Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

The B2B Roundtable

The B2B Roundtable artwork

How to Turn LinkedIn Engagement and Website Visits Into Sales Conversations, with Ariel Cohen

Nov 17, 202622 minAudio
ShareXLinkedInFacebookWhatsAppEmail

About This Episode People engage with your content, visit your website, and then go quiet. That activity can mean several things. How do you tell when it adds up to a useful reason to start a conversation? In this episode, Brian talks with Ariel Cohen about how Swan connects signals across people and channels to see which accounts are already paying attention. Ariel shares the story of a past prospect whose renewed activity prompted the team to reach out. When they asked whether the company was evaluating Swan or just enjoying the content, they learned the prospect was considering the new product. That conversation led to a demo. The conversation also explores how sales, marketing, and AI agents work from shared account history and outreach guidance, why agreeing on processes and measurement takes real work, and how a LinkedIn link and a voice note helped Swan capture event conversations and follow up more consistently. About Ariel Cohen Ariel Cohen is the founding GTM engineer at Swan. With a background in electrical and computer engineering and AI research, he began building growth systems while running his own business and experimenting with organic LinkedIn content. At Swan, he connects buyer signals, account context, and outreach workflows; his post about his Becoming 10x talk prompted this conversation. What We Cover What Swan was missing when LinkedIn activity and website visits stayed disconnected How engagement across several people reveals an account worth paying attention to How renewed interest from a past prospect led to a demo The difference between familiarity, curiosity, and readiness for a sales conversation How a shared GTM brain keeps people and agents working from the same context Why defining metrics and agreeing on processes can be harder than building workflows How a LinkedIn link and a voice note replaced badge scans for event follow-up A Few Things Worth Taking Away Look across the account, not just at one person. Engagement with different team members can be scattered across LinkedIn, website visits, and events. Bringing it together helps the team recognize patterns it would otherwise miss. Use interest as a reason to ask. Familiarity makes a conversation easier to start, but it does not establish buying intent. In Ariel’s account example, the prospect’s response confirmed that they were evaluating Swan. Shared context has to reach the work. At Swan, Jeremy, the VP of Sales, updates the outreach instructions that Ariel and their agents use. Shared account history also helps outreach pick up where an earlier conversation left off. Agree on the process and what you measure. Ariel says defining what to track and what counted as an MQL took substantial effort, and Swan changed the MQL definition along the way. Consistent processes and measurement matter when evaluating the results of an AI workflow. Capture the conversation, not just the contact. A LinkedIn link paired with a voice note helped Ariel record what someone asked for and when to follow up. He says the team followed up more consistently and converted more after adopting the approach. A Few Lines That Stuck With Me “The foundation of every successful AI implementation in a company, no matter what the size and no matter what the AI product, is having tight processes.” “When a salesperson now spins up a workflow, does his agent know about the change in positioning that we did last week?” “People that add you on LinkedIn, usually if you reach out to them and talk to them, they will respond.” “You just have to keep iterating. That’s the only thing that works, to be honest. My advice might not be relevant in a few months from now.” Resources Mentioned Swan, where Ariel builds GTM workflows Ariel Cohen on LinkedIn Sumble, the data company whose CEO discussed “signal zero” with Ariel Claude skills, discussed in the context of shared instructions for people and agents Chapters 00:00 Spotting Hidden Buyer Interest 02:45 Connecting the Buyer Signals 04:24 From Engagement to Demo 07:56 Curiosity or Buying Intent? 11:32 The Shared GTM Brain 15:26 What’s Worth Tracking? 18:33 Voice Notes Replace Badge Scans Transcript Brian Carroll (00:00) Welcome to the B2B Roundtable. I’m Brian Carroll, and I wanted to reach out to Ariel Cohen after I read his LinkedIn post about an upcoming presentation he’s giving on becoming 10X. Ariel works with Swan, and they were doing about 40 LinkedIn posts a month. And what they were experiencing is that people would like a post, they’d check out the pricing page, and then they would go quiet. So how do you know when all of those activities add up to the reason you should reach out to someone? Well, we’re gonna talk to Ariel about that. He is the founding GTM engineer at Swan. And Ariel, welcome to the B2B Roundtable. Ariel Cohen (00:46) Hi, Brian. It was great to meet you back then. It’s fun to be here. Thanks for having me. Brian Carroll (00:50) When you joined Swan, what made you think you were missing something here with how you were doing things through LinkedIn and the other types of signals that you were putting out there for people to engage? Ariel Cohen (01:02) Yeah, so I joined Swan around June. One of the reasons that Swan grew so fast was founder-led content, which is a great strategy, was a few months ago and still is today. We’ve actually doubled down on it last month. It’s going really well. So basically the the CEO, founder and CEO, his name is Amos, and he he’s been posting a lot on LinkedIn. He’s been very active. So Amos was already posting, and Swan is a product, so we had a lot of people sign up. We would get PQLs, so like relevant companies would would sign up. Some of them would ask for demos. But we noticed that - or I noticed that - a lot of engagements, both on the personal level and on the company level, from different people were just floating out there, and we weren’t really doing anything with them. And so over a few months, I created a system where we capture all these signals. And by signals, I mean likes, comments, profile views, connection requests. Incoming inbound connection requests and also DMs to make sure that we don’t miss anything on LinkedIn. And I cross-check that against other signals that we have, like website visitors or maybe webinar attendance, and other places, to really create a a better understanding of what accounts are actually in our orbit and what accounts are actually being interested, so that when we could go and reach out to them, we are much more likely to get a response. Brian Carroll (02:36) As you were putting things together in terms of the signals that you’re observing in LinkedIn, what was it that you could see that you’d been missing after you started assembling these things? Ariel Cohen (02:45) Well, first of all, we weren’t even tracking all those signals before. So just by doing that, you suddenly see, okay, this VP, this CRO, has been liking ten of my posts, and he’s also been viewing my profile five times. If I reach out to him, he’s probably gonna know who I am, right? So, like, just capturing that demand, being able to open open up conversations, both manually if you want to keep a full human in the loop or autonomously, depending on the a specific set of scenarios. And then the additional value that you can get when you’re tracking all of these is across the company. So I am also like a native LinkedIn creator. I have forty-five thousand followers, so I also post and and get good engagement. And so some people saw my posts and maybe engage with my post, and then maybe other people from the company engaged on Amos’s posts or on Jeremy’s posts. And so you kind of get this scenario where you have all these people. Maybe even different people within the company that know who we are. And that just just gives us a lot of other opportunities to to engage with our buyers that already are in our orbit. And when using that strategy, we’re able to get like thirty to forty percent reply rate, depending obviously on the persona and and such, and not all of them are in market, but it definitely helped us get more conversations, and we probably doubled - more than doubled - the MQLs since we started doing that on a company level, with the same input. Which occasionally, as we got better with the outreach, led to us doubling our demo - our monthly demos - without investing a dollar more. Brian Carroll (04:24) Wow. Could you walk us through an example account where this worked recently and, you know, what happened? Ariel Cohen (04:30) Yeah, so I - let let’s give a few examples, because each buyer - Brian Carroll (04:33) Yeah. Ariel Cohen (04:34) - journey is very different. So let’s say client number one is somebody that we talked to in July twenty twenty-five. Got on the call with them, they saw the product’s last version, they weren’t even aware of the new product and what it does, and we actually were on a call with them. And we - no, we knew what they were trying to do. Ah, and then suddenly you see a rise of interest within their account. So suddenly you see people checking us out, viewing our profile, maybe head of demand gen, maybe VP Sales, and suddenly you see this kind of company-wide demand. And then we we reach out to them, we say, “Hey, saw you like a couple of you guys - of you guys are checking us out. “Are you in market right now, or are you just, I don’t know, interested in our content?” In a friendly, not-so-intrusive way. And they were like, “Yeah, we’re evaluating you guys. We saw that you had a new version, and we haven’t talked for a while. Let’s chat.” Cool. That turned into a demo. Brian Carroll (05:34) Mm-hmm. Ariel Cohen (05:36) And then another example is maybe somebody that was looking at Amos’s content, for example, and then they sent him a connection request. So in my eyes, and we’ve seen this work, a connection request is a pretty direct inbound signal of personal-level interest. So not not necessarily - they’re not necessarily in market, but people that add you on LinkedIn, usually if you reach out to them and talk to them, they will respond. Right? So - Brian Carroll (06:01) Right. Ariel Cohen (06:02) Every exec or - or person in a target company that adds us, we know Swan scores them and we see if they’re ICP. And if they’re ICP, we immediately approve them and and reach out. So that has resulted also in quite a few demos. But it doesn’t end with LinkedIn, I would say. So to do a lot of with signal aggregation, right? So let’s say maybe you’re a company that doesn’t have good LinkedIn activity, but maybe you’re doing events. And maybe you’re also tracking website visitors, like probably like the more established companies, right? They could have like a dozens of events a year. And if you ingest all that context into your - what we call a GTM brain - and you know who you’ve been talking to in these events, and then you track them on the website, then you could see, okay, I talked to whoever in this event, and now they’re checking me out maybe three, maybe five months after, and now they’re in market. I already have a relationship with them. Let’s reach out. Right? And so this latent demand that comes from like stacking the signals, I believe, is something that that everybody should do, and and if not, they’re just missing out on so like easy money. Brian Carroll (07:10) With your upcoming talk, I’m I’m sure you’re gonna be talking to to marketers. What’s the advice that you’d give them? Ariel Cohen (07:15) Well, I think number one would be to look at what channels you’re already using and try to get the most out of them. So this strategy is what I would say is is the basis to every successful signal-based outbound motion. And I would probably try and capture the demand that exists already. Like you just need to look at what signals you’re - people in your orbit - already giving you and then try to just stack them. And there are a few ways to do this. There’s not one way. You could stack it in your CRM, you could use Swan, you could build it in a database. There are a lot of ways to to get to that result. Brian Carroll (07:56) And as you were thinking about the examples that you gave, what are the things you do to tell a difference between someone that’s enjoying your content from someone who’s actually ready, uh, you to engage and have a conversation? Ariel Cohen (08:10) Yeah, well, that’s a - that’s a great question. Basically, you’re saying, how do we know that somebody’s in market? Well - Brian Carroll (08:16) Yeah, exactly. Ariel Cohen (08:18) - you know, you could use whatever third-party data is out there to try and estimate it more. Like, for example, for each business, it means something else. For us, if the company is growing and they raise money and they’re hiring SDRs, they’re in growth mode. And then if they engage with us, they’re more likely to to be interested in talking to us because we’re a GTM company. But this - this data, or what - I’d recently had a talk with the CEO of Sumble. It’s a data company. And we talked about - I think he called it like the like signal - signal zero. That, like, that one thing that if if that happens, then you know that they’re in market and they’re relevant. And then you can use whatever mechanism you want to get fo - to get them to be familiar with you and to reach out to them. But yeah, for us it’s growth, growth initiatives, but it would mean something else from for each person and it’s something that you have to figure out. Brian Carroll (09:20) I’d love to hear from you. How did you get going in that? Ariel Cohen (09:23) I have a data science background, so I always like digging into these numbers and seeing what what customer journeys work the best and what signals - like, maybe we’re reaching out too early for certain signals, and and that’s hurting our reply rates. Maybe the type of outreach is maybe inappropriate, or maybe you should try a few variations until we crack and get the best numbers, and so. You could do that in a data-driven manner, but lots of times it’s just kind of like human GTM alpha - Brian Carroll (09:58) Mm-hmm. Ariel Cohen (09:59) - where you just kinda know, based on all your sales calls and just your business sense. Brian Carroll (10:04) What I would probably add: with you being so embedded in your organization’s GTM process as the founding GTM engineer, you probably see all of this. But I’m guessing, for other companies, sales reps probably know that information, maybe in some cases the SDRs. Do you think that’s true, or would you say something different? Ariel Cohen (10:24) Yeah, I would completely agree with that. Definitely for those larger orgs. Like, the people at the - at the edge that are closest to to the market and to what people are saying, they know. They also know what a good account looks like and what a bad account looks like, and they could propagate that information back up to marketing and and look at positioning and everything. And I think the the key to success here is to have good data foundations that allow everybody that wants, or that is allowed to, to access that information and and gain insights and and use it. For example, I analyze all of our sales calls. I ingest them every day into our brain. I derive insights upon them, and I get those fed to me every day so that I know, okay, like, this content is working, this content is not working, maybe our positioning sucks, maybe it’s good. All kinds of stuff. And so I keep my my like my finger on the on the pressure bar to try and be as close as possible to to the to the sales calls. I also do demos myself, but that’s because we’re we’re small. Brian Carroll (11:32) Right. And that makes a lot of sense. And something you mentioned: you talked about a shared GTM brain. And what does that change about how your team works together? Ariel Cohen (11:43) Yeah, so I think at the most simple level, it allows everybody to use the same information. Both for for humans and for agents. Right? So I’ll try to explain this through a bad example. Yeah, so imagine you have a company where you have ten reps, ten sales reps, and then you have your AEs, and then that’s like the sales department, and then in in customer support or in marketing, you might be working off different different - like, even skills, like Claude skills, or different - Brian Carroll (12:12) Mm-hmm. Ariel Cohen (12:13) - different knowledge, maybe different SOPs, right? And so when a salesperson now spins up a workflow, does his agent know about the change in positioning that we did last week? Or d - does it know about the the value proposition and the competitive analysis, and are we even doing the right thing? So that’s kind of like a lot of the mess that we’re seeing in larger orgs especially. And when you - Brian Carroll (12:37) Definitely. Ariel Cohen (12:37) - have that shared foundation, everybody could could act off of that with the same context, and context both for agents and also for humans. But yeah, that’s a game changer. Brian Carroll (12:49) If you were to think about how having this brain affects your day-to-day, what does your day look like as a GTM engineer? How might this translate for someone else that you’ve worked with once they have a GTM brain? Ariel Cohen (13:05) Well, that’s an interesting question. Day in the life. Let me think about that one. Well, first - Brian Carroll (13:09) Yeah. Ariel Cohen (13:09) - of all, I would say that one of the first things that I like to do in the morning is create content. I create all my content manually. Usually I start with thinking about maybe scribbling something and then kind of like finalizing it with with Claude. But yeah, that’s what I like doing in the morning. Then it really depends on the - it really depends on the day, but I’ll check on our workflows. I have agents that check all of my other agents, whatever agents they may be, or outbound agents or ads or, I don’t know, LinkedIn content and signal agents. Brian Carroll (13:43) Mm-hmm. Ariel Cohen (13:44) So I’ll take a look that everything’s working and nothing broke. I’ll look at our numbers. So yeah, that’s kind of like a - the monitoring aspect. And then I’ll also initiate things. I really like mini campaigns. So like to run these small experiments to see if they work, and then if they do, then scale them. Okay, so I’ll try to give you more more concrete examples on the brain specifically. Brian Carroll (14:05) Yeah. Ariel Cohen (14:06) Yeah, so first of all, our brain contains all of the account’s history and the client’s history and like the customer’s history. So - Brian Carroll (14:14) Mm-hmm. Ariel Cohen (14:15) - that gets aggregated, and I have all that. So if I’m prepping for a call, I just talk to Swan or my - my agent, and it knows all of that. That’s me pulling information out of there. And then also I can I can see the skills of everybody else in the org. We all work on them together. So - Brian Carroll (14:37) Mm-hmm. Ariel Cohen (14:38) - that’s that kind of shared foundation. Jeremy, he’s our VP sales, he will update like the reach-out skill, which I also use. So I’ll see what he did and and it’s working in all of our agents or outreach agents. We’re using it for that. So like the shared skills part is is one. Then another one is when I initiate outreach, so I can create incredibly personalized outreach at scale with our our account history. So when I launch campaigns, it knows to say, “Hey Brian, we chatted back in February. I saw that a few things changed, that you just launched a new product or or something. Is now a better time to talk?” Right? So things like that that take the whole context and history in account. Those are two z - two good examples, I think. Brian Carroll (15:26) Yeah. And as I’m listening to you, I wondered what’s surprised you since you started doing this, or maybe something you could share that was harder than expected? Ariel Cohen (15:38) I’ll start with what what is hard. So hard - Brian Carroll (15:41) Yeah. Ariel Cohen (15:41) - is tracking data and marketing. Yeah, that has been an ongoing challenge in the initial first few months. Like what do we even track? What’s important to track? What is an MQL? Like we also changed that a few times. Brian Carroll (15:59) Mm-hmm. Ariel Cohen (16:02) Yeah, so I spent a lot of time on like dashboarding and bringing everything to to a place where we have our scorecards, our metrics, we go over those and we’re happy and we know that the business is going well, the business isn’t going well. We have enough demos, we have enough qualified ops, we’re not. Yeah, so that was - that was a hard part. Or at least it it took a lot of effort to get - to get right, because it’s also a people thing. Yeah, so the foundation of every successful AI implementation in an a company, no matter what the size and no matter what the AI product, is having tight processes. Brian Carroll (16:37) Yeah. Hmm. Ariel Cohen (16:41) And that change management problems and challenges - and yeah, that could be the challenging part. And being very disciplined to follow those, you know, the SOPs and the standards that we set for ourselves is not easy. And it requires agreeing across the different - like sales and marketing and CEO and product marketing - and it it requires kind of negotiating and agreeing and and seeing what the best way of doing things or what everybody agrees the best way of doing things is, and then aligning the agents on top. Brian Carroll (17:13) And so as you’re working with the team and and getting processes and doing this work, I was thinking you have this upcoming talk for 10x and you’re gonna give marketers advice from your own experience, you know, both what worked and what didn’t work. And I wanted to ask you for for listeners out there that really want to embrace this, where do you think they should start? Ariel Cohen (17:38) So, what worked and what didn’t work. So what worked was initially tracking everything - Brian Carroll (17:44) Mm-hmm. Ariel Cohen (17:45) - and then understanding what works the best and calibrating, maybe removing a few things, and and and iterating. So there - the faster you iterate, the the better - the better it’ll be, I would say. And if you don’t iterate fast enough, things stop working, and, you know, we’re in a crazy market. Like everything is changing all the time. Buyer psychology is is changing fast. The whole product landscape is changing, so you just have to keep iterating. That’s the only thing that works, to be honest. My advice might not be relevant in a few months from now. Brian Carroll (18:16) It - I know that’s true with how AI is evolving and the change, the the speed of change that you talked about. And I wanted to ask you, was there something else as as you thought about our conversation today that you wanted to share or that I should have asked about? Ariel Cohen (18:33) So if we’re talking to marketers, I think that marketers have a very big pain point that they don’t know that is solvable around events. And I discovered this when we went to our first event, and - Brian Carroll (18:45) Mm-hmm. Ariel Cohen (18:46) - suddenly I saw everything that we can do once we have our processes and we have a system. I would say that, like, the whole lead capture and the outreach before events and the post outreach, it doesn’t have to be such a pain. It - there is a a a better way to do this. And we’ve gone into our our first event, and we’ve - we - you know, it wasn’t easy in the beginning, but then we just got better and better and learned really fast while doing it, and we got to a situation where we’re ingesting leads by talking on WhatsApp to Swan, or like our pre-event outreach really booked us a whole bunch of meetings because we reached out to like systematically with a correct context. We track event ROI after the event also in a systematic way, using like the agentic platform. We to get good ROI on on events, and I know that events are getting hotter now because people want to see people, and they’re kind of sick getting cold outreach. So we did book quite a lot of meetings before the event, but it wasn’t only that. It was how we ingest leads while we’re at the event. So you have those - Brian Carroll (19:57) Mm. Ariel Cohen (19:58) - those barcodes that you scan and then you have to do something with, right? I hate those barcodes. What a useless app. Brian Carroll (20:03) Right. I think everyone does. Ariel Cohen (20:06) Yeah, so instead of doing that, I just - I wanted to create a skill. So like we didn’t have a process for this. So I was like, okay, what do I need? I need to know - I I need to have that the person’s company and, like, whether they’re ICP or not. I need to have their LinkedIn and their email. I need to have the context of our conversation and what did they ask for? Did we book - Brian Carroll (20:26) Mm-hmm. Ariel Cohen (20:26) - a demo? Did they say, “Not now, talk to me next quarter”? Did did they even like it? Was it relevant for them? Right, and so I found myself using only the - their LinkedIn link. And then blah-blahing, like, just yapping all of my context into a voice note with that LinkedIn link. I told Swan, “Hey, this is the person I talked to, this is what they said, blah blah blah blah blah blah blah. Write it into our brain, and schedule outreach next week.” And that’s it. And then I said, “Wow, this worked really well. Let’s turn this into a skill.” And then - Brian Carroll (21:01) Yes. Ariel Cohen (21:02) - I told Jeremy, “Yo, like, you use this as well. Don’t use those barcodes.” And then the whole ingestion process got better. And then our - like, we converted more, and we were able to to consistently follow up more and not have to log in leads, you know, after the the dinner that we had. So yeah, that was - that worked well. Brian Carroll (21:22) I love that. And were you asking people for the LinkedIn URL, or were you able to use a skill where you just got their name and then your AI was able to reverse look that up? Ariel Cohen (21:32) Really need to know the company, and I just used a link because it’s a one-to-one match, so you don’t lose accuracy on that. Brian Carroll (21:38) Very cool. Well, this has been a great conversation. Every time I talk with you, Ariel, I feel like I learned something more. And I I know our listeners did too. And hearing how you’re doing this in practice and that you’re teaching from your own experience. So I wanted to say thank you for being on the show with us. Ariel Cohen (21:55) Sure. Welcome. It’s great chatting. Brian Carroll (21:56) Thanks, everyone, for listening to the B2B Roundtable. We’ll see you next time.

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Hosts and Guests

Nobody is credited on this episode yet.

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. Why GTM Alignment Breaks Down and How to Fix It with Martha Mathers

    Oct 6, 202626 minAudio

    About This Episode Everybody talks about GTM alignment. Sales, marketing, and customer success all working together around the customer. But actually making that happen changes how teams work together, and that’s where...

    Credit the people on this episode

  2. Stop Automating the Past: How AI Is Changing the Way We Work with Liza Adams

    Sep 14, 202638 minAudio

    About This Episode Most GTM teams are using AI to make existing work faster. But faster isn’t the same as different. And the teams that are actually pulling ahead aren’t automating the past - they’re reimagining the...

    Credit the people on this episode

  3. AI Scaled Activity. It Didn’t Scale Relevance with Craig Rosenberg

    Aug 18, 202631 minAudio

    For years, B2B teams tried to make SDRs more efficient. More calls. More emails. More activity. Then AI arrived, and many teams tried to automate the same model. But what if the problem wasn’t needing a faster version...

    Credit the people on this episode

  4. Why GTM Rebuilds for the Enterprise Start With Clarity, Not Campaigns, with Corey Livingston

    Jul 28, 202631 minAudio

    A company reaches a key moment. The strategy changes, the target buyer shifts, and leadership decides to focus on the enterprise market. But the go-to-market system underneath the business may still be built for an...

    Credit the people on this episode

  5. Why B2B Buyers Really Buy: The Hidden Buyer Journey with Scott Gillum

    Jul 14, 202624 minAudio

    Most GTM teams have gotten very good at tracking what buyers do. They can see form fills, intent signals, CRM activity, demo requests, and pipeline movement. But those systems often miss the people, pressure, risk,...

    Credit the people on this episode