-
769
01:57:42.430 –> 01:57:55.609
Julia Nimchinski: All right, we are transitioning to an all-star CXO Roundtable, and excited to welcome Mark Niemek, CRO of SalesLoft, to lead the roundtable on the agent-driven revenue framework.770
01:57:55.850 –> 01:58:04.859
Julia Nimchinski: With over 20 years in enterprise sales leadership at Cisco and Salesforce, Mark now leads global revenue team at SalesLoft.771
01:58:04.960 –> 01:58:14.839
Julia Nimchinski: And it’s helping companies to close more, forecast accurately, and coach teams to excellence. Mark, super excited to host you here. How are you doing?772
01:58:15.040 –> 01:58:18.460
Mark Niemiec: Awesome, Julie. So nice to see you, thanks so much for having us.773
01:58:19.670 –> 01:58:27.719
Julia Nimchinski: Yeah, this is a very long-awaited panel, an anticipated one. Let’s do a quick round of introductions.774
01:58:27.870 –> 01:58:28.860
Julia Nimchinski: Everyone.775
01:58:29.970 –> 01:58:31.090
Julia Nimchinski: James Roth.776
01:58:31.530 –> 01:58:42.419
James Roth: Sure. Hey, Julia, great to see you here, see you. It’s great to be here. Mark, always, always good to join with a friend, so excited for this panel. I’m the CRO at ZoomInfo.777
01:58:46.120 –> 01:58:47.869
Julia Nimchinski: Hannah. Yeah. Welcome.778
01:58:47.870 –> 01:58:52.469
Hannah Willson: to see all of you excited for this panel. I’m Hannah Wilson, I’m the CRO of Nooks.779
01:58:54.350 –> 01:58:55.490
Julia Nimchinski: Yeah, Senti.780
01:58:56.010 –> 01:59:00.609
Santi @ Momentum: Hi everybody, I’m Santis Farra-Saldonez, I’m the co-founder and CEO of Momentum.781
01:59:02.980 –> 01:59:03.900
Julia Nimchinski: Angela.782
01:59:04.600 –> 01:59:13.339
Angela Winegar: Hey everyone, excited to be here. I’m Angela, I lead growth and marketing at Invisible Technologies, and excited to be the solo marketer on the CRO panel.783
01:59:14.650 –> 01:59:15.590
Julia Nimchinski: And Aaron.784
01:59:19.000 –> 01:59:23.619
Julia Nimchinski: Gonna bring Erin onto the show. Boom.785
01:59:23.790 –> 01:59:25.080
Julia Nimchinski: Erin, welcome!786
01:59:26.510 –> 01:59:28.139
Julia Nimchinski: Isabree joining us?787
01:59:28.530 –> 01:59:29.740
Julia Nimchinski: Live TV.788
01:59:36.180 –> 01:59:36.680
Aaron McReynolds: Thanks.789
01:59:36.680 –> 01:59:38.079
Julia Nimchinski: Right on time.790
01:59:38.520 –> 01:59:44.110
Aaron McReynolds: Nice to be here. Nice to see everyone. Aaron McReynolds, co-founder, CEO of Alicia.791
01:59:45.430 –> 01:59:47.859
Julia Nimchinski: Amazing, what a panel. Mark, what’s in your mind?792
01:59:48.030 –> 01:59:59.399
Mark Niemiec: Awesome. Well, thanks, Julia, for helping everybody get introduced, and folks for being here today. You know, the topic that we’re going to discuss amongst the group here is AI and agents.793
01:59:59.400 –> 02:00:10.199
Mark Niemiec: And, in a couple different ways, you know, how we’re building those agents and those capabilities into our own organizations and our own, you know, go-to-market models, how we’re bringing some of those things to our customers.794
02:00:10.200 –> 02:00:23.639
Mark Niemiec: You know, how we’re making sure that they’re effective and efficient in delivering the value that, or the promise of AI, and maybe some tips and tricks from the panelists around how they’re, how they’re driving their organizations forward.795
02:00:27.910 –> 02:00:28.520
Santi @ Momentum: Exciting.796
02:00:28.520 –> 02:00:29.160
Mark Niemiec: Or…797
02:00:29.670 –> 02:00:54.660
Mark Niemiec: Awesome. All right. Well, thanks everybody for being here. Maybe just to get us started a little bit, we’re… the group of us agreed we’re not necessarily going to talk our books. We’ve all got offerings in the space, and we wanted to make this more of a conversation about what we’re doing as leaders to make sure that our organizations are becoming AI-first and embracing the agenda capabilities that both we’re developing for our customers, but also some of the experimentation that we’re doing internally and learning within798
02:00:54.660 –> 02:01:04.579
Mark Niemiec: with our own organizations. One of the things we believe is that the best ideas are coming from our teams, and all of us selling, you know, sort of go-to-market capabilities.799
02:01:04.580 –> 02:01:18.320
Mark Niemiec: Our subject matter experts are our employees, and so we have a great opportunity to learn from them as we teach them, but also then to collaborate and cross-pollinate great ideas that we find within our organizations. They oftentimes help us800
02:01:18.320 –> 02:01:22.720
Mark Niemiec: Come up with new go-to-market capabilities, new products, and new ideas.801
02:01:22.720 –> 02:01:46.640
Mark Niemiec: And so one of the things that we wanted to talk about, or start off on, is how we’re using agents and AI inside of our own organizations, and how we’re developing our own teams. And so maybe get the team started off with an easy question. You know, we’ll start off with, you know, how do you make sure that everybody inside your organization is leveraging the power of AI, not just the802
02:01:46.640 –> 02:01:58.980
Mark Niemiec: the few that are your top performers, but how do you get everybody moving? And Santi, you’re on my screen, maybe we’ll start with you. How do you make this an everybody approach versus just a handful of folks?803
02:01:59.330 –> 02:02:00.750
Santi @ Momentum: Yeah, so,804
02:02:01.490 –> 02:02:13.129
Santi @ Momentum: We are a… we’re a vendor in the AI space, so we’re very much AI-forward. We’re also a small company, so we see AI as kind of that edge that could allow us to overcome a bigger company out there we’re competing against.805
02:02:13.130 –> 02:02:21.960
Santi @ Momentum: So I think about this all the time. In fact, I’ve been interviewing my own customers and other vendors to see what best practices they see, and a pattern that I really like806
02:02:22.080 –> 02:02:36.900
Santi @ Momentum: is the idea that leadership is a key driver for transformation on this front, and that literacy of AI is essential at the top. So having CROs, CEOs, CTOs that deeply understand807
02:02:37.110 –> 02:02:52.330
Santi @ Momentum: the different tools and technologies at their disposal. So if you’re talking to a CTO, it’s probably Codex and Cloud Code. If you’re talking to a CRO, you’re looking at the offerings in the sales tech space to see what are your options. It is actually a counterintuitive808
02:02:52.410 –> 02:03:04.379
Santi @ Momentum: approach. You know, most CROs are really good at delegating. As you start going to the top of the org, you start putting in your lieutenants and the front lines to be the ones exploring, but the depth of transformation that AI can deliver.809
02:03:04.380 –> 02:03:17.170
Santi @ Momentum: really needs to penetrate all the way to the top. So that’s my… my personal advice, is get after ChatGPT, create yourself a Gemini account, go in 11 Labs and generate some voices. I think it can be really powerful for the whole org.810
02:03:17.720 –> 02:03:19.699
Mark Niemiec: Love it, love it. Hannah, how about yourself?811
02:03:20.110 –> 02:03:45.089
Hannah Willson: Yeah, I would totally agree with that. I’ve started doing some of that myself, and I feel like now, more than ever, I get so much value from these networking events, because I get to hear what others are doing, and I went to an event a couple weeks ago, and I heard the CRO of OpenAI there, and she was talking about how she built a chat GPT so people could, you know, send emails on her behalf, and they sounded like her, and you know, just even something small like that, I totally agree with you, Santi. It kind of sets the tone. If you’re gonna go and create something.812
02:03:45.090 –> 02:04:04.210
Hannah Willson: It, you know, serves as a good example for your team of what can be done, in addition to hiring great people. Like, I think it’s absolutely critical that we have folks on our team, whether they’re in RevOps or go-to-market engineering, that can build things quickly and that are meaningful for the rest of the org. So I think hiring is also really important for these skill sets.813
02:04:06.060 –> 02:04:29.880
Mark Niemiec: Yeah, for sure. We’re definitely seeing folks come into the hiring process with some skills and experience, different than we’d seen maybe 18 or 24 months ago, and we’re definitely seeing folks who have spent time in learning their own frameworks and have their own tactics and techniques. One of the people that introduced, the person that introduced Santi and I is Cal Norton, who many of us listen to and read his content, but he’s all about experimenting.814
02:04:29.880 –> 02:04:33.339
Mark Niemiec: And bringing these capabilities to the edge. So, he’s been a great example of that.815
02:04:33.340 –> 02:04:56.770
Mark Niemiec: Maybe get a little more specific and maybe ask some other folks on the call, what specific programs, maybe, has anybody implemented, in terms of bringing these capabilities to your team? Anything around forecasting, pipeline generation, outbounding, things like that. We’d love to hear how you’re bringing these capabilities to your team with more specificity. I don’t know, Aaron or James, maybe you’ve got something to share with us.816
02:04:57.420 –> 02:05:12.339
Aaron McReynolds: Yeah, yeah, you know, one of the things that I encouraged our team, probably 4 or 5 months ago, was, you know, don’t request headcount, request more tokens, right? And what that kind of enabled them to go and do is, the minute they got into a bind or a pinch.817
02:05:12.340 –> 02:05:35.479
Aaron McReynolds: felt overwhelmed, instead of always coming and saying, hey, we need to hire another engineer. It’s like, go find the applications, test them, budgets are there, and come back to me with the results. And I think kind of enabling and empowering the team to think this way, you know, we’ve seen a great shift, even, you know, a head of product jumped in the other day, got himself a few accounts, and he’s just now writing code. And he wasn’t doing that 6 months ago.818
02:05:35.480 –> 02:05:43.740
Aaron McReynolds: And so I think giving them the freedom and conviction to go and do it, try those tools, and implement it in the daily workflows is really helping us right now.819
02:05:46.260 –> 02:05:48.779
Mark Niemiec: Sure, yeah, I love hearing that. Go ahead, James.820
02:05:48.780 –> 02:05:58.869
James Roth: Yeah, I think… I think just kind of a touch on each of them. I think one of the biggest that we find when this first started, call it a year ago, I felt like we were the…821
02:05:59.620 –> 02:06:12.489
James Roth: C-suite that Santi was mentioning, we were telling everybody to use AI more, and it was more of a standing on a pedestal, telling, you know, I hate to admit it, but kind of like the old guy in the room now, like, use AI, kids.822
02:06:12.550 –> 02:06:28.229
James Roth: And, I think it wasn’t until we really started leaning into it as an executive team, and I think the biggest way we’ve driven it internally, and yes, we’re in the space too, I think we’re all trying to sell our own, call it agentic solutions, whether you’re AI-native or not.823
02:06:28.230 –> 02:06:40.199
James Roth: I think there’s a combination of things that we did that really started to drive internal adoption, and I think the biggest one is we created an incentive for solving problems with agents that we build internally.824
02:06:40.200 –> 02:06:49.779
James Roth: And when we started doing that at the executive level, I looked at my day-to-day, and I would say, everybody on here has probably dealt with this to some degree, Mark, especially in very large orgs.825
02:06:49.990 –> 02:07:08.219
James Roth: Probably 50% of my inbound slacks across a team of 2,000 GTM folks is, do you know this person? You’re connected on LinkedIn, do you know them? And I think 5 to 8 years ago, there was probably a connection where I actually did know them. Now, it’s completely, you know, I probably don’t know 90% of them.826
02:07:08.220 –> 02:07:27.610
James Roth: And so we created an agent that goes through first-party data, goes through calls, goes through emails, calendar, like, have I ever met with this person? And then it takes a variety of things, from job history, location, alma mater, so on and so forth, and it creates a relationship score. And again, this is not a ZoomInfo product we’re selling, so I promise this isn’t a pitch. This is truly just a problem.827
02:07:27.610 –> 02:07:33.799
James Roth: Where we knew we had the data across different systems, and we could at least Quiet the noise on…828
02:07:33.830 –> 02:07:45.099
James Roth: 80% of these slacks to say, James and Mark, unequivocally, they know each other. They’ve met 9 different times, they work in the same space, you know, all of these different data points, so it creates this829
02:07:45.100 –> 02:08:04.649
James Roth: James agent, so they can ask that. It spits out a score of, like, an 8 to 10 based on all of those attributes, and then it can go into the James emailer, if you will. And so I think it wasn’t until we started solving problems like that internally, and then getting them out to the team so they could see it, that trickle-down effect of, okay, what’s the most annoying part of my day?830
02:08:04.650 –> 02:08:17.279
James Roth: what is the least efficient part of my day, and having people solve those with agents they can build with relative ease. You know, that’s kind of the carrot. We started then incentivizing, we gave a bounty. If you have an agent that gets built.831
02:08:17.280 –> 02:08:29.139
James Roth: You know, if you’re an account manager, and you build an agent that does great prep, or great historical context, or great point of view, and then people start using that, there’s, like, an adoption and utilization threshold internally.832
02:08:29.140 –> 02:08:48.010
James Roth: once the company starts using that agent that that individual has built, we give them a SPF, we give them recognition, you know, all of those different things. That’s the carrot side of it, and I think now, in a world where AI can do a lot of the things that make your great salespeople, your great salespeople, like account planning, like account briefs, like prep, point of view, so on and so forth.833
02:08:48.350 –> 02:08:52.429
James Roth: The stick is, if folks aren’t leveraging those agents.834
02:08:53.450 –> 02:09:04.940
James Roth: I think it’s probably the only time in my career at least, there is… there is not a compelling answer as to why they would not be using that. It’s like trying to hit home runs against Mark McGuire back in the 90s.835
02:09:05.040 –> 02:09:15.469
James Roth: not using performance enhancers. So, you know, we’ve had a very aggressive stance on folks that are not using those agents for those particular use cases.836
02:09:15.880 –> 02:09:34.960
Mark Niemiec: I love that. The carrot and stick and the story around the connection, I think all of us get that question all the time, do you know this person? That was a great tutorial for all of us. Angela, how about you? From a marketer’s perspective, how are you thinking about bringing these capabilities to your team and putting some structure around it in the marketing side of the business?837
02:09:35.160 –> 02:09:49.430
Angela Winegar: Yeah, absolutely. And maybe I’ll just, like, dovetail into what James was sharing as well, and about how… how you incentivize learning, because I think it’s true across sales and marketing, but something we’ve seen be really successful is,838
02:09:49.430 –> 02:10:07.500
Angela Winegar: having people sit down and have, you know, one-on-one sessions where they go through things together. So, you know, to give an example of that, we had a marketing ops manager who was helping pull lists for an event, for our events manager, and the SER who was doing all the pre-event, you know, outreach and meeting setting.839
02:10:07.500 –> 02:10:30.690
Angela Winegar: And it was really interesting because our marketing ops manager used… no one in that group is technical, but used AI to generate a script that scrubbed the event page of, like, a thousand event attendees, and then got it auto-enriched. And it was funny, because we attended the same event last year. It took that team about two days to enrich that entire list, and, like, even just get the list of event attendees.840
02:10:30.690 –> 02:10:32.510
Angela Winegar: And…841
02:10:32.510 –> 02:10:56.090
Angela Winegar: having the marketing ops manager sit down with the events manager and the SDR, who were doing all of this pre-qual work, and just being like, here’s how I got it to generate a script. Like, you know, running a script on a page is not a GenAI workflow per se, but Gen AI is enabling non-technical people to run these really basic things that then, you know, take two days’ worth of work and turn it into less than an hour of work.842
02:10:56.100 –> 02:11:04.239
Angela Winegar: And so we have, basically, what has been really interesting to me is that, you know, Mark, you asked in your question, what’s a structured way of doing this?843
02:11:04.240 –> 02:11:28.560
Angela Winegar: the most effective way we’ve seen has been unstructured learning, of really just sitting down next to someone, watching them do it. And so what we’ve been trying to do is create a structured environment to celebrate those things and elevate them so that we encourage more people to have that kind of, organic learning, I would say. So it’s been really interesting of… I think for a while, I think maybe, you know, 12, 18 months ago, we were trying to do a lot more structured learning on it.844
02:11:28.580 –> 02:11:36.639
Angela Winegar: But it really is, you know, thinking back to early in our careers, I’m gonna date myself here, also as, you know, the old person in the room now, but…845
02:11:36.930 –> 02:11:53.420
Angela Winegar: I started my career as a consultant. The first 6 months, everyone knew your first 6 months, you were useless, and you just sat next to a senior associate, and learned how to do Excel, and learned how to make slides, and do all the other kind of workflow things, until one day you weren’t useless anymore.846
02:11:53.420 –> 02:12:02.800
Angela Winegar: And Gen AI is really interesting in that it’s really making us all think back to that beginner’s mindset, and be like, wow, I just need to sit down next to someone who knows how to do this.847
02:12:02.810 –> 02:12:14.200
Angela Winegar: in my case, it’s almost always someone more junior than me, but I’m learning from them, and how do I start to adopt that technology in that, you know, kind of organic use case-by-use case way? And I think that’s true across sales and marketing.848
02:12:14.620 –> 02:12:22.939
Mark Niemiec: I love that. Yeah, like, a bunch of comments about old people, I think we’re all roughly the same age. I don’t think I ever felt like the old person in the room, but maybe… maybe we’re finally there.849
02:12:22.940 –> 02:12:47.379
Mark Niemiec: I’ll tell you what we’re doing at Sales Loft, and then we’ll kind of jump into the next part of the conversation. You know, what we’ve done is we’ve had these, what we call these co-lab sessions, so collaboration, a play on collaboration and laboratory, where our teams are putting together what they’re doing in the market, and they’re sharing them with their peers and colleagues. And so, how are we creating tighter points of view? How are we doing better discovery? How are we doing better850
02:12:47.380 –> 02:13:11.419
Mark Niemiec: more on-point presentations. I think you take some of that together with some of the incentives that James talked about, and there may be that example of, you know, what does the unstructured structure look like together? Because there’s a lot of peer-to-peer sharing in these sessions that we do bi-weekly. We amplify and share them all across the company, but what we haven’t done is put the incentive structure in the system like they did at ZoomInfo, which I think is a super851
02:13:11.420 –> 02:13:15.550
Mark Niemiec: exciting idea for all of our employees, so that’s awesome.852
02:13:16.500 –> 02:13:41.450
Mark Niemiec: All right, so moving on a little bit in terms of AI time to value. Love to talk a little bit about what’s some of the ideas out there in terms of misconceptions around time to value. We’re hearing some companies talk about, you know, building agents, and it’s taking, you know, 6, 12, 18, 24 months to get to value. That certainly seems like a long time when we’re seeing some companies get from zero to, you know, $500 million in a year. -
853
02:13:41.450 –> 02:13:56.590
Mark Niemiec: in the AI space, and nobody’s waiting 18 months for those capabilities. So, we’d love the group’s perspective on, you know, what is the… what are some of the misconceptions about how long it takes to get value out of these AI capabilities? And Aaron, maybe we’ll start with you.854
02:13:58.280 –> 02:14:03.530
Aaron McReynolds: Yeah, you know, I think, you know, when we started to release our product, and we’re working on855
02:14:03.530 –> 02:14:19.490
Aaron McReynolds: interviewing CROs and VPs of RevOps, you know, this was one thing that just kept coming up. We’ve all seen, you know, the MIT study, 95% of POCs failing, and I think, you know, there’s kind of two pieces to it. One is the limitations of the technology itself.856
02:14:19.490 –> 02:14:42.209
Aaron McReynolds: And secondly, the implementation, and kind of how those expectations are dealt with. And so for us, you know, we don’t have CSMs right now, we’re a small startup, I am the CSM, and so I get to go in and sit down, watch the onboarding flow take place, see the type of questions and workflows that our customers want to see in their organization.857
02:14:42.210 –> 02:14:45.359
Aaron McReynolds: And what I’m finding, and I don’t know how everyone else feels.858
02:14:45.360 –> 02:15:07.330
Aaron McReynolds: But the expectations, and I think the reality right now, there is still a gap. And I think that’s why it does fail in some instances. And this is not, you know, in using our product, but in testing other products as well, of what are we really trying to get out of these POCs? What are those, you know, achievements? What are the KPIs that say, yes, this did work? But what I’m seeing is the appetite, you know, we always talk about.859
02:15:07.330 –> 02:15:08.980
Aaron McReynolds: Dog food, right?860
02:15:08.980 –> 02:15:26.340
Aaron McReynolds: the ability to just put anything in front of a dog, it’s gonna consume it, it wants it, it needs it, it doesn’t care about the protein mix, etc. And that’s what we’re seeing, is people are just craving for the technology to work, and so I think being able to go in, and I know most people on the call do this, actually sit with the customer.861
02:15:26.340 –> 02:15:40.470
Aaron McReynolds: figure out the expectations before launching into a POC. But then being able to, you know, essentially hold the hand for as long as possible until the outcomes are being delivered, and that’s something, you know, we’re obviously keeping a close eye on and tracking internally.862
02:15:41.550 –> 02:15:54.889
Mark Niemiec: For sure. What are some of those key KPIs that we should be thinking about as we start to put some, you know, some performance capabilities around these agents? What are some of the KPIs you all look for?863
02:15:58.250 –> 02:16:23.220
Hannah Willson: I mean, I think there’s a range of things we look for, and I would say it’s a pretty… it is a pretty big range, because there’s some things where you can see time to value really, really quickly. Like, in a matter of days, you start seeing, you know, more pipeline. There’s other things that might take a couple of months, so I think it depends on exactly what you’re looking at, and then the KPIs can be different, too. But I would encourage everyone, I mean, we really try and look at not just sort of, like, input metrics in terms of, like, activity864
02:16:23.220 –> 02:16:25.970
Hannah Willson: But actually more of those, like, output metrics.865
02:16:25.970 –> 02:16:50.010
Hannah Willson: Like, is it actually driving pipeline for us? Is it actually driving revenue for us? So beyond some of, like, are people using it? Do they like it? Do we actually think it has a tie back to revenue? And, like, you know, when we talk to vendors, really diving into that. And then also, when we’re looking at time to value, I know at my previous company, we tried a lot of different solutions in the early days, and some of them we actually had really successful POCs with, but then, like, the usage and the…866
02:16:50.010 –> 02:17:09.330
Hannah Willson: excitement kind of tapered off over time. So I also think just, you know, the other side of the coin to time to value, and I have conversations like this with, you know, people looking at our solution, is, okay, well, what’s going to happen in 3 months from now and 6 months from now, and having those proof points and customer stories is just as important, and when you’re looking at solutions, to ask for those kinds of things.867
02:17:11.080 –> 02:17:13.679
Mark Niemiec: For sure. James, I saw you come off mute.868
02:17:14.120 –> 02:17:20.970
James Roth: Yeah, I think to the MIT study in the 95.5, you know, I think two things. The first one…869
02:17:21.690 –> 02:17:28.479
James Roth: in terms of the time to value, and in terms of getting value out of just about any of these tools, you know, I do think870
02:17:28.580 –> 02:17:38.570
James Roth: and Aaron, I know you guys are very close on this as well, you know, the data foundation, getting that right, where a lot of these tools will sit on top of CRM, or sit on top of871
02:17:38.680 –> 02:17:50.050
James Roth: you know, some data that might not be ready for AI, if you will, and I think getting that data foundation right across your data lake, across CRM, first party, third party, so on and so forth, I think872
02:17:50.080 –> 02:18:08.010
James Roth: before folks can really hit that warp speed of getting value from all of these things, they’ve got to start there, and I think over the past year, we’ve all been on the other end of that really cool whiz-bang AI use case. Somebody on your team sees it, they show it around the org, everybody gets excited about it.873
02:18:08.010 –> 02:18:14.770
James Roth: They deploy it, and if the data foundation’s not right, then ultimately you don’t get that expected outcome.874
02:18:14.870 –> 02:18:25.820
James Roth: You know, I think on the metrics, once you do have that foundation right, you know, we’ve got a framework internally where there’s really 3 buckets that AI should be driving an outcome. It’s either headcount.875
02:18:26.080 –> 02:18:30.829
James Roth: headcount reduction. You know, increased efficiency, you know.876
02:18:31.190 –> 02:18:43.100
James Roth: major metrics, if you will, and then cost savings, which kind of goes into… into the headcount side, but I think cost savings can be more than just reduction of headcount. And so, I think…877
02:18:43.150 –> 02:19:02.529
James Roth: what we see is it’s not always that first metric that is the most important. Like, one of the most obvious is time saving, and I think that was the first 6 months of the AI boom, it was all about time saving, and so we’ve got an agent internally that we can go, any rep, any team, and say, show me all of the things878
02:19:02.530 –> 02:19:05.990
James Roth: all the time that they’ve spent. And so we have this use case.879
02:19:06.379 –> 02:19:26.380
James Roth: There’s an individual, he’s an enterprise account manager, he’s got 15 accounts, all large enterprise, and you can say, okay, if one of these accounts, show me what he’s done. 35 hours of actual face-to-face meetings, or Zoom meetings, and then 85 hours of prep, docs, decks, it takes all the data from, you know, Google Docs, all the decks he’s built in PowerPoint, so on and so forth.880
02:19:26.480 –> 02:19:38.229
James Roth: And, you know, that 85 hours of prep work for 35 hours of meeting, yeah, to cut that down to 10 with all of these generative AI use cases where they can get the account plan, they can get the deck, they get the point of view, great.881
02:19:38.250 –> 02:19:53.420
James Roth: But then what does he do with those 75 hours? And I think, like, the initial, great, this guy, his name’s Josh, it’s the example we always use, this guy Josh saved 75 hours, now what? If Josh is not allocating those hours to other, more productive things.882
02:19:53.420 –> 02:20:11.049
James Roth: now Josh just has more time and a better work-life balance. I’m not against that, but I think that’s where, when people deploy some of these products, or they go implement a bunch of these products with time savings as the only KPI, they’re missing when their CFO comes calling, saying, why are we spending two to three million dollars more883
02:20:11.460 –> 02:20:22.390
James Roth: But we’re not seeing any meaningful output change. We’re just seeing time savings. And so it’s basically, like, what is that? As you’re procuring or as you’re piloting any of these products, it’s like, sure.884
02:20:22.490 –> 02:20:42.240
James Roth: We’ll get those 75 hours back, and then we’ll see calls increase. We’ll see the 35 hours of actual in-person meetings increase. We’ll also see the ability to touch more of those 15 accounts increase, so potentially we could go from a world where I think most of us in the enterprise try to shrink account loads, so you can have better relationships, and closer to the customer, and you want to get account loads down.885
02:20:42.240 –> 02:20:59.790
James Roth: In this new world, can we take account loads up? So now, we’re not only getting more productivity in terms of call output, in terms of meetings, in terms of driving upsell growth, renewal, so on and so forth, but now we can also give this guy 5 more accounts, because he doesn’t have to waste 75 hours on each of them in just prep work.886
02:20:59.870 –> 02:21:18.970
James Roth: And so, now we are not only hitting output, we’re also hitting cost savings and headcount reduction, because we don’t need to go hire 100 more enterprise AMs. So, I think it’s really important to track the initial thing that you’re looking for, and then what’s that gonna yield, so that we all, as CROs or CMOs, Angela in your case, welcome.887
02:21:18.970 –> 02:21:26.409
James Roth: CMO to the group, you know, we can stand taller in our investments by showing the actual outcome, not just the time saving.888
02:21:27.060 –> 02:21:43.609
Mark Niemiec: Yeah, yeah, no, a lot of people are talking about that time savings. We’ve got to move beyond that into real hard dollars. Santi, you see this from the bridge, given what Momentum does. We’d love to hear your perspective on some of the insights around KPIs and most effective sort of metrics that you’re seeing at Momentum.889
02:21:43.610 –> 02:21:50.419
Santi @ Momentum: Yeah, one thing I wanted to share, kind of riffing off of James’s last bit, and Aaron as well, I think890
02:21:50.870 –> 02:22:02.320
Santi @ Momentum: given that we sell to revenue organizations, I think this is a pattern that is more present on the revenue side of the aisle, even marketing, but not as much as in engineering, because engineering is already pretty staffed up on891
02:22:02.340 –> 02:22:16.849
Santi @ Momentum: technical resources, but one new quality of this new trend is that agentic workflows, agents, this level of sophisticated automations, they’re inherently a technical challenge, and we see that892
02:22:17.090 –> 02:22:34.260
Santi @ Momentum: the time to value, or the ratio of success, tends to be a function of how technical are the resources that you’re putting behind that project, and how much ownership do they have to push it through to get it done. People forget that in the end, you’re building automations, they’re smarter automations, but893
02:22:34.390 –> 02:22:48.830
Santi @ Momentum: the person that builds that needs to think in systems, needs to understand foundational data, makes sure that the data is there, needs to think about scale. So, when we see failures, it’s because you’re grabbing somebody who is not really894
02:22:48.830 –> 02:22:56.030
Santi @ Momentum: that adept to think about systems and processes, and you put them to, you know, magically try to make something work with ChatGPT,895
02:22:56.030 –> 02:23:03.649
Santi @ Momentum: And it tends to flunk, but… but… so this is an interesting pattern. I think moving forward, organizations will start developing their own896
02:23:03.720 –> 02:23:13.619
Santi @ Momentum: ops teams that are a little bit more technical and a little more staffed up from an engineering standpoint to execute those projects, because the projects pay off if you do them right.897
02:23:14.660 –> 02:23:25.609
Aaron McReynolds: For sure. Real quick, if I can, you know, one of the things we’ve seen is, like, even discovery questions have changed, right? We always talk about appetite, preparation, and trust.898
02:23:25.610 –> 02:23:36.620
Aaron McReynolds: Right? As we go into these calls, everyone says they have an appetite for AI, and they want to go and change the world with it, but how prepared are you? To James’ point, like, what is the level of your data? How clean is it?899
02:23:36.620 –> 02:24:01.449
Aaron McReynolds: the AI can only get the results it can get on the preparation of that data, and then trust, and the trust comes into the outcomes, working with the teams to develop those outcomes, and so even from, you know, that discovery standpoint, almost disqualifying early, not everyone is ready, not everyone has that appetite, not everyone can go and just deploy this as quickly, and I think that is having a significant result on the outcomes of a lot of these POCs as well. And Angela, I’m so sorry for cutting900
02:24:01.450 –> 02:24:02.560
Aaron McReynolds: off. Sorry about that.901
02:24:02.560 –> 02:24:26.410
Angela Winegar: No, all good. I was just gonna add to this, because I love this discussion so much, it’s something we think so much about. We promised we wouldn’t talk our book, but Invisible is an evals and annotation platform, and so the KPI would push everyone on is, you know, are your models actually getting better? I think what makes Invisible unique in the space is, you know, the surges and Mercors of the world also work with the major model providers, as do we.902
02:24:26.410 –> 02:24:29.800
Angela Winegar: But we work a lot on the enterprise side.903
02:24:29.800 –> 02:24:41.950
Angela Winegar: And exactly to James’s point, I would say 90% of enterprises that we work with, they actually don’t have any issues with their Gen AI models, they have issues with their data foundation. But once you get your data foundation in place.904
02:24:42.010 –> 02:24:53.200
Angela Winegar: the question becomes, what is your loop between your underlying data that’s feeding the model, and then the evaluation of, you know, is it actually… is your model working? Like, can you… can you genuinely answer that question?905
02:24:53.200 –> 02:25:04.480
Angela Winegar: And when, especially, you know, when your reps realize the model isn’t working, is that data getting fed back to improve the model and show where it’s breaking down? So I would actually argue, you know, I think906
02:25:04.480 –> 02:25:28.390
Angela Winegar: we all know benchmarks don’t work, custom evals really have to be the standard, but then do you have the custom annotation, you know, the ways to point out where you can improve the model, and do you have that as a virtuous cycle and a loop connecting your data, the evaluation, the annotation, and then feeding back into the underlying data? So, my push would be, you know, any KPI really needs to be based907
02:25:28.390 –> 02:25:39.489
Angela Winegar: at the workflow level, but across an organization, I think one way to look at it is, do you have models in place, time to value, kind of all those baseline, expectations, but then908
02:25:39.490 –> 02:25:49.909
Angela Winegar: are your models actually improving over time? Is that time… is that, you know, time savings? Is it getting shorter and shorter every time? Are you layering more and more of these different workflows in place?909
02:25:50.200 –> 02:26:11.570
Mark Niemiec: Yeah, for sure. So we’re… not everybody’s at the same place on the call, and so maybe there’s some folks who are earlier in this journey than the rest of us, and I talk to a lot of CROs, and they say, oh, no, no, we’re using GenAI, and, you know, all my reps have ChatGPT, and it’s working great. So maybe we want to ask the group a question about, you know, proprietary models versus some of the, you know, more, you know, more sort of910
02:26:11.570 –> 02:26:20.560
Mark Niemiec: horizontal or generally available models out there. You know, how are you using them both in your environments? What do you lean towards in terms of911
02:26:20.560 –> 02:26:36.110
Mark Niemiec: proprietary model that’s got guardrails on it, or giving your reps, giving your reps, you know, access to some, things like ChatGPT. Maybe a little pro-con of what the difference is in your, in your organizations. I don’t know, Santi, you came off mute, what do you… what are you, what are you seeing?912
02:26:36.780 –> 02:26:49.020
Santi @ Momentum: I’m a… I have a strong opinion here. I think all the foundation model companies out there, they’re spending hundreds of millions of dollars training very sophisticated models. It takes them a year to put them together.913
02:26:49.430 –> 02:27:07.560
Santi @ Momentum: I just think it’s very hard to consider that a company like ours, or even, you know, a public company that is not dedicating billions of dollars into this journey, could accomplish anything close to something like ChatGPT or Clotson. So I’m a big fan of saying, use vanilla models, and the real edge914
02:27:07.560 –> 02:27:15.529
Santi @ Momentum: Is on the data you feed the model, the context you provide, and the engineering of the prompt, and the sophistication of the task it accomplishes.915
02:27:15.530 –> 02:27:28.829
Santi @ Momentum: So, my team is, you know, we switched, we could… currently we’re using Gemini, but we’re kind of switching between Gemini, ChatGPT, and Claude for most of the work we do. And then, of course, specialized tools, and the way they differentiate is916
02:27:28.900 –> 02:27:45.330
Santi @ Momentum: Are they integrated into our stack? Do they execute complex tasks that my team needs that perhaps would be cumbersome to do by hand on top of ChatGPT? That’s, of course, what we ultimately sell from the momentum standpoint to sales teams, but it applies to the whole organization.917
02:27:46.010 –> 02:27:47.420
Mark Niemiec: For sure. Hannah, how about yourself?918
02:27:47.770 –> 02:28:12.410
Hannah Willson: Yeah, I mean, I guess I would just kind of build off of what James and Angela were saying in terms of the data that you’re putting into it, and, you know, without going into our product, I mean, one of the things I do think about a lot is, like, what is the unique IP that we have as a company that will differentiate us? So, for us, it’s like all those conversations that our AEs are having, you know, we do a lot of work training our AEs, training our SDRs on what questions to ask, what are, you know, how do you elicit a good response.919
02:28:12.410 –> 02:28:15.629
Hannah Willson: What is that data that you’re capturing in all those conversations?920
02:28:15.630 –> 02:28:24.930
Hannah Willson: And so for us, like, that data is really important, that that feeds back into all of the models that we have when we go back out to clients, and when we understand a little bit more about their business, so…921
02:28:24.930 –> 02:28:38.059
Hannah Willson: I think it’s not as much for us, like, what model are we using, as opposed to, like, what data are we putting into it? And specifically, what data are we capturing with the human side of the job that, you know, is unique to us, and we think is our unique IP.922
02:28:38.990 –> 02:28:51.000
Mark Niemiec: For sure. Aaron, how about for your organization, in terms of the general models, or something more specific, and how you harvest the information on the back end to be able to learn more quickly inside of your business?923
02:28:51.000 –> 02:29:13.100
Aaron McReynolds: Yeah, I mean, this is exactly kind of where we sit, you know, we always say ChatGPT was built for 8 billion people and for taco recipes, right? It’s not about the models themselves, it’s what we’re feeding it, it’s the context behind that, and so that’s kind of what we focus on, is building that orchestration layer of the MCPs to be able to get the right context at the right time from the right applications.924
02:29:13.100 –> 02:29:17.569
Aaron McReynolds: And then the results are gonna be better. And so, I don’t think there’s much more I can say about what Santi said as well.925
02:29:18.300 –> 02:29:26.040
Mark Niemiec: For sure. Switching topics a little bit, we hear a lot about, you know, either fully autonomous or human in the loop.926
02:29:26.040 –> 02:29:46.830
Mark Niemiec: I’m curious what folks are thinking about there. I know maybe a couple years ago in the start of this was all about augmenting human capabilities, and as things have changed and we’ve looked for more efficiency, we’re saying, hey, like Aaron, you said, you know, how are we replacing the, you know, the people with these tokens? So, what’s the group’s view on either fully autonomous or human in the loop in terms of927
02:29:46.830 –> 02:29:48.760
Mark Niemiec: How you deploy these capabilities.928
02:29:50.540 –> 02:29:53.760
James Roth: I mean, I’ll jump in with, I think, one of the funniest929
02:29:54.230 –> 02:29:58.550
James Roth: Probably unknown paradigm shifts that we’re seeing today.930
02:29:58.700 –> 02:30:17.270
James Roth: If you look back 12, 18 months ago, it was AI’s gonna replace salespeople, and that was, you know, a lot of noise around that, and I’m sure what many feel, if not all, is the most obvious thing that AI has done is really impact inbound.931
02:30:17.570 –> 02:30:27.760
James Roth: from, you know, AIO, AI search, people just using the summary in Gemini, or people just going to GPT, and none of those… none of those have yet monetized932
02:30:27.890 –> 02:30:41.060
James Roth: search engine optimization of old, and so there’s no way to pay your way up top, and so you saw Monday announce, a couple weeks ago, heavy PLG, heavy inbound company, HubSpot, I think, said they were down 70% on inbound.933
02:30:41.100 –> 02:30:57.950
James Roth: And so a lot of firms are now moving heavier outbound. And so they’re actually hiring more salespeople to go outbound, because I think everybody is in the midst, depending on where they are in the journey, of like, hey, our inbound volume is down 25, 30, 40%, and Angela, I’m sure934
02:30:57.950 –> 02:31:01.220
James Roth: you know, you’re in a lot of CMO roundtables where this is a hot topic.935
02:31:01.230 –> 02:31:08.579
James Roth: And everybody’s now going, hey, we gotta fix Reddit, we gotta, you know, YouTube, what are people saying about us, what ChatGPT is searching, and…936
02:31:08.630 –> 02:31:27.270
James Roth: You know, I think in that shift, we’re seeing, again, more humans being hired in outbound, and so to the fully autonomous versus human in the loop, I think this concept of intelligent outbound is bringing great salespeople back into the mix, and I think what at least we’re seeing from a hiring perspective937
02:31:27.270 –> 02:31:36.489
James Roth: You used to have to make this trade-off of, hey, this person might not be as charismatic, might not have as high EQ, IQ, etc, but they’re willing to make a thousand phone calls in a given day.938
02:31:36.660 –> 02:31:42.670
James Roth: And I think now, with Gen AI, you can hire this different class, if you will, of sales folks.939
02:31:42.670 –> 02:32:01.289
James Roth: Because so much of that drudgery that used to make sales a horrible place to be can be automated, and it can be a bit easier, and so you’re really looking for that quote-unquote human in the loop from an outbound, where, yeah, if you can get the right place, the right time, the right message, the right contact, all of those informational points.940
02:32:01.360 –> 02:32:12.589
James Roth: And just, you know, your job is to be great when you finally connect with Mark. You know, I think that’s where the human in the loop, the not being replaced by AI piece, especially in sales, we’re seeing that941
02:32:12.620 –> 02:32:21.520
James Roth: like, by a long shot. I think in certain areas, I forget who mentioned it, but there are low-hanging fruit areas where it is just primed for fully autonomous.942
02:32:21.730 –> 02:32:34.349
James Roth: password reset and support, you know, those areas, I think, has probably been the biggest and most obvious use case in terms of support team shrinking for the very low-end Tier 1 support.943
02:32:34.350 –> 02:32:42.130
James Roth: And I think then repurposing the great Tier 1 folks into Tier 2, Tier 3, as these products, inevitably, everybody’s bringing AI into their products.944
02:32:42.130 –> 02:32:56.080
James Roth: They get more complex, there’s a pathway for those folks to move up into bigger and better roles while you fully automate the Tier 1. So I think it really is use case dependent in terms of where can you fully automate and where do you still need a human in the loop.945
02:32:56.810 –> 02:32:58.710
Mark Niemiec: Awesome. Hannah, would love to hear your view on the…946
02:32:58.710 –> 02:33:11.960
Hannah Willson: Yeah, I was just gonna say, I mean, I totally agree with that, and I think there is something fundamentally different about the sales use case. Like, if you think about customer support, you know, you have a question, you want that question answered, maybe, you know, maybe somewhat similar in, like, the legal space.947
02:33:11.960 –> 02:33:31.399
Hannah Willson: Whereas sales, I mean, you typically have a problem to be solved, but you’re not like, oh, this is the solution I want. You know, you’re looking at a variety of solutions, there’s a lot of nuance to which solutions are going to fit which parts of your problem. So I think that the use case for AI and sales just is different, so I think humans are going to stay in the loop much, much longer than some other functions.948
02:33:32.230 –> 02:33:32.850
Santi @ Momentum: No, normally.949
02:33:32.850 –> 02:33:33.610
Angela Winegar: I’ll say.950
02:33:33.800 –> 02:33:35.319
Angela Winegar: Go ahead, Santi.951
02:33:35.830 –> 02:33:54.190
Santi @ Momentum: I’ll riff on… I’ll finish that thought, because I’m… I got opinions there, too. I think not only are they going to stay in the loop for longer, I think humans are becoming more critical in that loop. Buying is a trust act, right? When I commit to a vendor and I sign a contract for one or two years.952
02:33:54.280 –> 02:34:12.059
Santi @ Momentum: what I’m doing is I’m deciding if I trust them to, you know, be the ones delivering on the other end. And in the world in which content has been completely democratized, and the world is overflowing with blog posts, and articles, and all the things that in the past you would use to derive trust in a brand.953
02:34:12.260 –> 02:34:22.970
Santi @ Momentum: it becomes then a human relationship. So I do think, and I agree that there is a function to, like, how important the job is if you do something that is, you know.954
02:34:23.030 –> 02:34:35.019
Santi @ Momentum: not as important, then you can fully replace with AI and run fully automatic, and that’s what we push for our own customers when we’re doing things like data extraction from calls. But when you want to buy software and the contract is large enough.955
02:34:35.210 –> 02:34:48.859
Santi @ Momentum: there is no going away from having a person be the one on the other side you talk to, and you know they’re making a commitment to deliver before you sign… before you sign, which is quite counter, what’s the thought that you were, I would say.956
02:34:49.320 –> 02:34:59.959
Mark Niemiec: Yeah. I, Angela, why don’t I go to you, but, you know, we all thought Excel was going to replace accountants, right, in the late 70s, and there’s more of them than ever before, so I think it’s a long game, certainly. Angela, what do you think?957
02:34:59.960 –> 02:35:09.900
Angela Winegar: Totally. I was just gonna add, I completely agree with everything that’s been said, and I’ll add, it’s adding some interesting friction to the sales and marketing relationship from the standpoint of958
02:35:10.240 –> 02:35:28.260
Angela Winegar: I now… my best channels, I am investing a ton in field marketing, in hosted events, alliance marketing, PR, getting people up on stages, especially our senior sellers. And it’s adding some really interesting friction around routing, because it used to be, oh, we have an inbound lead, and it just gets routed to the next person, you know, territory owner.959
02:35:28.260 –> 02:35:33.380
Angela Winegar: And now I have this decision of, hey, I got us a speaking slot at this particular industry event.960
02:35:33.380 –> 02:35:39.120
Angela Winegar: I have two senior sellers who could be a fit, maybe one hasn’t had an opportunity in a while, but maybe one’s better on stage.961
02:35:39.120 –> 02:35:54.109
Angela Winegar: Or, you know, hey, there’s… I’ve gotten a lot of signal that says this account is very, very warm. It’s technically assigned to this person, but this person is in a closer geography, and I know they will get on a plane to go meet with them in person, which makes such a big difference at our ACVs.962
02:35:54.110 –> 02:36:13.180
Angela Winegar: So that has been a really interesting thing for me, and kind of a… just wanted to share, you know, the friction point of, I think we are going to have to rethink about routing and how we handle, especially as we move to this world where, you know, exactly to Sandy’s point, at these large deal sizes, nothing replaces in-person connection.963
02:36:13.180 –> 02:36:35.940
Angela Winegar: And so, you know, our… to James’s point on EQ really deeply matters, that… that element, I think, is constantly something that our marketing team is talking about, is like, hey, we… we think we got this PR slot, who… it could… kind of fits better for this person, given their background, but, like, we think this person would actually be more charismatic on stage. How do we… how do we think about those decisions, and how do we work across964
02:36:35.940 –> 02:36:39.570
Angela Winegar: A go-to-market organization to drive the best outcomes for the organization.965
02:36:40.240 –> 02:36:59.210
Mark Niemiec: Yeah, for sure. We’re seeing a lot of different people take the stage in our organization and others, and seeing us amplify the voices of our sellers and our RevOps people in ways that we didn’t plan to before, so I think you’re definitely onto something there. There’s different makeup of what the go-to-market team looks like, especially as we go forward in this space.966
02:36:59.660 –> 02:37:01.779
Mark Niemiec: One of the things we… Oh, go ahead.967
02:37:01.780 –> 02:37:14.380
Aaron McReynolds: I was just gonna add one thing to James’ point. I think, yes, we’re gonna see an increase in headcount for sales reps that closes. Over a year ago, I talked about this concept of sales becoming like the NFL.968
02:37:14.380 –> 02:37:27.390
Aaron McReynolds: Right? Everyone plays football in high school, everyone can catch a ball, everyone can run. You get to college, it’s just better and better, and only the select few make it. And, you know, people were selling in 2018, 2019, 2020,969
02:37:27.390 –> 02:37:49.490
Aaron McReynolds: Budgets were open, products were selling themselves, and now what we’re seeing is you have to fight for every dollar, and it’s the best of the best, and it’s those that do the extra sleep, that run the extra laps, that go that further mile, that adopt AI to make themselves more efficient. Not only will we have more, but I think we’ll start to see only the best of the best kind of stay in these, you know, high-end enterprise sales roles.970
02:37:49.490 –> 02:38:01.749
Aaron McReynolds: And people like James gets me more selective, right? Interview questions have changed. How do you play AI? How are you more efficient? How are you better? How do you catch better than the person next to you who can also catch? So, just want to add that real quick.971
02:38:02.160 –> 02:38:08.159
Mark Niemiec: Yeah, I think that’s certainly, certainly happening, and, you know, one of the statistics we talk a lot about is.972
02:38:08.160 –> 02:38:30.570
Mark Niemiec: you know, customers are 80% or 85% of the way through the buying process, once they want to engage a salesperson. And when I first heard that statistic, I thought, well, I used to have 100% of the sales process to influence my buyer. Now I’ve got 15%. So to your point around the NFL, I gotta make the 15% really count, because I don’t have all the other opportunities to engage that customer. I gotta nail it on the first conversation or the second conversation, because973
02:38:30.570 –> 02:38:35.980
Mark Niemiec: my, my opportunity for, for failure is… is… I get, I get so much fewer of those opportunities.974
02:38:37.760 –> 02:39:02.529
Mark Niemiec: All right, one of the things we did at SalesLoft, I think, is super interesting that we came up with through an experimentation, is we built what’s called the Office of Pipeline, and it’s really an intermingling of marketing and our sales organization and sales programs. And so what we did is we took some of the SDR capabilities, the ability to analyze data from our RevOps team, the ability to create programs and marketing activities and events from marketing.975
02:39:02.530 –> 02:39:26.180
Mark Niemiec: and put that together in the Office of Pipeline, and to deliver that stuff to our sellers into their environment, so they can go do that outbound much easier with much less friction. So that’s been a really interesting hybrid role that’s been created, where it’s part data scientists, it’s part, you know, former maybe BDR, and it’s part marketer, all coming together into a hybrid role that has proven to be really, really interesting for folks in their careers.976
02:39:26.180 –> 02:39:38.130
Mark Niemiec: but also interesting for our sellers to engage with those folks, because they get a real partner in building pipe, and they’re doing that stuff on their own. So, it’s become an interesting experiment internally for our folks. -
977
02:39:40.160 –> 02:39:59.009
Mark Niemiec: All right, so maybe to round it out, as we think about maybe one piece of advice that you all would give the folks on the call in the session about how to best leverage AI in your go-to-market environment, we’d love to get your best piece of advice each, and that’ll take us home on this one. James, maybe we’ll start with you.978
02:39:59.380 –> 02:40:03.270
James Roth: Yeah, I’ve got a real-time one that some people might get a chuckle out of.979
02:40:03.490 –> 02:40:18.070
James Roth: You know, in that period of time, call it 9 months ago, when we were pushing hard to drive AI adoption, and, you know, we had our own product, and that was great, but really driving more innovative, you know, agent building internally.980
02:40:18.560 –> 02:40:36.539
James Roth: I think we skewed heavily on the incentive for the agent building, and at one point, our account management teams call it 450, and there were close to 5,000 agents that account managers had created within the organization, and so we probably had…981
02:40:36.650 –> 02:40:56.379
James Roth: A couple hundred that were account planning based, a couple hundred that were point of view based, and as they were building these, you know, at first, we were highlighting these great agents being built, and then we realized, like, I don’t really care if these 450 are great agent builders. I want them to be using the best agents that the best people have built.982
02:40:56.560 –> 02:41:14.650
James Roth: And so we kind of flipped that on its head, and we started a full enablement, you know, we picked the 8 to 10 per segment, per vertical, per function, said these are the best as of today, and the folks that built the best ones, now they’re on this Tiger team, and they are the ones that can go tweak and change and kind of improve or add anew.983
02:41:14.650 –> 02:41:24.409
James Roth: But, you know, shifting that focus, I think so many people to the earlier conversations have been just like, use AI, use AI, get AI, use AI, be great at being an AI person.984
02:41:24.490 –> 02:41:37.900
James Roth: When you think about most of the org, I want our account managers to be leveraging AI to drive great outcomes of things they weren’t doing before. Become a vertical expert in a week versus having to be in that vertical for a decade. Like.985
02:41:38.000 –> 02:41:47.240
James Roth: I think incentivizing the outcomes, incentivizing the use of great agents versus our entire company needs to be great agent builders.986
02:41:47.240 –> 02:42:02.969
James Roth: That was a real-time miss that we had. It was great to light the pilot, to get everybody in there thinking about it and playing with it, but then we realized that what we really want is these folks to be using it and experts in using it, versus building it. So, that’s one piece of advice, real-time, that we learned the hard.987
02:42:02.970 –> 02:42:03.609
Mark Niemiec: I love it.988
02:42:03.910 –> 02:42:05.950
Mark Niemiec: Love it. Awesome. Santi, how about yourself?989
02:42:06.620 –> 02:42:24.049
Santi @ Momentum: I do have two. One is very short, though, and seconding on a lot of thoughts everybody shared, to make a really strong point, you gotta get your CRM data under control, full stop. It’s unacceptable. Before, it was all about pipeline reviews. Now it’s about building workflows and actually getting results, so that’s number one, real quick.990
02:42:24.050 –> 02:42:34.890
Santi @ Momentum: I have a fun one. Run an AI hackathon. We just came back from an off-site. I got this shirt there, it was at the beach, so everybody got a linen shirt. We’re a small company, so 100 people.991
02:42:34.890 –> 02:42:50.040
Santi @ Momentum: But we ran a company-wide hackathon. We sat SDRs next to developers, next to RevOps, next to marketing, and we had a lot of really fun exercises, from designing the mascot for our company, to putting together a 2-minute ad video.992
02:42:50.080 –> 02:43:00.450
Santi @ Momentum: And just seeing the level of learning between all these different types of people in my company. Some of them are very AI-forward, others had never used ChatGPT before.993
02:43:00.740 –> 02:43:14.649
Santi @ Momentum: come together, use 11 Labs to put together a podcast ad, or using Google VO3 to generate a video. I think it’s a really… again, it’s not a repeatable motion, but I think it can get you from 0 to 50 at a pretty high scale.994
02:43:14.650 –> 02:43:21.519
Santi @ Momentum: of the organization. It’s worth doing every time if you can find a day or two for your company to take a break.995
02:43:22.080 –> 02:43:31.060
Mark Niemiec: Yeah, we love doing things like that. It just sort of breaks through some of that tension, service tension that may exist. You don’t do it all the time, but it gets people inspired a little bit. Hannah, how about yourself?996
02:43:31.060 –> 02:43:45.250
Hannah Willson: Yeah, I think one thing I’ve learned is, you know, in the early… I’ve been at my current company for about 6 months, it was just like, try everything. Like, let’s have people develop things, let’s… there’s a million things we could do, and all of these things are cool and fun, and they’re fun to build, and they, you know, they get some impact.997
02:43:45.250 –> 02:44:09.130
Hannah Willson: I think, you know, where we’ve come is, like, and our CEO talks to me about this a lot, is like, okay, what are the things that are actually… like, what are the actual problems? Like, what are the things we need to fix? Like, how do we prioritize it, and how are we either building or buying things around that? So just, like, being really clear about what are those biggest areas of impact for the organization, whether it’s efficiency, or pipeline building, or whatever it is, and putting more of your resources there, because998
02:44:09.140 –> 02:44:17.790
Hannah Willson: I think we spread ourselves a little too thin with some of the things that we were doing in the past, and yeah, that focus is important, and centering on the actual business results.999
02:44:18.080 –> 02:44:24.599
Mark Niemiec: Yeah, yeah, staying… keeping people focused, I think we can get a little experimental sometimes, and go off the butterfly path.1000
02:44:24.600 –> 02:44:39.780
Hannah Willson: Sometimes it’s worth it, because you come up with really cool things. You’re like, wow, that’s actually… that took half a day, and now it’s saving everyone hours and hours. So I definitely don’t want to limit the experimentation, but just being really thoughtful about, you know, why are we doing this experiment, and what do we think it’s going to drive?1001
02:44:40.150 –> 02:44:42.299
Mark Niemiec: Love that. Aaron, how about yourself?1002
02:44:42.300 –> 02:44:57.010
Aaron McReynolds: Yeah, Mark, forgive me for using a British reference here, but, you know, you go to London, you’re on the tube, every single time you get on the train, they say, mind the gap. And I think that’s where AI comes in. It’s like, it’s doing those gaps, the things we don’t think about as sellers.1003
02:44:57.010 –> 02:45:04.440
Aaron McReynolds: the little bits, updating the CRM, sending the follow-up email, etc. I think that is where the impact comes in. And if you still have gaps.1004
02:45:04.440 –> 02:45:09.469
Aaron McReynolds: You’re not applying AI well enough right now, so… little takeaway. Mind the gap.1005
02:45:09.890 –> 02:45:24.669
Mark Niemiec: I love that. Yeah, it makes all of us a little bit better by checking all of the things we might not maybe consider, maybe some of that great research, or what are the things in my process that I haven’t really fully baked out. So, mind those gaps, I love that. Angela, how about yourself?1006
02:45:26.030 –> 02:45:44.670
Angela Winegar: Yeah, I’ll just, maybe a slight hot take. I think we’ve all also seen that data on, what Gen AI is doing to, you know, new grads, and I would say I’m actually very bullish on hiring super smart, super scrappy folks out of college who are really native in these tools.1007
02:45:44.670 –> 02:45:54.729
Angela Winegar: and maybe haven’t seen how a normal go-to-market org should function, you know, how does that sales, marketing, RevOps relationship all work?1008
02:45:54.730 –> 02:46:08.480
Angela Winegar: which I think would lead to my second piece of advice, which is, you know, I would really consider being really flexible on org charts, titles, roles, and responsibilities for the next few years, because I think what we’ve been seeing is,1009
02:46:09.240 –> 02:46:18.200
Angela Winegar: our go-to-market, our traditional titles and swim lanes have been very designed for the past era. They have not been designed for the Gen AI era.1010
02:46:18.200 –> 02:46:43.130
Angela Winegar: And as a marketing leader, I’m really struggling to figure out, should I only be hiring senior folks who know how to use these tools? Oh my gosh, actually, maybe a senior person who really understands how product marketing should work, with a junior person who really understands gen AI tools and pairing them together, and that’s actually a really good pod. Like, there’s all sorts of experimentation that we’re doing, so maybe this is going, anti-Hannah’s point on, like, be careful of the experimentation. I would say we’re really experimenting on future of org charts.1011
02:46:43.260 –> 02:47:07.110
Angela Winegar: But to Hannah’s point on always being intentional about why are we doing this, what is the root problem we’re trying to solve, I think I would say really don’t get too stuck in, you know, structures, titles, things that have come from a different era, because I think the next few years are probably going to break a lot of org structures and create a lot of new roles that we, you know… I think we’re starting to understand what they look like.1012
02:47:07.130 –> 02:47:22.619
Angela Winegar: And why there’s value in having an engineer sit in sales and marketing and, you know, a true engineer and building out these tools. But that’s something we’ve been playing with and haven’t quite figured out, but it’s been a very enlightening experience going through that journey.1013
02:47:23.010 –> 02:47:47.749
Mark Niemiec: Yeah, this sounded like a little bit of a plug for the humanities degrees that folks might be in college getting these days. Like, dispense with some of those deep, maybe technical, or rigid sort of ideas of what the world is, because it’s all changing in front of us. So, I loved hearing so many of the things here around, be intentional, be focused, have the right outcome-based KPIs, make sure that we’re maybe not living in the past in terms of org charts and structures, and the1014
02:47:47.750 –> 02:47:50.710
Mark Niemiec: Office of Pipe example for us is one of those.1015
02:47:50.710 –> 02:48:09.960
Mark Niemiec: Where it’s not a role that we would have necessarily hired for previously. And when we look for who the folks are that were going to be successful, it was definitely a cross-section of skills and experiences that may not have fit in a traditional sales role. So, I think we’re all seeing this stuff real-time. Julia, I think you’re going to take us some, get us some questions from the audience, if there’s any out there.1016
02:48:10.190 –> 02:48:21.329
Julia Nimchinski: Yes, what an amazing session. Thank you so much, Mark. We have some questions here from Hans, and he’s asking about agentic sales training and sales coaching. What are your thoughts? What are you doing at Sales Love?1017
02:48:21.730 –> 02:48:39.739
Mark Niemiec: Yeah, so we use quite a bit of it. It’s interesting. One of the things, I’ll talk at a sort of a high level, one of the things that our, we do a lot of is we do a lot of training, we do a lot of knowledge checking, and one of the things that we had done when I first got here was those managers reviewing and viewing those presentations by the reps.1018
02:48:39.740 –> 02:48:44.539
Mark Niemiec: And what we found was our reps were actually much more comfortable reviewing those into AI tools.1019
02:48:44.540 –> 02:48:56.800
Mark Niemiec: They feel a little bit less judged, a little bit less, you know, maybe a little less criticized in doing those in front of your manager. And so, we’ve been… it’s been requested by our team to give those capabilities1020
02:48:56.800 –> 02:49:15.300
Mark Niemiec: to more of our sellers, and every time we do a knowledge, you know, knowledge transfer, we do a training, the rep’s actually asked to be, judged by AI. And so it gives it maybe a little bit less personal feeling in some senses, but a little bit more psychological safety. So, that’s been one thing we’ve done that our team has really reacted well to, but curious what other folks are seeing.1021
02:49:16.860 –> 02:49:29.969
James Roth: I think somebody mentioned it earlier on email writing, where you can, you know, basically coach an agent to write emails like you do. We do the same thing with coaching. There’s a me coaching agent, we have a CEO coaching agent, and we basically coach 50 calls.1022
02:49:30.120 –> 02:49:37.719
James Roth: we train the model to understand how we would coach, and then we put that in place, and it’s automated. I would say that’s a fun one.1023
02:49:37.740 –> 02:49:55.319
James Roth: I think more foundational, we also, to the Data Foundation, you know, we use Docket, we use Seismic, we use, you know, from an enablement standpoint, but I think building all of those into the data foundation, pulling in from first party, pulling in from CRM so you understand win-loss, op generation, conversion rates.1024
02:49:55.320 –> 02:50:10.230
James Roth: What good looks like at scale when you can take all the first-party information, and then you can take the outcomes, and you basically have, this is what great looks like, this is what mediocre, this is what bad looks like, but you have all of that sitting on a data foundation that then pumps out within your either internal chat1025
02:50:10.230 –> 02:50:20.970
James Roth: whomever it might be, or into your actual product set. I think it’s probably the best time to enable at scale I’ve ever been a part of. But that’s just an interesting thing we’re doing.1026
02:50:21.520 –> 02:50:22.240
Mark Niemiec: Love that.1027
02:50:22.870 –> 02:50:40.269
Santi @ Momentum: I do want to talk about coaching as a… coaching is an interesting one, and we do it ourselves, but it is a bit of a breakthrough use case, because coaching before the age of AI was just so poorly done. It’s just so labor-intensive. No company out there with a high-performing team had the time to really1028
02:50:40.270 –> 02:51:00.239
Santi @ Momentum: go deep and coach. If you have a rep that is taking 6 hours of calls a day, no manager, no enablement team would have the time to actually give full coverage to that rep. But in the age of AI, you can listen to every call. You can analyze every single line they said. So it’s like, you’re going from an age in which you were only able to watch one game1029
02:51:00.240 –> 02:51:15.050
Santi @ Momentum: for the team, and try to guess if they were, you know, in the running for winning the championship, to being able to watch every game, and all of a sudden, you can give a complete read of how they’re performing. So I do think it’s a use case that everybody should be looking out for, because it’s a high-yield one.1030
02:51:15.420 –> 02:51:35.159
Hannah Willson: Yeah, I totally agree with that. Just to take that analogy one step further, I mean, it’s kind of crazy that we used to just, like, put AEs on the field, like, biggest call, or cold call, or whatever it is, you know, without any practice, really. Maybe some back and forth, role-playing with the manager, and now you can just do this, you know, 10 times a day with bots. It’s pretty cool, and we’re definitely leveraging that as well.1031
02:51:35.810 –> 02:51:36.580
Mark Niemiec: For sure.1032
02:51:39.080 –> 02:51:54.680
Julia Nimchinski: Amazing. I’m curious to hear your thoughts, Mark, and everyone, really. We just had Mark Roberge, and he’s releasing a new book, The Science of Scaling. It’s out now, I believe you know this, Santi, since he’s an investor, but yeah,1033
02:51:55.020 –> 02:52:09.850
Julia Nimchinski: On the topic of the science of agentic scaling, and generally agentic sales, Mark, how do you see the future configuration of a sales team? What’s the ratio between, you know, humans, I don’t know, agents, and tech?1034
02:52:10.370 –> 02:52:34.520
Mark Niemiec: Yeah, I think there’s been, as we talked on this call, that, you know, sales was, you know, sort of fully human-led, and there was these ideas that, hey, it’s going to be, you know, fully agentic, and it’s going to replace the humans. Those are usually never the answers. They’re usually somewhere in the middle. And I think we’re finding that. We’re finding that the answer is becoming… these are capabilities that will replace some of the mundane, maybe less fun tasks that are part of this part of the business.1035
02:52:34.600 –> 02:52:58.149
Mark Niemiec: But they’ll make the parts of the business that are uniquely human better. The questions will get deeper and more thoughtful, the preparation will get better, the outputs that customers expect will become… the bar becomes higher, and so folks will expect more value in a smaller, shorter engagement. So, you know, I think the future of this business will look like you’ll have really, really high-powerful1036
02:52:58.150 –> 02:53:22.459
Mark Niemiec: highly powered, you know, account executives working with your most important customers. I think you’re going to be seeing a lot of inbound be groomed through, you know, through AI agents and things like that, and you’re going to be seeing a lot more, maybe support roles that are helping to the reps to understand, the deeper data analysis, the deeper insights that we’re, we’re garnering for our customers.1037
02:53:22.460 –> 02:53:31.530
Mark Niemiec: So I think it’s going to become a hybrid where those high-end sales folks are going to become even more and more important. Maybe to Aaron’s point, it’s going to feel like the NFL at every turn.1038
02:53:34.480 –> 02:53:35.390
Julia Nimchinski: Anyone else?1039
02:53:36.450 –> 02:53:49.659
Aaron McReynolds: Yeah, I think I’ll maybe just chime in and say, you know, I think we had these expectations when AI SDRs hit the market that you weren’t going to have to hire SDRs. I don’t know. Anyone put your hand up if that’s your reality? I don’t think it is.1040
02:53:49.660 –> 02:54:07.409
Aaron McReynolds: But the one thing I’m learning is, you know, never bet against the technology either. And so it’s more of, like, how do we plug that in? How do we empower the humans that we do have to become better? But yeah, I’m optimistically cautious, I think, on where that technology can go. But, you know, we talked about this on the panel.1041
02:54:07.410 –> 02:54:25.579
Aaron McReynolds: at length about, you know, there’s no excuse now to be the best of the best, and everyone can do that. Even to the last question on sales coaching, you don’t have to go knock on your manager’s door or wait until you’re not doing well to get the coaching. You can go and do that on your own. You can use AI to go and train yourself and become better every single day, and so…1042
02:54:25.580 –> 02:54:30.780
Aaron McReynolds: Yeah, I think, you know, it’s interesting to see where the technology will go in the very near future, but…1043
02:54:30.780 –> 02:54:33.969
Aaron McReynolds: Right now, you know, still a bit of a gap, I think.1044
02:54:35.640 –> 02:54:37.909
Julia Nimchinski: We have quite a… Angela, go ahead.1045
02:54:37.910 –> 02:54:50.869
Angela Winegar: I was gonna say, I can speak to this a little bit from a marketing perspective, because I think it’s a little more varied, but, as I mentioned before, we’re investing a lot more in field marketing, in-person events, like, anything that is organic relationship building, I think is huge.1046
02:54:50.870 –> 02:54:59.250
Angela Winegar: And then about 5% of my team right now are engineers, and I think we’ll probably ramp that up significantly as well, just because1047
02:55:00.290 –> 02:55:16.289
Angela Winegar: so much of that sales-marketing relationship is reliant on the handoff process, and a lot of the handoff can actually be automated, but, you know, to the last question on where are we investing in Agentic, I think intent-based marketing is a term that folks are really throwing around, but1048
02:55:16.290 –> 02:55:41.269
Angela Winegar: how do we get the signals to our sellers that they need to prioritize their leads appropriately? And, you know, there’s all sorts of, you know, oh, this person from this company commented on this thing on LinkedIn, you know, and then how do you marry that with, like, oh, and they also went to college with one of our senior sellers, and, you know, how do you marry those types of signals? I think is something that we are investing a lot in, and I think the makeup of our organization is going to change.1049
02:55:41.270 –> 02:55:43.379
Angela Winegar: Change somewhat dramatically as a result.1050
02:55:45.770 –> 02:55:46.470
Julia Nimchinski: James?1051
02:55:46.700 –> 02:55:50.260
James Roth: Yeah, I mean, I think we touched on it briefly, but…1052
02:55:50.430 –> 02:56:13.509
James Roth: you know, I started my sales career in 2010, and I think most AEs, most salespeople were full cycle, where you had to go find it, you prospected it, you ran the deal, you kind of grew up, you worked in the very low end of SMB. I think we’re gonna see a shift back to that. I see a lot more companies going towards at least some form of full cycle AE.1053
02:56:13.630 –> 02:56:31.260
James Roth: you know, you go back, I think Aaron brought up 18 through 22. There was so much inbound, there was so much budget, ZERP, all the good things that we fondly look back on. You could have this machine that basically no one was actually prospecting. You had SDRs, they were primarily catching inbound.1054
02:56:31.260 –> 02:56:43.780
James Roth: So I think prospecting comes back in a big way, and I think we’re seeing that, and I think, you know, to the overarching… now, that might last 9 months to 12 months. I think whoever can call 5 years, please let me know, and…1055
02:56:43.780 –> 02:57:00.480
James Roth: I would love to invest alongside of you, but I think ultimately, for the next year, that’s probably what it’s going to look like, the ability to go reinvest in prospecting alongside a full-cycle AE. I also think, just one take on it, you look at most sales teams that have an SC,1056
02:57:00.480 –> 02:57:05.649
James Roth: you know, we’ve kind of said, hey, you don’t need to know the product that well. An SC’s gonna do most of the heavy lifting.1057
02:57:06.380 –> 02:57:26.040
James Roth: And when you think about what the rep needs to be accountable for in addition to that, it’s almost impossible. You need to know the vertical, the sub-vertical, you need to know the segment, you need to have read their 10K, and this is one of 50 accounts that you have, and you need to be an expert across all of these different things. So companies will go out and they’ll verticalize, and they’ll say, hey, here’s my retail seller.1058
02:57:26.040 –> 02:57:45.510
James Roth: There are little things that we’ve done, but I think the biggest change, I think, in Agentic AI sales land is that we’re actually giving salespeople a chance to be good again, because we’ve overwhelmed them with tools. It was like, hey, well, we’re gonna give you an SC, and then we’re gonna give you 20 different tools that are gonna help you do all of these 20 different things.1059
02:57:45.520 –> 02:57:53.310
James Roth: And when you look at the makeup of great salespeople, not always are they gonna be the sitting in front of 3 monitors and an expert across 20 different tools.1060
02:57:53.310 –> 02:58:17.800
James Roth: And so I think, irrespective of whatever the AI tools are, the ability to input all of that knowledge and all of that context and all of that personalization so that they can do what they do best, which is get out to the field marketing events that Angela’s throwing, which is go build relationships with customers, and an SC can help you on the product side, sure, but I think it’s so much harder to the 85% that Mark brought up, the fact that it is hard1061
02:58:17.800 –> 02:58:26.820
James Roth: harder to sell anything than ever now. You have to have a point of view. You have to speak their industry, their language, their sub-segment vertical. You have to know their acronyms.1062
02:58:26.820 –> 02:58:41.339
James Roth: It is almost impossible to be a great salesperson today, just given all of those shifts, and I think AI is the thing that should help them to be able to show up with the industry landscape, with the point of view, know their 10K, know that Mark was on this podcast.1063
02:58:41.340 –> 02:58:53.499
James Roth: whatever it may be, so that I can just focus on what I do best, which is connect with Mark, build trust with Mark, present us as a trusted partner, etc. And so I think how that manifests across different go-to-markets.1064
02:58:53.500 –> 02:59:08.749
James Roth: I don’t know, but I… I do know there’s gonna be a lot more outbound with AIO, because I feel like that’s just gonna get bigger and bigger and bigger, and whether they monetize search engine optimization or not, who knows? So it’s gonna continue to be a black box for at least the next 12 to 24 months.1065
02:59:09.760 –> 02:59:10.960
Julia Nimchinski: Phenomenal, phenomenal.1066
02:59:10.960 –> 02:59:31.949
Mark Niemiec: Yeah, James, we’re seeing the same things back to, back to prospecting, back to full cycle, and I think of what you just described as putting everybody in an Iron Man suit. It’s gonna have to make everybody as powerful and capable as they can be. So thank you to all, everybody who joined us, Aaron, Angela, James, Santi, Hannah, appreciate you joining us.1067
02:59:31.950 –> 02:59:35.129
Mark Niemiec: Julia, I think you’re going to take it over to a demo after this, right?1068
02:59:35.530 –> 02:59:43.539
Julia Nimchinski: Definitely. Mark, lastly, where should our community go, and what’s on the sales love roadmap, as far as what you can share?1069
02:59:43.880 –> 03:00:03.300
Mark Niemiec: Awesome. Yeah, well, we’re super excited about the combination we announced about a month ago. We are coming together with Clary over the next couple weeks, and we’re in the final process of that. We really look forward to sharing the full vision of that revenue orchestration lifecycle management platform that we’ll be bringing to the market, and it’s the best of both.1070
03:00:03.300 –> 03:00:23.299
Mark Niemiec: both worlds. There is another group of agents coming out. We announced a group back in early May that have been really well received by our customers, doing a lot of the automation work that a lot of the folks on the call talked about, and another group coming out as well in about two weeks that we’re excited for the market to react to. And so that’s where we’re going, and we’re really, really looking forward to it.1071
03:00:23.690 –> 03:00:36.509
Julia Nimchinski: Amazing. Excited to see this in action, and thank you, everyone, once again. Now we have Shane McLaughlin, Principal Sales Engineer, SalesLoft, and yeah, let’s dive into the demo. Welcome, Shane. How are you doing?
Register now
This event focuses on the following skills:
Prompt Engineering
Demand Generation
Analytics
Conversational Intelligence
Prospecting
Storytelling
AI Agents Orchestration
Prompt Engineering
Demand Generation
Analytics
Conversational Intelligence
Prospecting
Storytelling
AI Agents Orchestration
Prompt Engineering
Demand Generation
Analytics
Conversational Intelligence
Prospecting
Storytelling
AI Agents Orchestration
Prompt Engineering
Demand Generation
Analytics
Conversational Intelligence
Prospecting
Storytelling
AI Agents Orchestration
Demand Generation
Analytics
Conversational Intelligence
Prospecting
Storytelling
AI Agents Orchestration
Prompt Engineering
Demand Generation
Analytics
Conversational Intelligence
Prospecting
Storytelling
AI Agents Orchestration
Prompt Engineering
Demand Generation
Analytics
Conversational Intelligence
Prospecting
Storytelling
AI Agents Orchestration
Prompt Engineering
Demand Generation
Analytics
Conversational Intelligence
Prospecting
Storytelling
AI Agents Orchestration
Text transcript
CRO Roundtable — The Agent-Driven Revenue Framework: Building Your AI-Powered Sales Vision
AI Summit held on Sept 16–18
Disclaimer: This transcript was created using AI
Watch. Learn. Practice 1:1
Experience personalized coaching with summit speakers on the HSE marketplace.