, exactly.

Speaker 1: 11:07

The first exposure I had to that cause. My experience in marketing has really been more on the technical side, you know, operation side, historically, even when I was doing some B2B or B2C stuff. But I remember I had a friend who at the time was a well, a friend of my wife, who was a, was the become the cmo at um, a large um retailer that we'd, uh, you find on corners of lots of places in the country. I'll leave it at that. You can probably figure out who it is. But I remember talking to her one time about what I saw at the time.

Speaker 1: 11:45

This was when first windlakes, paid search was coming around and web analytics were really coming into their own and I was like as, as somebody who was like, really keen on data and insights and using that to inform and make you know, understand what's working, not working. So I think it's going to be really transformational and useful and it was interesting to get her her inputs Like I think she still felt like I think her fear was that that was interesting to get her inputs. I think she still felt like I think her fear was that that was going to suck all the actual creativity and risk-taking that might come with being in marketing, and so I actually think there's a balance to strike there. Right, you need the data, right, because you should want to know what's going to work, but at the same time, I feel like there's still you need to leave room for experimenting with stuff that may not even be measurable.

Speaker 2: 12:38

Sure. Well, and I would also say there's one of the things that we're finding. You know, there's been a lot of solutions out there that aggregate data and put it into some nice charts and then that's the offer, but they're not solving the problems of understanding why channel sometimes is merely it's all about who you're marketing or selling to Sure the wrong vendors for the intended audience or the audience that they've created is not. It doesn't fit all these different vendors and they should really segment it into three. And all this stuff I'm seeing over and over again from customer to customer, and all this stuff is trainable in AI to make recommendations and give specific actions to drive the efficiency of all your vendors and channels. It's not about just killing this vendor because the numbers look bad. It's like why do they look bad first?

Speaker 1: 13:59

Yeah, I think that's good. I think the best marketers out there will be keen to understand what's working and not working, and I like that point about why. So, with Channel 99, you're going back to this challenge of marketing analytics In terms of like, was there a particular trigger that led to you go like, oh, it's time to solve this problem, and I'm assuming there's, like you saw some sort of opportunity for it. Like, what was the origin story of that then?

Speaker 2: 14:31

yeah, you know a lot of it was just. You know, I was still a running demand base and I worked with a lot of different CMOs and a lot of big enterprises and they all shared similar frustration there. You know, they have not just demand base as a vendor but of course lots of others, and everybody has their own KPIs. It was hard to make sense of. You know different KPIs, including demand bases, and I'm like gosh, you know. So I was just polling a bunch of people Like what if there was a platform with that allowed you a true apples to apples comparison and did it in a way that it was more financially oriented?

Speaker 2: 15:14

that tell you how to invest. And, believe me, I'm like CFOs. When I talked to them, they're like this is beautiful, like I don't care about MQLs. I really don't. I just want to know how do we generate opportunities and what does it cost per opportunity because I can plan around that. I can't plan around mqls because everybody keeps changing the definitions and so, um, that's when I really said you know, the market is now ready and I want to um, ignore some of the, the learnings from kind this, this world of attribution that I sometimes refer to as legacy attribution, that just relates, relies on visits and touches, and you know some people are paid on that. So I get it, but it's not predictive and I, you know, very intentionally hired the co-founder of visible as our CTO, knowing that he probably shared some of the same frustrations as me, kind of building that solution.

Speaker 1: 16:13

Yeah, I can imagine. Yeah, it's interesting, even in places where I've been, where I'd say across what would be the equivalent of revenue operations, right, so marketing and sales operations and the leadership there, there was a fairly I think people were fairly comfortable with the definitions of the different MQL sales accepted, lead and so on. Sales qualified all the way down to pipeline opportunity. There were still nuances at this company that were like even I it took it was like something, I'm missing something here, right, right. So there was always, even on a regular basis, going back and going like how is this defined again? And how is like how does it go from one stage to another? Because, um, in in one, and there was one particular sort of transition point in where literally nothing changed for a couple weeks because somebody was on vacation and I and I was like what, wait a minute, this is like someone's manually reviewing these one at a time and us in making a determination this one person, no backup like that's crazy yeah, yeah.

Speaker 2: 17:28

Well, michael, one of the things that you know not to go back backward in time, but I was always frustrated as a ceo at demand base. You know we had sales operations and marketing operations and operations and finance. Now I'd ask each person how many customers do we have? Like the most difficult question to answer at any company I've been well, I would get three different answers, but the worst answer was always well. What do you mean by that?

Speaker 1: 19:00

that's, but that's what I say. Like that I I've said this before right, like when, when, when I've been asked, uh, as an ops person, to go like we want it, we want to send all our customers an email, I always sort of like I get this. Like my body tenses up because I know what I have to ask is like, well, how do we want to define that? When you say customer, do you want so, say we've got a company and in our database we've got 100 people, do you want all 100 people to get it, or do you just want one person or a handful of people? Like, how do we determine that? That's if we can get down to the one company, right, usually, you also have examples of the obvious, like the IBMs right, where you have different variations of the name or something, and it is incredibly hard to answer that question at way too many places.

Speaker 1: 19:50

I've only been one place where they actually put a lot of effort into cross-functionally right Sales, marketing, finance and I call it business operations collectively to come up with common IT common definitions, common and it common definitions. And so when they, when they were pulling data, it would have been essentially like a cdp right, but mostly driven by financial data. There was a lot of effort put into like okay, this is how we define a customer. You know, and I don't know that it was the. It's not necessarily a right definition, but it was one that everyone agreed to and understood what that definition was, and that was more important than the particulars of the definition.

Speaker 2: 20:32

Yeah.

Speaker 1: 20:33

Yeah, so sorry, off on a tangent, because as soon as you said that I had that same reaction like oh, I don't want to have that question. Yeah, so Channel 99, you've described it to me as AI driven marketing investment decision platform. Break that down. What does that mean?

Speaker 2: 20:58

examples. We wanted a system that could make recommendations not only about where to invest, but also where not to invest, but also how do I make improvements and not just give the answers but show people why, based on results and their data and industry benchmarks and things like that, and their data and industry benchmarks and things like that. And so you know, if I want to increase my spend in LinkedIn 10K per month, what are the outcomes I can expect?

Speaker 1: 21:39

So there's an element of predicting as well, yeah.

Speaker 2: 21:43

Like scenario building.

Speaker 2: 21:45

Exactly no-transcript to make predictive outcomes, but it's not going to be based on visits, it's going to be based on a like the last six months, when you change this, this was the outcome, but you got to be able to do that across all vendors, all channels or saturation curves. You know it gets a little bit more complex. The other thing I would I would say is you know, we really had to solve this world of the direct traffic and understanding what is the true source of that traffic. Is it organic, social? Is it ads that people don't click on but they come to my website direct, or is it the content, syndication, all this investment that's going in there?

Speaker 2: 22:57

To me, if you don't unveil that source, all your numbers, all your dollar per visits and all these things are going to be wrong. And you know, we've, we've shown that like view through attribution is six or seven times that of click through, and so that effectively lowers your dollar per visit for a particular vendor by like 75%, and so this is going to give you a totally different answer unless you solve for that. And so that's one piece. Um, you know, the how we do predictive is another, and building a really robust account identification engine was another piece that we invested in early Um cause I I knew that that would be critical and be able to do it without third party cookies and things.

Speaker 1: 23:48

A couple of clarification things for me. So when I think you alluded to it Maybe I'm misunderstood, but you talked about traditional or most attribution models today are not how they predict is different. My view is I don't think any of them are predictive. Most of them are backwards-looking, as opposed to, as opposed to like no predictive nature at all. Okay, so that's interesting.

Speaker 1: 24:17

And then the other is like what I, when I hear you describing this, like you didn't even mention ai, really, right, so, um, ai is just sort of a tool that enables some of this is the way I'm hearing it. But, yeah, I, I, I think of, I was going to say scientific method, but really I think it's more like economic analysis, right? So you're, you're not doing cause. I think of scientific method. You basically control everything except for one variable, more or less right, what AB testing should be. But what, if I understand right, economic analysis is more about like, you've got a whole bunch of variables that have variability and, um, you're trying to find, like, what are the ones that actually are? Are predictive or influence on whatever the outcome is you're trying to do by trying to eliminate the impact of the others. Is that kind of what you're talking about doing?

Speaker 2: 25:13

Exactly exactly. So. Imagine a world where you're a platform that's connected to all your media platforms so Google, reddit, microsoft, all these different vendors and channels and it's constantly logging changes of spend and and your crm system and changes of business outcome. So this huge messy database, sure, but then you layer ai on top of it. It can then trigger alerts around performance. It can answer questions like hey, what were the three changes we made back in Q2 that had the biggest pipeline impact in Q4? Like that would take somebody like you know, that's a science project and so that's all.

Speaker 1: 26:01

Well, my experience with that would be is like, even if they could do it once, the immediate thing would be hey, there's something in here, Could you go like dig into that particular piece of it, right? So it wouldn't be just a one-time thing. It would be like it would take two, three, four iterations to even get to something. Where who has time to do that?

Speaker 2: 26:21

Yeah, and so the world of sucking in all this data from all these vendors unifying in an account-ric world based on actual results, not what they're telling you necessarily because, they're always going to be different.

Speaker 2: 26:37

um, you're going to get a totally different answer and you're right on your assessment of, you know, ai is the thing you kind of layer on top of all this once, once you get it together, and that's why I, like I have this whole like five step thing that we've gone through. Step five is where we're going with the AI and, you know, think of training AI based on white marketing, white papers, but also like playbooks, like how do I improve my LinkedIn campaign. We've we're training it on all these exercises. Okay, my LinkedIn campaign.

Speaker 1: 27:09

We were training on all these exercises. Oh, okay.

Speaker 2: 27:10

And think of it as we build our community of customers. Think of the AI getting smarter and smarter and smarter. It won't be perfect at first. We won't cover every kind of campaign improvement possible at first, but it's getting's getting really, really great. Because I find myself, you know people naturally ask me hey, chris, like I know you know, you worked at demand-based. Like we have this demand-based campaign is any ideas on how to make it better? And I find myself going through the same four or five steps in their instance of Channel 99 and show them. Like you know, I would take these 120 companies out of the audience and use Facebook ads, because they're generally too small for demand-based outreach. And then Facebook gets better and demand-based gets better and they didn't spend any more money, right? So there's a lot of things like that that are surfacing and you know we'll be diving into paid search and, of course, LinkedIn we mentioned a few times but there's, you know, a lot of other platforms and programs.

Speaker 1: 28:19

Got it, you also. So you also mentioned that you've one of the things you can, you I can't, couldn't tell if you can do or you are doing is pulling in some sort of like benchmarks, like we're like what are you, how are you sourcing that? Is that going to be based on performance of things through your customer base? Is it through third party stuff? Was a combination like what is that? What's that from?

Speaker 2: 28:43

yeah, so think of, um, there's a bunch of ways we're doing it, certainly like KPIs and averages, but think of also like 95th percentile and like you shouldn't make investment decisions hoping that you're going to be better than the 95th percentile. So there's, there's always like saturation curves by different vendors, where the incremental benefit gets can get less and less. You may not sure at that point, but all that stuff, most companies don't have enough data to understand where those points are, and so if we can kind of help build those through the, the collective of customers, in a very anonymous way, it just gets better and better, and so it's a pretty exciting. You know pretty heavy math, as I think I mentioned to you kind of jokingly. It's like this is more calculus, this is not algebra, right?

Speaker 1: 29:41

Well, and I think you know one of the other things is that it's popped in my head.

Speaker 1: 29:46

It actually ties back to the sort of calculus kind of um model as well.

Speaker 1: 29:51

Is or thought process is, um, yeah, attribution, traditional attribution modeling is looking for as much like trying to build as much kind of a almost a direct linkage between actions activities that a customer prospect took to specific opportunities, right, whereas I think what I'm hearing you is you're not trying to build that linkage right, you're looking more at the outcomes, what are the inputs and activities kind of as well, and saying, based on these activities or combination of activities that we did and how customers reacted to it, the downstream impact was this kind of revenue impact or pipeline impact, which can kind of and one of the things that I think is a challenge, especially at B2B companies where they have a long sales cycle. It's complicated and it's a high value transaction, right the time lag between some of the activities to when you see that the downstream benefit is really long and it tests patients that I think most places don't have. Are you like, is this helping to solve for some of that and kind of break?

Speaker 1: 31:15

that need to do that direct level of tracking.

Speaker 2: 31:19

Exactly the, and it's important for me to say that, like I'm not super anti like first touch, last touch, all that I mean we will. We will have that data for people, because some people, frankly, are paid on it and so they need to understand it. Um, all I'm suggesting is that, like, as you agreed, it's not predictive um and so, um, while we'll have some of those models, I just don't know the benefit other than kind of understanding, like what really initiated this deal is. Is it, was it a sales outreach or you know, was it, uh, an email or was it uh, you know, they saw an ad and then or they came to it.

Speaker 1: 32:02

they came to a little dinner we did two years ago exactly, yeah, right, um, so keep going, sorry.

Speaker 2: 32:12

No, no, that's okay, it's. You know, sometimes how you present the data. People can see the patterns and understand oh, this is what's happening with all the things that move into pipe quickly. Well, the AI doesn't need to see the patterns visually. They can interpret the patterns in the data. They can interpret the patterns in the data. And so there's a lot to be learned. And there's still a lot of companies that battle over it was sales generated or marketing generated, and I think any CEO would be frustrated over those battles, right?

Speaker 1: 32:49

Because there's a role for both.

Speaker 1: 32:54

I would think sales leaders and marketing leaders should be frustrated with it. Like I'm a big believer, like, at the end of the day, like we should all be focused on getting the most value for the company, regardless of what our role is, and sometimes that means you need to give on something while your colleague does, and sometimes that means you need to give on something while your colleague does. Yeah, it's interesting because I just talked to somebody yesterday, a younger person in his career, who was asking for some kind of mentoring stuff, and we talked about attribution, because I think my position about attribution is fairly well known, but I'm not against it either. Attribution's fairly well known, but like, I'm not against it either, right, but I also wouldn't if I was to, I would challenge leadership if they came to me and said we need to get attribution reporting, because I'm like I don't think you really do Right. I think there are other ways you can demonstrate value. I think there's other, better ways, like the opportunity cost of investing a bunch in all that may or may not be useful, especially given all the limitations we know about.

Speaker 1: 33:55

At the same time, I think there's value in it if it's used in a way that makes sense with an audience it makes sense, right which is not the CEO and the C-suite or the board, right? That is not. They don't care, to your point, right? The CFO wants to know, like, are we going to generate more pipeline and revenue? That's what they care about.

Speaker 1: 34:17

So again, I could go on a rant about that, probably for a few minutes if I was let loose. All right, so okay. So so far, like I don't know that any of this kind of stuff about the limitations or challenges with attribution modeling would be a surprise to our audience and marketing ops. What I suspect is many of them are maybe not as familiar with things like economic analysis and the things that go along with that kind of what you're describing is you're doing with Channel 99. How can our audience listen like, think about this stuff and apply to? Well, I mean so, like, let's be honest, like a lot of marketing ops folks, I think feel like they're not considered a strategic partner to the business, let alone just like marketing in general, and so how can they use this kind of idea and insight and maybe help elevate their perception and add value to the company and be seen as strategic and start getting into those conversations with leadership?

Speaker 2: 35:32

That's a great, great question. I think it's different depending on the scale and sophistication of the company, because I think when you get to kind of larger enterprises, marketing ops is a lot more sophisticated and strategic because a lot of the blocking and tackling around all the data and unifying has been put in place. And so I do think you know, with you know, the advent of AI, I think it'll make marketing ops a lot more efficient and productive and have a lot more. They'll have a lot more insights into making recommendations on what to improve and at the end of the day, you know, hopefully the marketing performance increases, right, like my cost per pipeline opportunity goes down, my customer acquisition cost is lowered. I mean, that's what everybody should be driving to.

Speaker 2: 36:29

And I got to believe, you know, some of the more mid-market companies where marketing ops is spending an inordinate amount of time like joining data and all this and then they present to the CMO and they want it sliced a different way and it's back to the drawing board, right, and I'm sure that's not very fun to be in that position versus. You know, you're coming to the meeting with a lot of great insights on what we should do next. Here are my top three recommendations for lowering customer acquisition cost. From a marketing influence standpoint, that stuff is all possible, and so I think we're at the just very beginning stages of this.

Speaker 1: 37:12

Yeah, I mean to me, I just heard two things that I think I want to make sure you tell me if I'm way off base here so I don't mislead our audience, but I felt like I heard two things that were implicit in what you said that our audience should be thinking about. One is understanding what the. I'll do it in two parts Right, understanding how the, how your company makes money, right, and then what are the drivers? What are the metrics that are drivers behind that? You mentioned cost per acquisition, cost per opportunity created. Like what are those metrics that are actually drivers for the business? So you need to understand those.

Speaker 1: 37:50

As opposed to visits, views, mql those may be important too, but these other ones, when you're talking to other people outside of marketing, that you need to understand what they care about. That's the first part. The second part is, I think, then, regardless of you using attribution, you're doing your own analysis or using a tool like Channel 99, you need to be able to understand what the data is telling so that you can interpret it in ways that you can then produce useful recommendations or insights. So those are like the two things. Like to me, that becomes down to understanding at least some basics about statistics and data analysis. Am I like? Am I off base?

Speaker 2: 38:34

no, no, you're totally right. But I can guarantee you, if you go to the CFO and you say, hey, out of our customer acquisition costs, 50% is pipeline acquisition cost, he or she is going to listen and then the next question is going to be like how do we fix that? That's way too high. They're going to say it's too high, even if it's average for most companies. And, um, just in the world of b2b there is just a lot of opportunity.

Speaker 2: 39:04

B2b marketing like I always say this as part of a subsidy it's hard, like most companies are only trying to reach three or 4% of the companies out there to buy their product or service and most of the technologies being used are still kind of consumer tech and so learning where to tweak the dial so you can really get it more pinpointed is is really important and there's just a lot of opportunity to keep on continuous improvements. And I will say that for paid search. But I'll also say it you know, abm is supposed to be a hundred percent focused on target accounts. It's not, it's not where it needs to be, and I also kind of say ABM, you know, is it's, it's a, it's a thought process, not necessarily just a technology, because some of the principles you should extend to all your vendors and channels right.

Speaker 1: 40:00

Yeah, yeah, I agree. Okay, so maybe a couple more, one or so more questions and we can wrap up. But you kind of mentioned, well, maybe that there might be some, you know, additional capabilities, that sort of AI driven capabilities that you're expecting to come from, coming to the marketplace and in channel 99, what kind of what are some of those things that you're thinking about that are going to be coming down the road?

Speaker 2: 40:38

the road. Yeah, so I think it's. You know the sophistication and how good the llms are kind of trained to give you the answers and recommendations. Um, and the more people that come into the community and just more education, it gets better and better. So everybody kind of benefits as a whole.

Speaker 2: 40:52

Um, and you know, the more we extend our technology to different channels you know whether that's you know Reddit or Twitter and things like that you know it'll learn how to make those things better. And so you know, I always look at this, I always come back to it depends who you're marketing or selling to. And if you break that down into a single account, if the AI can learn what's the most effective vendors and channels and mix to market to a single account, right, then you have a world where you could upload a list of, let's say, 112 accounts and have it tell you exactly. You know what might be the most effective ways to reach and engage those companies. And of course, this is all nuanced by you know, do you have a high ASP product or low and what in it? You know your industry and things like that.

Speaker 1: 42:08

Yeah, you know, sort of um queries into data that are like based on the way we would actually ask questions, as opposed to here's how I'm gonna have to do an sql and join data and all that right, it's gonna, you're gonna ask it a open-ended question and it's gonna be on the scenes.

Speaker 2: 42:27

Do some of that, do some of that right, and then it's up to you to again be able to understand and interpret the output yeah, that's right, you know, and that's interesting that you bring it up that way, because it was probably about eight months to a year ago. Um, the engineers were describing to me that you know, we're just training the llms to learn how to query the data and I'm like, wait, let me understand a little bit more. And then, so as I learn more about that, we started writing kind of the playbooks from which the LLMs learn from and then query the data and come back with the answers and really, really powerful and it's just going to get better and better and really, really powerful and it's just going to get better and better. But, as I said earlier, if you don't have the underlying data kind of connected right and you haven't solved for this direct channel and you're only identifying 20% of your traffic, like more data is always better.

Speaker 1: 43:30

Right. Well, I would say cleaner data is always better, but I always tell people, don't wait to start doing some of these things until your data is quote clean, because it will never be as clean as you want it, and especially in the B2B world, it's just by the nature of all the sort of hands are in the middle of it, the challenges with the technologies out there. It's just never going to be quote right. Um, so don't don't ever expect it to be so, because if that's what's holding you back right, then you'll never get to it, because you'll never get to that point where you feel like it's right enough. So, um, this has been, it's been interesting. Is there anything that we hadn't touched on yet that you want to make sure that our audience heard about?

Speaker 2: 44:19

Well, I just think you know, especially today in the market, you have to be able to do all this stuff and make it super, super easy for people to get started, like, how do I just get started and experience the platform without spending any money and really low-touch self-serve configuration. And I think you know, certainly we built our solution that way and I think a lot of the you know the kind of companies in the marketing analytics space at least the newer ones are pretty easy to get started. So that's refreshing that you know some of the older kind of solutions out there. That's not the case.

Speaker 1: 44:54

So yeah, I would encourage anybody just to get started because it's pretty low friction yeah, I'm, I'm, I'm excited about technologies like channel 99 and some of the others that are coming online that are helping with analytics and insights, because I think there's been a lot of limitations on it, as we've talked about throughout here, and it's why I'm actually encouraged by where I think AI is going to take all this. So, chris, thank you for sharing. This has been really interesting. I always love hearing these stories. If folks want to keep up with you or learn more about what you're doing, or what about Channel 99, what's the best way for them to do that?

Speaker 2: 45:39

Well, certainly I'm on LinkedIn. They connect. We're going to be at the Mopsapalooza event hosting a dinner, so like we'll be doing a Mops leadership activity. So love to meet anybody while they're down there and I really enjoyed the conversation, michael.

Speaker 1: 45:58

Yeah, I appreciate it. Well, thank you. We'll make sure that we share that info once this gets out there. So again, thanks to Chris for joining me and sharing all this info, and thanks to our audience for continuing to support us and provide your input and suggestions. As always, if you are interested in being guest or have an idea for a guest or topic, feel free to share with mike, naomi and me. And until next time, we'll talk to you later. Bye, everyone.