The new fuel for your GTM: dark social content & identity resolution with Kevin White, Head of GTM Strategy at Common Room

Introduction 5-10 years ago, the user journey was very linear. If someone is researching your product, they're landing on your website, then they'll fill out a form. Since there's been this trend away from someone revealing themselves within the product and then moving to these dark social channels, that makes it really hard for the traditional way of doing marketing or go to market. And that is an area of opportunity because if you're in any go to market team, you want to meet people where they are and engage with them on those channels. They're just in this channel that is hard to traditionally capture interest and intent. Those signals are now trackable in the dark funnel like they never were before. 1 interesting take here is that marketing's new role could be more about how does marketing help generate signals happening on LinkedIn or happening on Twitter or some other dark social channel. It's this new channel to identify qualified leads. Speaker 2 Welcome to the GTM Engineer, where we share the hidden stories, tactics, and mental models behind the rise of AI and sales, marketing and operations. I'm your host, Noah Adelstein. This episode was a fun conversation with Kevin White. Kevin runs GTM strategy at Common Room, which is a software tool helping GTM teams find intense signals, track users across different platforms, and turn that data into full funnel campaigns. Before Common Room, Kevin was Head of Marketing at Retool and before that he worked at Segment. In this conversation, we talked about why data is the most important differentiator in your cold outbound. What differentiates the best market orgs that Kevin works with at Common Room and how to build and evaluate GTM engineering skills? I found Kevin's view about the increasing importance of dark social content and identity resolution to be innovative and refreshing. He and the team at Common Room spent a lot of time thinking about how to find and harvest demand in the places where their prospects already exist and interact. Whether it's LinkedIn, GitHub, Sub Stack, or anywhere else, They think deeply about how to tie those social profiles back to specific prospects, which helps level up the entire cold outbound in intent motion for their customers. If you want to check out more content from the GTM Engineer, follow us on Spotify and subscribe to our Sub Stack. In the show notes, you'll also find a link to join the GTM Engineer Lab, a collection of GTM and aspiring GTM engineers that receive custom content, opportunities to interact with others in the lab, and much more to come. With that, here's Kevin. Hope you learned something useful. Speaker 3 Kevin, welcome to the GTM Engineer. It's great to have you here and thanks for making the time. Speaker 1 Awesome. Yeah, thanks for having me. Excited to get into things. Speaker 3 I've seen common room in the news a lot and I know you guys work with a lot of high growth and and successful companies. And then I've also seen a lot of your personal LinkedIn posts about how you think about GTM, so really looking forward to the conversation. Speaker 1 Yeah, I feel like the, it's funny you say the news because I don't know if we've had any articles published, but I guess the news is now the LinkedIn news feed. So good that we're showing up there. I feel like that's our main distribution channel right now. Speaker 3 Yeah, yeah. Seeing you guys all over LinkedIn, maybe we just start. Can you share a little bit about your background and what led you to Common Room A? Speaker 1 Lot of where I cut my teeth is in SAS and specifically starting off in the marketing OPS. When I actually started it was like 2010 and I'm like dating myself here, but it was like I joined this company, it's called Gig, yet actually got acquired by SAP long, long time ago. But at that time they were like, hey, we bought this thing called Marketo. Can you just come in and like figure out how it works? And I was like learning how to do marketing operations via Marketo at the time. But I think as I progress my career, whatever like leading growth segments was like a huge learning for me and then also leading marketing at retool. And so, you know, through those experience, I just have seen a lot of different pains, a lot of different models like product LED model or sales LED model, all this kind of stuff. And that's actually what drew me to joining common room is just pattern matching and having just empathy for what's happening for marketing and go to market teams and sales teams and the ability to do some of these things that we were kind of like hacking together in my segment days. And then seeing common room at the platform as a vehicle for making that just a lot easier. And I feel like the dream of a marketer or like a go to market engineer is like, how can I have my team, my people on the front lines reach the right people with the right message at the right time. And we've we have promised over promised that a lot on the especially on the like Mark Martech front, probably guilty as charged here. But I do see that as the market is developing like we are getting like actually pretty close to that. And common room is a platform for doing that. And then another part of the reason for me joining and why I have comfort in a lot of putting myself out there on LinkedIn is just because I have all this like I'm using the product, I have all the empathy built in and I just really feel comfortable sharing about what's possible now. And it also excites me. And that is a big reason for joining the company is that I could see myself as a subject matter expert that can share about this. And that just goes like a, a really, really long way. If you're joining a company where you you get it, you get the problem, you get the pain point, you can speak to the audience. I feel like that has been a really big contributor to my happiness and I'm glad that I joined the common room so that I could do that versus working out with data engineering company, which I've had in my past, which I can do, but it's not like I'm the person that feels that pain so much, if that makes sense. Speaker 3 Can you just give a quick overview of what common room does? What Common Room does We view common room as a platform for pipe Gen., and when you think about that platform, there's three different pillars to it. One is the ability to capture buying signals and capture data from any different digital breadcrumb. There is everything from website visits to job changes to product data to open source data. We're essentially building this platform for capturing any buying signal, which on its own is not super interesting. It's just a signal I think are going to be commoditized, but everyone's going to need signals in one place. The second pillar of our platform is this identity resolution and waterfall enrichment layer, which we call it Person 360. It's essentially like taking all those signals, unifying them into a single unified profile, and then filling in all the blanks that you would need for prioritization like their person's name, their title, company funding, all this kind of stuff. And then if you think about when you combine those two things, you can run all these different types of outbound plays on top of it that are can be really hyper sophisticated and high converting. And so that's like the last pillar is the action layer. So it's combining the signals with the prioritization of who you know about this person or their journey or the company and then running an outbound player outbound motion as the final pillar. So that's like the execution side of things. And then if you think about that as a complete like data foundation, like AI running on top of that and generating emails or doing account research for signals like becomes the interesting thing too. So that is a somewhat long winded way of describing like all that we do with Cumber, OK. Speaker 3 Go. So first, finding and aggregating data and intent signal, second, building a single identity to track all these data intent signals, and then third, figuring out what to actually do with all of them and run campaigns. Speaker 1 Yeah, exactly, exactly. You got it. Speaker 3 You worked at these two high growth companies that segment and retool experience all these pain points probably directly saw a common room and you made the decision to go spend time there actually evangelizing and working on the thing that. Speaker 1 Solving that actual pain. Yeah, yeah, yeah. Speaker 3 Yeah, that makes sense. I remember reading one of your LinkedIn posts that talked about how you think about data and this sort of usage and aggregation of data as the primary differentiator for go to market teams as opposed to things other things happening in AI like AI avatars or AI videos and images. Data is the Oil for AI Would love to dig into that a little bit. Speaker 1 I do think, you know, as worrying as it sounds, the data that you capture and the data infrastructure layer that you have that go to market is built on top of is actually your, it's like your own proprietary data. It's your own product data, it's your own website, but it's it's your own, you know, people engaging with you on your own social channel, things like that. Data that you own becomes like a pretty big differentiator in a world where, you know, data and 3rd party data sources are pretty publicly accessible. It sounds cliche, but like data is the oil for AI or for these signal based go to market to work the way it should work. I think the biggest like stark contrast in what's not working and why data is such a big differentiator is if you look at the AISDR market, which, you know, has a ton of promise and I think it'll actually catch up and get there. But that market has, you know, went through this hype cycle where everyone's like, oh, this is great, we can replace humans. AISDR agents are going to do the same thing that you would hire like a predictable revenue SDR team to run and they're just going to do it better and for cheaper. And the problem with that premise was just like the data that is being used, the data that's trading those STR that it that it that the AI sitting on top of is pretty commoditized. It's looking at a LinkedIn profile or like doing a Google search for this company. But then outside of that, the different things I talked about, like what's their product? What's their product usage? What's who's changed jobs, all that stuff it doesn't have access to. And so like it's just going through this commoditized surface level personalization and then doing that at scale and then sending a ton of emails, the people who are the recipient and everything sounds the same. And the law of like bad click throughs where it's just like we, we kind of like ruin that Channel or that Channel is ruined. And so therefore, I feel like if you, if you do have the right kind of data that's fueling that, that kind of outbound motion that's bespoke to not only what data you have, but also like how you can connect that data to the pinpoint of the end user. That's where we actually see people say like, Oh yeah, I, I, I feel that that's my pinpoint. I'm going to, I'm going to take an action, I'm going to reply versus this commoditized personalized thing that's very like service level. So I think the market will catch up and we'll get there. But I do think that's like a that's a good example of like how data is so critical and people don't think about it as much. Speaker 3 Yeah. And so there, there's one piece in there that's like a lot of this data is commoditized at this point, assuming you're a business that has the means either financially or internally with people that know how to go get it. How to stand out in the data world And so then given that, how do you think about what, what, what can actually help you stand out on top of just having access to a lot of these classic intent signals? Speaker 1 Yeah. The thing that I'm always looking for is, OK, your product solves for some pain points. And then I like to ask, OK, what is the pain points that your core customer, what puts them in market for like having that pain and signal like a job change or funding or anything like this? This is commoditized. But if that signal means something like if someone changes jobs and they were formerly using your product, that means something. It's like, oh, they might want to take that product with them to their new company. So that's where I'm like, OK, if you're looking for these and then maybe that's not solving a pain point, but it is getting to the point of, OK, this is the specific thing of the specific reason. That's your own proprietary data and you're connecting that too. Oh, this person might also have this pain. That's what I'm always looking for. And it's not that it's the data could be somewhat commoditized or the signal could be somewhat commoditized. But if you could connect it to a specific reason, a specific thing that your product can help with or and alleviate that pain, that is the thing that's differentiated. And so I'm always looking for that kind of nugget when I'm doing advisory or helping our customers out is what is the trigger that's putting that company in market And can we go, we probably can find the easy way to go find that signal, go fetch that signal, return it and then use that to our advantage. And that is the missing piece of just like generic commoditized personalization versus personalization that matters. I don't even call it personalization. I like to call it context or relevance versus personalization because it's like really connecting to you like a business need or a pain point versus just saying, hey, we went to the same alma mater, whatever. Speaker 3 But then how do you think about the levels of this where it's anybody that changes jobs, you could reach out and and say, hey, you're in a new job, you're probably looking for a new software, our product can help you like that, that kind of does that or like funding, hey, you just got a bunch more money you're trying to grow like our product can probably help you. So I guess it's how do you decide what is enough and meets this unique threshold when you can fabricate this? Yeah, like it's this pain and it ties to our product. Speaker 1 Yeah, I do think that those signals that you mentioned or those triggers that you mentioned, if something like that happens, you can guarantee that they're going to receive a lot of the same people reaching out with the same kind of message. And I would argue that signal on itself is maybe not enough to make a even reach out. You maybe want to add on more layers to that. They just got a new job and like they put opened up like job, like usually when someone changes jobs and then you want to reach out to them, they're a manager. So they're like are opening other job recs. And so maybe that thing combined with, OK, this person just got a new job, they're a director and they're hiring A-Team. And then the job listings that they have for that team have certain skill sets or certain things. And like those two combinations are what put them in market for your products. Because there's, you can identify the pain there because whatever skill set or whatever title they have in that job and the number of job openings could all be certain things that like an, oh, this equation of these things stacked together, we call it a signal stack, means that like this person is really highly likely to be in market for something that your products can help out with. And so I'm always looking for that like next level thing too. It could be, like I said before, like you're their previous product user. And so that's another stack to that like job change or that funding kind of thing. Speaker 3 Yeah. Or there's some company level characteristic they just don't put up a new office or they also got funding and John like put up with that. Speaker 1 Yeah, and there's this concept that a friend of mine, Brendan Short coined this term micro campaigns and what a micro campaign is and what I'm what we're looking to build for like these plays for our go to market teams is this sweet spot of like something timely happened. It's very specific, the criteria they mentioned like title, role, job listings, all this kind of stuff. And you're, what I'm trying to do is get to like a really small cohort, like a micro campaign of 100 to 50 or so different prospects that like have this really unique thing in common. And then from there you can really tailor the message, really touch on the pain point versus if you're doing a job change only kind of campaign, it would be like maybe there's thousand people who change jobs who are in market for your product and you're just blanketing them with a generic outbound sequence like that is not going to work. What's going to work is really like getting deep and the getting going a layer deeper to understand like what this micro campaign is and like what the pain points are in like tailoring the message more to that. I've seen that work a lot better than the generic type of plays that we all default to. Speaker 3 Yeah. And what you're basically just saying is that you have to get more sophisticated with the data that you're using to reach out to companies. And so a couple years ago, job change was sophisticated enough. Now everybody's running job change. And so it's like you have to add in these other these other data points, stack them on top of each other. And until you can get to the higher level of sophistication and differentiation with combination of data points mixed with your pain points to sell the product. How to break patterns I would say it's more what makes you stand out. What you're trying to do is break pattern and that's level of specificity where you're layering different things on top really helps a break pattern of, oh, I saw this e-mail. So like all these emails that you're getting are all these outbound, they're getting on LinkedIn or whatever, all feel the same. And if you can stand out where it's like you hit, you send a message that's, oh, this really, this person really nailed it. They get me and like, I really feel this pain point. Like that's the level of detail or they'll bar that I try and evangelize so that people like hit that bar because that's what's going to make a difference if you're pattern breaking all the hundreds of emails that people are getting everyday. Speaker 3 Yeah, yeah. So in theory, that sounds awesome. How to find the threshold of good Would love to talk a little bit about how you think about this in practice. And I think there's like two things. 1 is how do you figure out what the threshold is of good in these situations? So it's like you find some combination of data points, you think it's compelling, you run a campaign. How do you figure out is this actually something that's meaningful and that we should continue to use? Speaker 1 Yeah. I mean, the answer to that is just like, you don't know, you don't know, you hypothesize. Like I've come up with, in my time doing growth experiments, I've come up with experiments that are like, Oh my God, this is going to completely denial it. It's going to crush it. Like I know it's going to be great. And then you put it into practice and it just completely flops or there's no difference in the variable versus the control. And. Speaker 3 Any examples? Do you have any examples of that recently? Oh. Speaker 1 Let me see, let me see. Yeah. So I we recently did a a exercise where we're going to use AI generated contents and take the research that our product did and then put that research generate different pages for it. We did a Forbes 100 account research. Essentially it's like we're trying to capture search results, people searching for tell me what Stripe, who's in the Forbes Cloud 100, Tell me what Stripes business bottle or something like that. And so we, we generated all this, all these pages for the Cloud 100 and I was like, oh, this is going to completely crush it. It's like these really long tail keywords that we would rank for. It's going to crush it on SEO and we published it and it's we do. We definitely got impressions and we're definitely getting impressions. We're getting more clicks and maybe I just, you know, give it more time because SEO you need to wait for like people to rank. But it's just something that I was thought was going to explode and it like hasn't really. And so maybe it's just other people are on to those tactics or whatever. And that's an example where recently we we did that. You can check it out at common room dot IO slash research, I guess, if you want to look at it. But but yeah, that experiment hasn't come landed in the way that I was hoping it would land. Speaker 3 Yeah, yeah, I've definitely had experiences like that too. So it seems like what what you're saying, correct me if I'm wrong, is like high velocity of tests try lots of different things and then it's just set your barometers for what success looks like. So you know, whether it's something that's like. Speaker 1 Yeah, yeah. I don't know if I don't know if velocity is yeah, you want to learn fast. It's more just like to know what works you to like actually go and do stuff and look at data and re see what's actually working or not and then use that as a feedback loop to to reinforce like what worked, what did it. And so it's really just, yeah, like actually going out there and doing something and like looking to see how it performed is the way I would uncover what's working or what's not. But yeah, I start with an informed hypothesis of what you think would work or what you've seen work in the past. And then maybe one example of that could be we're reaching out to people who our active product users or they signed up and they fit our ICP like that play is working really well and to run a separate, like a different camp, different play that might work well. Oh, let's, let's segment this and layer on people who are also visiting our terms of service page or docs plus plus are active in the product. And then that's like a, but it's a branch that's not like a complete unknown hypothesis. Like you have some sense of this is working already, let's expand, let's branch off of that thing that's working and run a new experiment. And that should give you a better, I would imagine would give you a better outcome. Of course, you need to like put it to the test and like actually look at the data, of course. And but yeah, yeah, that's how I'd go about it. Speaker 3 Yeah. OK. That makes sense. And then the second thing that I think this brings up for me is the more data points you stack, the smaller the audience gets to reach out to. The trade-off between reaching smaller audiences and reaching larger audiences And so there's this inherent trade off that you're making of let's go build this custom campaign for the smaller subset of companies versus let's let's target more companies, less differentiated. And I think like historically, a lot of companies think about it is we don't want to go after these like really tiny sub segments with micro campaigns because it just doesn't scale. How do you think about that? And do you think just the access to data and the need to be able to run these is just more important? And so like that makes the stronger case. Speaker 1 Yeah, I don't think I buy the like that doesn't scale argument so much. Maybe more in the Paul Graham school of thought here, which is do things that don't scale and then you'll be able to figure out how to scale them. Once you figure out that they work. My default is to go to let's figure out what things are working. It can be a micro campaign and then, yeah, you're only reaching 100 people and maybe your SDR team or your outbound AE team is a dozen people or so. But I think once you do crack that code of, oh, this is working, then stacking on top of that becomes much, much easier. So you come up with more and more ideas and it can then scale. So I always start with come up with a hypothesis and audience and use a spreadsheet. Let's use ChatGPT and then run outbound sequence just with e-mail. Like you could start with the most basic stuff possible. Always find people are like, oh, we're going to dry up this channel. We're going to dry up this signal, but once you find something that works, I feel like it just is a great mechanism for ideation of, oh, that works, Maybe something else that's similar and tangential that will also work. And I can come with more and more ideas and just continue to stack on top of what we've already learned. If you scaled something that doesn't work, you're just burning money and burning your brand affinity. And so start with what'll work, and then you'll figure out how to scale. I feel like there's always the opportunity to figure out how to scale. Speaker 3 And so if we take this job change mixed with now they're hiring a bunch of people, example that you were giving a little while ago. Let's say you run that, you make that work. Then your point of view is OK. Even though in a given month there aren't actually like that many people changing jobs in your ICP that are now hiring, you can now come up with new ideas that are slight abstractions off of that original idea. Speaker 1 Yeah. Or maybe it's like, oh, they got a promotion and now they're hiring. Or maybe we want to target a different industry of people who change jobs because that's tangential. Or maybe it's people who change jobs who are also engaging with a competitor on a different channel. And that's just the job change trigger that you're stacking on top of that. And so I feel like they're just, there's enough data out there that you can continue to scale and continue to layer on things and come up with ideas. And so, yeah, I've, I have yet to find a time where I've like at a shortage for ideas or data. And I just like, oh, I want to do more and automate more and do things with humans on my team that are the hard things. Figure out how to automate that, put it into our stack of this is a pipe chain play that's on autopilot essentially. And then continue to do more on the like things that don't steal side of things with the team. Speaker 3 Yeah, yeah, that makes sense. Examples of non-obvious data points Do you have any other examples that you've seen on your end or with any of you guys as customers of non obvious data points or combinations of them that have been successful? Speaker 1 Yeah, that's a good question. One anecdote that I like to share here is and it really highlights the fact that often times data points are specific and bespoke to your the company that you work at. And so the example here is, is from retool. And when I was at Retool, we relied, we had a lot of people signing up for the product and using the product. And one of the signals that we identified there was that we would find some people would sign up in with a personal e-mail often times. And they would just connect a data source like Postgres or Snowflake and just load a ton of different data into to ritual and it would just spike usage like crazy. And then a day or two later or maybe even a few hours later, that same usage will just flat line and go straight to the bottom again. And so it's like you saw this really big spike in usage and we're like, what the heck is going on here? And we then started reaching out to some of those people and found out that, oh, that is actually someone who works at a really large enterprise. They want to test that this product can scale the way it does. And so we were ignoring them before. And then once we found out like that was what was that behavior was a signal of, and like they work at a large company and they're just putting the product to the test, then we saw that as a signal of, wow, whenever we see this activity like that is a really high indicator that our sales team should reach out to the person, regardless of any kind of qualification or anything about that, that we found on the enrichment side. And I feel like there's just things like that that are specific to your own company. And if you know your data, if you know your users, there's always things that are bespoke to your company and your product. And that's the type of stuff that makes you stand out more, makes you stand out or makes you more sophisticated than someone else who's just using this generic website, job change, funding type place. Speaker 3 Yeah, totally. So one other thing that you and I were talking about a little bit on the data side is dark social content. Dark social content Could you talk a little bit more about what that is and how you think that ties in? Speaker 1 Yeah, I think it's, I think it's, it's a interesting change shift into user behaviour and then also the way that user behaviour can be captured or harness on the marketing front. And So what I mean by that is call it like 5-10 years ago, I feel like the user journey was very linear. It was like someone is researching your product or landing on your website. Maybe you can use like clear bit to understand the companies on your website, but then they'll fill out a form and then they will you'll get signal from your CRM and then with the LinkedIn feed or Slack, Slack communities or GitHub open source activity, which has been around for quite a long time. Actually. These are all channels where the your audience is doing that same research, but they're doing it outside of what's in your control and your purview. It's not first party data, it's like second party data, I would say. And because of this kind of like explosion and users doing research on these different other channels are doing their due diligence almost like online word of mouth. There's like since there's been this kind of trend away from someone revealing themselves within the product and then moving to these dark social channels that makes it really hard for the traditional way of doing marketing or go to market because you don't you're not like counting MQLS, people who are downloading an ebook, they're not doing that anymore. And that is an area of opportunity because if you're in any go to market team, really you want to meet people where they are and engage with them all in those channels. And that is where a product like Common Room can identify what's happening and identify the person behind these dark social activity. Like if someone's engaging with a competitor on Twitter or engaging with your LinkedIn posts, like things like that, where it's OK, you're a marketer, you're educating the market in these dark social channels. You're not like forcing people. They'll give their e-mail address over, but we can't identify. We know who these people are. It's almost like this new channel to identify qualified leads that haven't exactly like raised their hand or filled out a demo form. And so it's like going up the funnel a little bits to say like, oh, these people are in market, they're just in this channel that is hard to traditionally capture interest and intent, but we can with a product in like common room and pull it in so that it's adding to our top of funnel. And essentially, like 1 interesting take here is that marketing's new role could be less about getting people to raise their hand and fill out a form and enter into your CRM and more about like, how does marketing help generate signals? Because those signals are now trackable in the dark funnel like they never were before, or it was. It's a lot easier now with a product like Color. Speaker 3 And so what's something marketing could do to try to actually generate that or are you saying more find it? Speaker 1 So a good example is actually like our LinkedIn strategy, which is we don't have anything that's gated. We don't we're not pushing to like people to leave LinkedIn. I think, oh man, I can't remember her name, but she works at Spark Toro, Amanda Nattivi dot I think coined this term zero click content. And so that's like what we that's what we are creating or that's our awareness approaches. Just create the zero click content where you're just like in the feed, you can consume this, you learn about common room and then maybe you'll do like a, a branded search or you'll go directly to our website. And so as a marketing team, we're doing that versus trying to drive people to our site to fill out a form. And because we can capture all of that signal that's happening on LinkedIn or happening on Twitter or some other dark social channel like that is our charter is like generate as much demand there. And then we'll capture it either when people come in and raise their hand, or we can capture it through common room when it makes sense. Speaker 3 I see. Are there a finite number of these channels? It's like Slack channels, LinkedIn, Twitter's, their exes are, I mean. Speaker 1 I could probably think of like a dozen. Yeah, I can probably think of a dozen off the top of my head. There's if you're, let's go through a few examples, if you're in a commercial open source company, there's a lot of activity on a GitHub repose and maybe your Devrel team is like your marketing team that's like engaging and generating interest there. There's Slack communities, there's Reddit, there's all the social channels essentially, yeah. And then there's what else sub stack. I don't know. This is like how I think about marketing in general is OK, there's a bunch of watering holes out there. What watering holes are your audience engaging with? It could be a convention of like real estate in investors are real estate agents. And if that's the channel, that could be your dark social channel that you like tap into. Speaker 3 Yeah, to your, the summary of what you're saying is like finding people is easier than it's ever been across all of these channels. And so it's if you put your marketing effort into finding more people that engage with the content that would indicate that they're in market or that they're right fit to buy, whether it's your content that you're posting in these channels or it's just pulling off of other people's content, then you're increasing the scope and the number of people that are actually in market or showing intent signals really materially. Speaker 1 Yeah, Yeah. It's very first principle of marketing as the final, where people final, where your audience is, engage with your audience on that Channel and then maybe even create demand within that Channel is as as simple as that. Identity Resolution Yeah. And so then that ties into identity resolution probably a lot. And I know you have a lot of thoughts on that. So how do you guys even do that? And what do you think the future of that even looks like? Yeah, yeah. Speaker 1 I feel like this is the unspoken. I like to call it the unspoken linchpin of go to market is identity resolution. And yeah, if you have one channel that really matters, like maybe it's not as important, but I feel like, OK, what is identity resolution is essentially saying, OK, we know, we know with a high level of confidence that this person is who we say they are, who we've identified. And that could be someone who is engaging with LinkedIn comment content. It could be someone who's active in a GitHub repo. It could be someone who's on your website. And all these channels, all this signal from a single person in across different channels. If they're in a silo, if you have to jump from one tool to another or they're creating lots of duplicate records in your CRM, that is a a recipe for chaos and mismatched alignment with what's happening with the customer journey and where you're reaching out to that customer on their journey. If you don't have all of those things unified into a single user profile. And so as the channel explosion happens of there's more signal happening in more channels, the need to identify and resolve that a user's identity from 1 channel to the next becomes like really critical. And it's a hard engineering problem to solve. I know from like time at segment that was a huge challenge that we helped out with. It was more on like the product and website events, merging side of things. But having with confident knowing with confidence that this one person is who they who you think they are and mapping their the log of their whole entire journey is what makes it possible to do that kind of a micro campaign signal stacking side of things that we were talking about earlier. Because then you can with a high level of confidence, Add all these filters and Add all these different activities to come up with this bespoke micro campaign that's executable on your go to market side of things. So, yeah, I know identity resolution is like a boring thing. It's all this is a technical thing that's not so interesting, but it is a very critical linchpin to go to market, I think, and not a lot of people talk about it. So that's my spiel. Speaker 3 Yeah, so dark social content, tracking, all of these intent signals, it's like it only matters if you actually can tie it back to a single record. The challenges of tracking users across social channels Yeah, exactly. Yeah. Speaker 3 Yeah. What do you think are the challenges there? Because my initial reaction is doesn't seem too hard, like you just find the person tied to an e-mail address in the LinkedIn profile and then you track it over time. What are the what are the things that make it hard? And then what are the tips you have to figure that out if you're a business? Speaker 1 Yeah, I would say like the primary key is a one of the hard parts there. And so the primary key is essentially like, oh, this unique value is what stitches together all of this user's journey. But that used to be a lot easier when you had this person's e-mail address because they would fill out a form and then once you have that form, you could match it to all these other different things. But as we meet the customers where they are in different channels like LinkedIn, GitHub, Twitter, Slack, like you might get their e-mail, it might be a personal e-mail, might be different than their work e-mail. You might just get their handle, you might get their name and their title. And joining on all of those things that are differentiated becomes a really hard engineering problem to solve. And so that's where it's all you need to have not only the ability to identify who this person is, but then to enrich their information and have a primary e-mail key, primary user ID key, all this kind of stuff where it's all we know who this person is, who they say they are. And we're going to merge all these things together into a single unified profile. That part is becoming a lot harder as the surface area for signals and channels where you're identifying with where these users are acting on is happening. So I think that's where the harder engineering problem is happening right now is like there's just a myriad of channels and like you get different data points across each channel and having and merging all that together is, as you can imagine, probably not so easy. Speaker 3 Yeah, like how do you turn someone's sub stack into knowing what their e-mail address is and where they work? Speaker 1 Yeah, or or Twitter handle or GitHub handle or their personal e-mail address with their work e-mail address. So this is like a lot of examples there, yeah. Speaker 3 So part of it's like how do you actually set up the underlying system from an engineering perspective? What are the other tips or the things that you've seen the companies that are the best of this do? Best practices for tracking users across different point solutions The other thing we didn't quite touch on too is that if you're tracking users in different point solutions, like a website visitor point solution like RB to B or a social tracking system or like the social engagement system. And these are all like different like silos that like those tools like don't communicate back and forth to each other to say, oh, this user is visiting the website who's also engaging on social, who's also using our product. If those are all three different products, it's really hard to tie it together, stitch together all those pieces. And So what I've seen really good data teams do is take all that data and dump it into a data warehouse. And then they have a data scientist that's stitching together, OK, we have all this unique information about different users. What can we do as a data science team to connect all the dots here and unify this user profile into a data warehouse? And then from there, our data warehouse is going to be the source of truth as if you're an engineering team or if you are doing this right, you essentially have a data source of truth that the engineering team is like doing all this identity resolution for you. That was the way that we were doing it or that we saw good teams doing it when I was at Segment. And now with common room, it's like that is something that we built in from the product by default. And so if you don't have that data science team or if you're not pulling in signal into a data warehouse, which a lot of teams don't do that, we have it all together in one system. That's also much more familiar and helpful for a sales Rep or marketing person to interface with then like a data warehouse. Like I don't know how many marketers or or sales people are like jumping into ABI tool and using sequel on top of a data warehouse. That just doesn't really happen. Speaker 3 Yeah, Yeah. OK. That makes sense. Any anything else on data, dark social content, identity resolution that we haven't hit on? The Data Layer No, I think what is coming through as a theme of this conversation is that the data layer and this kind of we were talking about before of the data layers, like what signals are you capturing? That's data and then being able to like unify those signals into a user profile and then enrich them. Like this is all just data. It's all data. That is the foundation of how you would run the right kind of go to market motion on top of. And if you have that, if you have your ducks in a row there on the data side, then it just makes everything else like the execution, the automation, the hard hitting outbound like all that stuff becomes much easier if your data foundation is in good order. Speaker 3 Yeah. And it seems like there's one other piece that is how you combine and structure and think about the combination of these data points because finally the underlying basic ones easy. Then there's one level higher that's more challenging on sub stack and acts and you know those. But then it's what are the combinations of them that actually drive interesting campaign intent? Speaker 1 That's where the experience human part of things, the the softer skills come in where it's like, OK, what are the things that we need to combine and mix and match to come up with a hypothesis and then test that hypothesis. Right now, it's very much like a human coming up with a hypothesis of, oh, this combination of things or this layering of things is where we want to run a run a outbound player pipeline play against and let's try that. But we do have the data foundation layers there so that we don't have to spin our wheels trying to even just execute this thing. Speaker 3 Yeah. So you guys work with a lot of companies that you know are growing quickly and are at the forefront of this. What sets the best GTM motions apart What do you think set the GTM motions apart of the companies that you think are doing the best? Speaker 1 Yeah, Yeah. We heard you have the privilege of working with a lot of really great teams. One of the the patterns that we see with the customers that we work with is that they have a lot of data, they have a lot of signals. It could be like APLG company or it could be a company that has a large brand presence on LinkedIn or something. They just have a lot of data to work with and they know that data is a competitive advantage for them. Or it could be a competitive advantage for them because it's their own first party data or they're generating that data themselves. And so they come to us or that they know like we're sitting on this gold mine of data and we know we need to like use that data to inform our go to market teams to take action in a more efficient and like prioritize way so that we are efficient with our go to market motion. And so that is like a, a common theme across all of our customers that we work with. It's just there's a lot of data. They know that data is valuable and then they're at looking for help of like, how do we make sense of it? How do we stitch it together? Where have you seen other companies use this data that is similar to our business model or our motion? And then like how do we execute on top of that? I think that's where I see, you know, a, a product like common room or the modern way of doing this is like, you know, you have a lot of data at your fingertips. Let's take that data and use it in a way that's more informed versus this kind of cold outbound scattershot approach to go to market. Speaker 3 And what do you think even the best companies are, what their biggest weaknesses are, where you feel like the market just hasn't gotten to yet? The biggest weaknesses in the market Yeah, I mean, at the same time, I'm like, oh, these companies are really at the forefront of of go to market. I still feel like you see the brand names that you would recognize and you're like, wow, they must have an amazing go to market team and function and marketing operations and all this kind of stuff. And I would say that every company, no matter how recognizable or how fast they grew, has growing pain, pain points, the same kind of problems. And yeah, I think that you'd just be surprised at how some of the basics of just like stitching the data together or doing an account prioritization list or scoring and things like that are just left behind or forgot about, or maybe they have a shiny object syndrome. And so I feel like that's the thing that is a constant challenge of, hey, let's just have someone help us reset and figure out what are the things that we can be doing to give us the quickest wins and the most leverage. And then from there build a foundation to like expand. Speaker 3 I was talking to Emre from waterfall dot IO and he was telling me that he thinks 25 to 50 companies have figured out all this stuff in a really sophisticated way, and that there's just all of this room for most of these businesses. Speaker 1 No one, no one has it figured out. No one has it figured out. We don't have it figured out internally at cover. I think we're doing the right things, but like, there's always room for improvement. There's always things that we could be doing better. There's always ways we can make our data more defensible or find different types of like Nuggets in our data. And yeah, I, I, I just, I don't think if there's ever like a perfection, it's always just like a constant learning curve and a constant like adjusting your priors and changing things in the, the market's also changing super fast too. So maybe if you do nail it or have it all figured out today, in a year, it's going to be much different. So it's always this constant evolution of adaptation, learning, feedback loops, all that kind of stuff. Speaker 3 Yeah, totally. Best account-based marketing With some of your customers or maybe other businesses that you talk to, what do you think is some of the best account based marketing that you've seen? Speaker 1 Going back to what we talked about earlier of finding this specific trigger that puts an account in market is where I see the best account based marketing because it's this ever like listening and evolving thing. And if we were to use an example of we work with this like home buying company or it's like a real estate market company that's like very different from our traditional customer set. But they are pulling in signals from different new sources and job listings that are specific to like the real estate market. And so there's not a lot of other companies that like care about that specific thing. And so they have this signal that they know puts a customer in market for them or a prospect in market for them. They're able to have this Evergreen account prioritization. And they know when an account or a person within that account hits this specific like trigger or bar, then that is now becoming the account that they're going after with a very specific reason and message. And so that is where I'm seeing like the most advanced stuff is like, where do you match the pain that your product relieves with someone who then becomes in market for that pain? And can you make that a constantly refreshed kind of motion where you know the best accounts are being serviced to your reps in real time? Speaker 3 Yeah. Speaker 1 Or close real time, I should say. Yeah. Speaker 3 That makes sense. Changing gears a little bit, if you were a business and you were very behind on all this stuff, didn't have people proactively thinking about finding new intent, maybe you were like on some legacy data provider tools. Advice for businesses behind the curve What advice would you give those businesses for where to how to get started? Speaker 1 Whenever, whenever I get asked questions like this, I'm like my, my answer is usually the same, which is instead of listening to me or listening to a podcast or reading an article or anything like that, my advice is to just go out there and do it and learn. And I feel like 80% of what I've learned in my career or in in figuring out what works is actually by doing versus listening to other people. And so, you know, maybe it's helpful to get some ideas from different resources, but then like once you have that idea, don't get hung up on perfection. Start putting that that idea to action. And you'll learn so much more from doing that then by trying to architect the like perfect thing. So my advice is just go out there and do it. Speaker 3 Yeah. Finding the right talent And how would you think about finding the right talent and also help you do this? Speaker 1 Yeah, that is a interesting question because you know, it's kind of funny how people put up like marketing racks or go to market racks where they're like, we want someone who has 10 years experience and knows HubSpot or Mercado or knows this traditional way of doing things. But then if you think about what skill set you need right now, it's almost like the skill sets have changed in the last two years where it's like you need to have some sort of idea of how to use, how to wield AI and use it to scale yourself. You need to know how to tap into all these signals. And what we're looking for in employees is certainly like changing. And so if I were to give advice to someone like, Oh, this is how I can grow my career. It's more like learning this new skill set and then also going out and doing and experimenting with things. And like you'll, you'll also learn, but you'll set yourself apart from someone who's been doing the same old predictable revenue slash marketing OPS playbook for the last 5-10 years. It's something that I think about for myself a lot too, is like, how can I Polish my, my skill set? How can I be at the forefront of things? And it's a fun time because I like learning thankfully. But if you have curiosity, I feel like you're naturally gravitate towards doing these more the cutting edge type of like things. Speaker 3 Yeah. And you've built a couple of your own sort of a vibe cutting tools, right? Vibe tools I had a lot of help on the front end engineering side of things, but myself and a person named Josh Lind on our team, we have a vibe coded a good amount of stuff. One example there is I often get asked a lot about what plays should I get started with? Like what are the best plays that are working? And the answer is unfortunate because it really depends on where your signals coming from, the size of your company, the size of your team. There's a lot of different factors that play into which signal plays should we start with? And so we have this kind of like repository of knowledge of I, I have a spreadsheet of 100 signals and like, you know, categorized by, are they product signals? Are they like open source signals? Like all the kind of stuff. And essentially, like, if we can gather enough information about company and they're go to market motion, I can filter that list in a certain number of ways to say, oh, here are the best. Here are the signals that I would start with. Might as well start here. But we had a thought like, oh, we can actually use AI to to figure out all the stuff for us. And so like as long as we have someone's domain, a company domain, we're having like Gemini go search for information about that domain. And then we're returning information about what their business model, how much website traffic do they have a product LED motion or not, stuff like that. And then that's like our decider type of AI. And then we'll then have I think it's open AI mini or ChatGPT mini 4 dot O. We'll then parse all that into Jason and then render it onto a a page or read render into sorry, markdown. So we can use it for a website. And then that is just like recommending place for people who put in their domain and they'll recommend place for them. I encourage people to go to play Gent PLAYGENT dot AI and try it out because this is a tool that we vibe coded. And I don't know do anything like this in this kind of curiosity is the type of thing that like I would encourage people to do to like learn and make themselves more hireable in the future. Speaker 3 Basically what you're saying you built is all you do is you input your domain and then it's finding all this information about your company and then suggesting data points in outbound. Speaker 1 Yeah. The thing I miss is it's looking at our content, it's looking at that spreadsheet, it's looking at all this, all these things that we've these playbooks. Speaker 3 That we've created like what is possible. Speaker 1 Yeah. What's possible And it's matching this business model, this signal that we've identified from this domain to the signals. And then it's like surfacing up the signals that they should. You would start with. Speaker 3 Nice, I like that. Changing gears maybe one more time there, there are a lot of new software tools or tools that have been around even for a handful of years that are all that like how to use AI and your go to market and they're at the cutting edge of all this. Which companies will win in the long-term? How do you think about which companies are going to win in the long term common room versus all the other tools on the market? Speaker 1 Yeah, I really have no idea. If I did, I would not work at common or I would just be an investor, I think. But no, yeah, it's really hard to say. I do think that there are maybe a few key things that I would look for. One is just like again, but going back to data foundations is like I, I do believe the data is like a differentiator for AI or for AI products. And so is there some sort of like unique knowledge or unique data set that a company has or can capture that differentiates them from anyone who can just spin up a tool that's using non proprietary data? And so that is one thing that I would look for is like, is there a competitive note here? The other stuff is just, I think I think product market fit is somewhat easy to identify, especially especially for companies like Open AI or Anthropic or whatever. It's like they're just continue to grow. They continue to have new use cases, expansion, stuff like that. And this even goes from the vibe coding companies like Lovable or Cursor, things like that. And so I feel like there's something there and you can see that in their growth numbers and also probably like their expansion. That's one thing I always look for and if a company is going to be successful or not is that if their NRR is increasing or like 120% or higher is kind of like a good benchmark for that. I think these companies have a much higher NRR than that actually. So, you know, just just looking at things like that is where I would place a bet. But then by then like everyone else knows that too. So you know it. Maybe it's too obvious, I don't know. Speaker 3 Yeah. What if we just take a step back? Unsolved problems in GTM What do you think are basically like the most important unsolved or still being solved problems in GTM that software can answer? Speaker 1 Yeah, interesting question. I still feel like we haven't, yeah, we haven't cracked like the the automation code, but I don't know if we ever will get there. I feel like there's, there's still humans in the loop and I just feel like the, the LLMS are not at the level where it's, you can fully relinquish all trust to them. The other thing I haven't quite seen yet is the, and like, this is kind of like the area that we're building towards with common room is OK, We have all this data, we have all this, we have all these learnings from that data. When we see something new or when we, when we learned something new that works, how can we just like surface that up or suggest or recommend new plays? But this kind of like we were talking about earlier, it was like, we definitely have a human doing that hypothesizing right now. But if you can essentially have a super data scientist that's an AI layer on top of your data, that's just seeing all this stuff happening and being like, oh, this is the best account, here's why. And just kind of putting all the stuff up on a platter. I feel like that's what we're building towards. But the market is definitely not there yet. I haven't seen anything even come close to being able to do that for go to market. People definitely say that they do that, but I just it's not a reality yet. Speaker 3 Yeah, you're just getting at, is there a space for a tool that does this almost just end to end? Speaker 1 Yeah, yeah, we're just. Speaker 3 Reaching out to people. Speaker 1 Yeah, you just interface with it and it's just giving you all these suggestions. You don't have to do any digging or building or anything yourself. Maybe there's something, maybe there'll be something like that. Let's like the ChatGPT or Vibe coding equivalents to for go to market. But if I knew what that was, and I think we're like, we're getting towards building that a common room, but if I knew what that was, I would do that or invest in that or something like that. I just, I think it's hard to, it's hard to foresee exactly like what that thing will be, you know, But I do think that like the data layer, the data foundation is what's critical for that thing to work. Speaker 3 Yeah, I agree with that. Getting into just the final couple questions, what is your favorite software tool that you think most people haven't heard of or that's underrated? Favorite software tool Yeah, interesting. I so I've tried some of the vibe coding tools. They're fun. Actually, the tool that I really enjoy is this tool called air OPS. That's what we built plagiant with. And what I really like about it is this kind of this AI orchestration. And it also like I, I think using a tool like this really helps you learn how AI works, where it's like you're tinkering with different models, you're adjusting prompts to get different outcomes. And it helps you understand, Oh, this AI thing is not like the scary amorphous thing. It's more just like, oh, I by by tinkering here or changing, changing this LLM model or changing this prompt, I get this different outcome. And like that type of that type of learning, I don't know, I feel like it's fun and it's like a new world. And I really like air OPS as a way to use that. It's this orchestration of if this than that like canvas of building different work flows with AI. Speaker 3 OK, nice. And what's a prediction you have about the future of GTM engineering? Predictions for the future of GTM engineering The future of GTM engineering, I would like for identity resolution and enrichment to be like the thing that people care about the most. I think that's a hope that's not going to come to fruition. No, I think I don't know about GTM engineering in general as being AI. Think of it as big data back in the day or, or maybe AI now is that it's just going to become like a default part of our workflow, a default part of who we hire. And it's going to be, it's not going to be a specific thing that you're looking for. It's just like everyone's going to need it or have it. And so I feel like it's a term right now that's like useful for like, oh, this is a modern different way of doing things. But I think that's going to be like the default and it's just going to become go to market. You don't need go to market engineering or go to market AI or anything like that. Speaker 3 Yeah, just as like the common way of thinking of all these businesses. Cool. Yeah, that makes sense. Great. Thanks so much for making the time to chat. And yeah, this was fun. Speaker 1 Awesome. Thanks. Speaker 2 That's it for the pod with Kevin. Don't forget to follow us on Spotify and subscribe to our sub stack. You can also fill out the form in the show notes if you want to join the GTM Engineer Lab. Until next time.