Fireside Chat with Chris O’Neill & Russell Scherwin — The Agentic Flywheel: Trust, Timing, and the Future of GTM Personalization

1
00:00:06.420 –> 00:00:26.260
Julia Nimchinski: And we are live! Welcome back to day two of the Agentic Distribution Summit, where we are getting a first-hand look at how companies and leading CXOs, VCs, and analysts are transforming from manual GTM into Agentic Distribution. As always, please follow along the conversation in the HOC Slack.

2
00:00:26.490 –> 00:00:31.849
Julia Nimchinski: And we’re super excited to kick things off with the Agentic Flywheel.

3
00:00:32.200 –> 00:00:37.780
Julia Nimchinski: Featuring Chris O’Neill, CEO at Growsloop, and board member at Gap.

4
00:00:38.100 –> 00:00:46.390
Julia Nimchinski: And this conversation is going to be led by Russell Sherwin, facilitator at Force Management and former CMO at IBM Watson.

5
00:00:46.580 –> 00:00:49.570
Julia Nimchinski: What a chat! Welcome to the show! How are you doing?

6
00:00:51.160 –> 00:00:53.269
Chris O’Neill: Doing great. Thank you so much. How are you doing?

7
00:00:54.130 –> 00:01:04.380
Julia Nimchinski: Super excited for this, and yeah, for our community, please ask questions, we’ll leave 5-minute Q&A in the end of this conversation, and yeah, let’s dive in. Russell!

8
00:01:06.110 –> 00:01:22.710
Russell Scherwin: Thanks so much, Julia. Chris, it is such a pleasure to have you here with us, and great to meet you as well. So, let’s just get started. You know, both of us have been through a couple of hype cycles, and we know that sometimes meetings can get distorted. Why don’t you start by explaining to the audience

9
00:01:22.710 –> 00:01:26.810
Russell Scherwin: what an AI agent is, and how they collaborate with us.

10
00:01:27.810 –> 00:01:34.080
Chris O’Neill: Sure, Russell. Yeah, I’ve been through my fair share of hype cycles, from dot-com busts to mortgage meltdowns to now,

11
00:01:34.080 –> 00:01:49.559
Chris O’Neill: what I think is probably the largest seismic shift in AI. I am unapologetic about it being real, probably overhyped in the short term, and underhyped in the long term, but to your question, agents are essentially fueled by data of all sorts.

12
00:01:49.560 –> 00:02:02.059
Chris O’Neill: the best type are first-party data for companies because it provides context. Agents are masters at taking context and completing task-specific workflows, or steps in the task-specific workflows.

13
00:02:02.190 –> 00:02:16.229
Chris O’Neill: The agents have autonomy in that they can have agency, that’s hence the name, and then they can work together in what I sometimes refer to as agent swarms to complete those workflows. So at the heart of it, it’s really taking context,

14
00:02:16.510 –> 00:02:25.439
Chris O’Neill: A whole bunch of data, task-specific algorithms, to complete… to complete, either discrete steps or whole workflows together.

15
00:02:27.900 –> 00:02:34.320
Russell Scherwin: Love it. How do you see them… how do you see them working with us, and how have you seen them working with us?

16
00:02:34.570 –> 00:02:47.590
Chris O’Neill: Yeah, the way I think of agents is really alongside. So, there’s a lot of hype, of course, and a lot of headlines written about how agents are going to take over everything, and, you know, there’s some degree of truth, of course, in that. Large portions of

17
00:02:47.590 –> 00:02:59.519
Chris O’Neill: jobs will simply go away, or almost every job will be reinvented in some meaningful way. So, I think of agents much the same as you would, like, having an infinite number of interns.

18
00:02:59.520 –> 00:03:08.569
Chris O’Neill: that you’d hire, and you integrate into a specific workflow, right? So, if we step back and think about what has fueled

19
00:03:08.570 –> 00:03:23.799
Chris O’Neill: step function changes in artificial intelligence, whether that’s going back to, things like AlphaGo, Google’s famous Leap Forward, whether it’s self-driving cars, or whether it’s even Billy Bean using new techniques to reinvent

20
00:03:23.800 –> 00:03:32.160
Chris O’Neill: how baseball and sports management was done. There are really 3 or 4 ingredients. It’s a step function improvement in data.

21
00:03:32.180 –> 00:03:49.390
Chris O’Neill: It’s a step function improvement in task-specific algorithms that says you do this, then this, then this. And the ability to test that, that’s an important part, to say, hey, are the agents actually doing the task-specific things right? And then the last ingredient is courage.

22
00:03:49.390 –> 00:04:00.769
Chris O’Neill: And that’s really why I think it’s so important for leaders to, you know, gain facility and fluency in AI, and then have the courage to say, there’s just simply a better way to do this.

23
00:04:00.770 –> 00:04:18.649
Chris O’Neill: Knowing that it’s not gonna be perfect, the road’s going to be twisty, it’s gonna be bumpy, but that’s really how all major breakthroughs have, have been fueled, with a combination of those three things. So agents have a starring role in all elements of that.

24
00:04:19.480 –> 00:04:44.310
Russell Scherwin: Love that, I love that, because I took two things away from that answer I want to probe a bit deeper into, but I want to make sure I got that right first. First of all, what a great quote, or terrifying quote, for some of my MBA students in my class. Agents are an infinite number of interns. I’ll put that one to the side for now. But the other key one is what… there’s four layers. There’s the data layer, there’s the workflow or application layer, there’s then testing on top of that, and then the courage to make

25
00:04:44.310 –> 00:04:45.319
Russell Scherwin: it all work.

26
00:04:45.320 –> 00:04:56.950
Russell Scherwin: And so when I think about that, that’s not too different than a, let’s say, an operational stack might have looked in the 90s, which is, you got your data tier, you have your application tier, and humans on top of it.

27
00:04:56.950 –> 00:05:09.689
Russell Scherwin: optimize the combination of humans and workflow on top of hopefully good data. We’ll get to that in a second, too. Yeah. And then have the courage to drive outcomes through it. So, let me throw this question at you.

28
00:05:10.770 –> 00:05:22.940
Russell Scherwin: Management could be said it’s about orchestrating good intent. We all have good intent, but orchestrating or orchestrating outcomes is about driving workflows and behaviors

29
00:05:23.150 –> 00:05:25.709
Russell Scherwin: That consists of team members to get there.

30
00:05:26.050 –> 00:05:32.690
Russell Scherwin: Let’s hit on capacity planning, or let’s hit on orchestrating outcomes. First of all.

31
00:05:33.160 –> 00:05:41.629
Russell Scherwin: What… what’s the new ceiling? Let’s start here. What’s the new ceiling in positive business outcomes that can be created

32
00:05:41.760 –> 00:05:45.339
Russell Scherwin: As we infuse agents into our workflows.

33
00:05:46.120 –> 00:06:01.070
Chris O’Neill: Yeah, and that’s the exciting part, Russell. It’s really driving outcomes. And, you know, through the lens of marketing, it’s always been a little elusive, right? It’s like, you know, the famous quote that says, like, you know, half my marketing works, but I just don’t know which half. And so.

34
00:06:01.070 –> 00:06:07.229
Chris O’Neill: What I think we’re chipping away at is the ability to have greater precision in outcomes.

35
00:06:07.270 –> 00:06:25.520
Chris O’Neill: And, you know, I’m a big believer, and why this is so exciting is… is the velocity that it enables. And I’ll unpack the term velocity because it sometimes gets interchanged, often gets interchanged with the notion of speed. But velocity is a term from physics that’s basically speed

36
00:06:25.520 –> 00:06:33.580
Chris O’Neill: times direction, or direction times speed, right? And often companies and individuals even overlook that important point.

37
00:06:33.770 –> 00:06:37.979
Chris O’Neill: Meaning, you have to be clear about where you’re going.

38
00:06:38.060 –> 00:06:55.190
Chris O’Neill: You also have to be equally clear as to what success looks like, and what problems you’re actually solving, right? This is not about just taking a new technology and trying to apply it everywhere, and that’s really one of the things which I think fuels hype cycles in the wrong direction.

39
00:06:55.190 –> 00:07:09.409
Chris O’Neill: Right? It’s a technology in search of problems, and I think this is so real in that if you are clear that your job as a company, among many things, is to basically clearly identify a problem that matters for a group of customers.

40
00:07:09.410 –> 00:07:15.680
Chris O’Neill: And then find a way to solve it. Not just launch a feature, but actually solve that problem.

41
00:07:15.830 –> 00:07:24.500
Chris O’Neill: And the pace with which we’re able to do that now, by working backwards from customer insights. You know, artificial intelligence agents have a fundamental role.

42
00:07:24.500 –> 00:07:40.590
Chris O’Neill: to basically not just scour the web, but to go everywhere, inside your customer… inside your internal data sets, and many, many companies are doing that, as well as we used to do with just Google. They’re gathering information from inside lots of disparate systems.

43
00:07:40.590 –> 00:07:56.750
Chris O’Neill: And there’s the rise of data clouds, the snowflakes, and the data bricks and Google and Azure, etc. Now, that is really fueling this, because it’s about gathering that information, right? Allowing you to get from insight to impact faster, right? Compressing that cycle.

44
00:07:56.750 –> 00:08:04.960
Chris O’Neill: You know, right now, it’s too manual, it’s too slow, and these tools and data sets are fragmented all over the organization.

45
00:08:04.960 –> 00:08:28.890
Chris O’Neill: You know, this is a path to a better way in that getting clear on where you’re going, the direction, and then really doing everything in your power to really work customer back to make sure that you’re reducing the time and the space between that idea, that insight, and then the impact that you can ultimately drive. That’s what the great companies are doing, or at least the ones that are separating the best from the rest, so to speak.

46
00:08:29.080 –> 00:08:33.929
Chris O’Neill: So it’s an exciting time, and I’m happy to drill into some specific examples if you’re interested.

47
00:08:34.320 –> 00:08:40.740
Russell Scherwin: Yeah, I do want to go there. I want to just, again, bring up a really cool quote that I think everyone should be writing down, which is…

48
00:08:40.890 –> 00:08:46.189
Russell Scherwin: you know, it’s speed… it’s velocity, not speed. It’s kind of like I told my kid last night, my 14-year-old.

49
00:08:46.600 –> 00:08:52.859
Russell Scherwin: You gotta work hard. Don’t work hard unless you’re also working smart, right? There needs to be intentionality to it.

50
00:08:52.860 –> 00:09:08.679
Russell Scherwin: Let’s talk about… I mean, we’re talking about driving greater outcomes, and I know Growth Loop, you’ve got some great name brands on your website. Talk to some outcomes. At the end of the day, for marketing, I just want to drive more consumers into my funnel, and I want to convert more of them.

51
00:09:08.910 –> 00:09:18.940
Russell Scherwin: I don’t think Agentic, or the buzzword of the day, is gonna change that, but talk about some of the outcomes you’re seeing this new technology in the mix driving.

52
00:09:19.710 –> 00:09:28.839
Chris O’Neill: Yeah, so if you think about, like, where the company started was really inside of Google, and for a very innovative company, which Google clearly is.

53
00:09:28.890 –> 00:09:41.649
Chris O’Neill: We felt that the way in which marketing was supporting business outcomes was a little bit convoluted. And not because people had… weren’t clear on what success looked like, or even had good or bad intent, like, it wasn’t… it wasn’t that at all.

54
00:09:41.650 –> 00:09:49.450
Chris O’Neill: It was that there was really these broken workflows and siloed information all over the place. So if you were trying to drive

55
00:09:49.450 –> 00:10:02.469
Chris O’Neill: business outcome for a specific business line, Google, right? The marketing person, the product marketing person, had to go over to the data team and line up and queue up their request alongside all the other things going to that data team.

56
00:10:02.470 –> 00:10:19.129
Chris O’Neill: You know, and then it would say, oh, this is like a little Goldilocks. I mean, it’s too big, too small, not really relevant to move the needle. Just to identify the first step, which is to say, who might be amenable to some ideas that we have to grow the business, or grow this category, or this product in this case.

57
00:10:19.130 –> 00:10:30.770
Chris O’Neill: And that literally took days and weeks, and sometimes in some organizations, months. So, it starts with reducing the back and forth by basically saying, hey, let’s put the data in an accessible spot.

58
00:10:30.770 –> 00:10:36.510
Chris O’Neill: In this case, it was Google BigQuery. And then let’s allow marketers, in this case, to self-serve.

59
00:10:36.680 –> 00:10:55.329
Chris O’Neill: Okay, what’s the problem that’s being solved there? And it goes from basically weeks and months in some cases, just to define an audience, who you should be speaking to, and what you… forget about for a minute just what you’re going to offer them, but just to say, hey, how do we sign? So, first and foremost, it’s dramatic efficiency from

60
00:10:55.330 –> 00:10:58.260
Chris O’Neill: You basically, what take… used to take weeks and months.

61
00:10:58.260 –> 00:11:01.410
Chris O’Neill: With an agent trained upon the data.

62
00:11:01.410 –> 00:11:12.020
Chris O’Neill: That happens instantly, while you’re sleeping. You wake up and you’ve got 10 ideas, right, that aren’t just pulled from the air. They are informed from the market, from your competitors.

63
00:11:12.020 –> 00:11:32.610
Chris O’Neill: and from your actual data and actual previous results to say, here are 8 ideas, 10 ideas, whatever it is, that we think you should consider running. Now, the human comes in and says, okay, which of these, knowing what I know about our customers and what we’re trying to accomplish, which ones might be the most amenable to testing? So that’s one, is just saving so much time, right?

64
00:11:32.610 –> 00:11:37.689
Chris O’Neill: Like, from hour… from basically weeks and months to basically hours and minutes and even seconds.

65
00:11:37.740 –> 00:11:38.810
Chris O’Neill: That’s one.

66
00:11:38.930 –> 00:11:40.940
Chris O’Neill: The other notion is.

67
00:11:41.110 –> 00:11:56.200
Chris O’Neill: And almost all of our customers are doing some version of that, right? Whether it’s all the baseball teams that we have, the professional sports teams that we’re so lucky to have, whether that’s Google, whether that’s, more recently, Costco and Albertsons, and, you know, we’re so fortunate to have a lot of these big brand names that have trusted us to be a partner.

68
00:11:56.410 –> 00:12:06.080
Chris O’Neill: But where it gets even more interesting is to say, okay, great. What are we going to offer, and what channels, and when, right? So you start to think about all the different surface areas of interactions with your customers.

69
00:12:06.080 –> 00:12:18.180
Chris O’Neill: And this is when it becomes hard, right? Because you’ve got to basically figure out what are the surface areas, right? Email, paid ads, the social, etc. Like, literally infinite number of channels, it seems like.

70
00:12:18.180 –> 00:12:28.550
Chris O’Neill: And how do you do it with consistency so that you can basically do two things, right? You can start to optimize the lift of what types of offers

71
00:12:28.580 –> 00:12:43.020
Chris O’Neill: work in which channels, and then you feed that back into the cloud to understand what actually worked. And you do that increasingly with personalized messaging, right? That’s been the Holy Grail for a long time.

72
00:12:43.020 –> 00:12:57.620
Chris O’Neill: right message, right channel to the right person, with the right offer. That’s been elusive for all the reasons we started to talk about more. But basically, you orchestrate across all these channels, very complex channels, and you integrate all the intelligence of a company.

73
00:12:57.620 –> 00:13:07.759
Chris O’Neill: You have these data science teams that are working away, doing a fantastic job. They know propensity to purchase something, they know price elasticity, but all these things live in the ether.

74
00:13:07.760 –> 00:13:24.809
Chris O’Neill: So it’s about stitching those in, or weaving those in to an orchestrated journey that’s personalized to that person. And how is it personalized? Because it’s the first-party information that a brand has on their customers. And that’s the goldmine. That’s the goldmine that has not been properly harnessed, or at least fully harnessed.

75
00:13:24.810 –> 00:13:30.720
Chris O’Neill: Until, until now. So that’s, that’s a bit of a tour of the things and the specific things that

76
00:13:30.720 –> 00:13:44.750
Chris O’Neill: are allowed. So it’s fast… faster and less fragmented workflows, and then it’s Lyft by basically being much more intelligent about which channels and which offers are working, so that you can double down on the ones that are, and then you can discard the things that

77
00:13:44.750 –> 00:13:54.260
Chris O’Neill: or failed experiments. So, I’ll stop there. I just had a bunch of things, Russell. I want to allow… catch my breath, and you can challenge any of that, or take us in a different direction.

78
00:13:54.260 –> 00:14:08.739
Russell Scherwin: I tell you what, let’s do this. I love what you said. I’m gonna… you reminded me of a story, gosh, going back almost 6 years right now, that I’ll tell you, because I feel like you’ve given me so much. And then we’ll pivot into how we do this at scale. So that’s kind of where this story’s gonna take us.

79
00:14:08.740 –> 00:14:19.649
Russell Scherwin: Story I’ll tell you is very similar when you… on the topic of personalization. This goes back when I was leading the e-commerce business at IBM. We had this customer, Carhartt, which I’m sure you’ve seen plenty of their stuff.

80
00:14:19.650 –> 00:14:21.399
Chris O’Neill: I know Cohart very well, yes.

81
00:14:21.770 –> 00:14:23.310
Russell Scherwin: Do you remember the movie Interstellar?

82
00:14:24.480 –> 00:14:27.270
Chris O’Neill: I didn’t watch it, but I’ve heard of it, yeah. Tell me more.

83
00:14:28.160 –> 00:14:37.570
Russell Scherwin: Well, it’s a must-watch. I’ll tell you it’s a must-watch, but in the movie, Matthew McConaughey was one of the stars, and for the vast majority of the movie, he was wearing a Carhartt jacket.

84
00:14:37.930 –> 00:14:48.249
Russell Scherwin: And at the time, Carhartt was a legacy commerce customer of ours. We owned their e-commerce platform. And imagine what happens on Carhartt’s site

85
00:14:48.340 –> 00:14:59.220
Russell Scherwin: And their e-commerce sales. You know, when Matthew McConaughey… and by the way, this was a record-grossing opening weekend. What do you think happened to revenues on Carhartt’s site that weekend?

86
00:14:59.530 –> 00:15:02.600
Chris O’Neill: Well, did the sights stay up? That’s my first question.

87
00:15:02.600 –> 00:15:04.470
Russell Scherwin: The site did… yes, the site did stay up.

88
00:15:04.550 –> 00:15:11.250
Chris O’Neill: IBM did a good job. I mean, it would just go… it would go… it would go basically straight out, right? It’d be exponential growth.

89
00:15:11.680 –> 00:15:18.040
Russell Scherwin: So here’s what happened. The site traffic went up, but that’s a vanity metric, as we know. Revenue?

90
00:15:18.150 –> 00:15:24.090
Russell Scherwin: Barely blipped the needle, because that jacket was buried, like, 11 pages deep.

91
00:15:24.220 –> 00:15:42.720
Russell Scherwin: And this was back before we were scaling AI, but we started applying, and so what we did is we applied what would today be called an agent to look at things that are happening. What’s… what kind of anomalies are happening? Why is everyone searching Matthew McConaughey that would have alerted someone over the weekend to go move that from page 11

92
00:15:42.720 –> 00:15:50.079
Russell Scherwin: up to the forefront and create a hero graphic. And so, when I heard your story, it went back to that, but back then, we weren’t doing that at scale.

93
00:15:50.490 –> 00:15:52.740
Russell Scherwin: You know, when I look at your organization.

94
00:15:52.760 –> 00:15:57.460
Chris O’Neill: You’re making those tools available, not to superhumans, but to…

95
00:15:57.460 –> 00:16:00.760
Russell Scherwin: an everyday human like myself. Talk about… Yeah.

96
00:16:00.860 –> 00:16:07.479
Russell Scherwin: now I’m the CMO at Carhart, I’m the CMO at Gat, which I know is one of your portfolio companies, or it’s a company that’s on the board of.

97
00:16:07.720 –> 00:16:21.479
Russell Scherwin: what questions should the board be asking a CMO, and what questions should the CMO today be asking their teams to make sure this technology is, which can be used at scale, is getting harnessed?

98
00:16:22.100 –> 00:16:41.610
Chris O’Neill: Yeah, I mean, that’s a great example, and maybe we can come back to the role that agents play on the other side. So this is more future-looking, but I do believe there’ll be agents on both sides of a transaction, so to speak. So, we can put a pin in that and come back, but to your question with CMOs, like, I think,

99
00:16:42.110 –> 00:17:00.179
Chris O’Neill: the very first question as a board or senior-level person in the company should be asking is very specifically, where, with specific examples, will this show up as a benefit on your profit and loss statement? Not hand-wavy, we’re gonna, you know, experiment our way into it. No, no, it’s going to move this

100
00:17:00.180 –> 00:17:07.540
Chris O’Neill: This metric, which will drive top-line sales because we’ll be able to do X more experiments.

101
00:17:07.540 –> 00:17:23.379
Chris O’Neill: And we’ll be able to do it with Y less humans in the loop, or whatever. Like, there’s different variants of where it shows up. But number one is be very specific about where it’s going to be. Secondly is around AI fluency, right? How and where are we using AI?

102
00:17:23.380 –> 00:17:29.360
Chris O’Neill: And how and where are we investing in our teams at all levels? By the way, all the way up to a board.

103
00:17:29.360 –> 00:17:45.119
Chris O’Neill: all the way down to frontline, frontline employees, how and where are we investing in AI fluency and being very specific about this? Not just like, hey, we’re using AI. No, it has to be, like, similar to how a job ladder is treated.

104
00:17:45.120 –> 00:18:02.419
Chris O’Neill: Right? And this is something we’re experimenting with, and I know Zapier’s here, and I drew great inspiration from Zapier, and my friends at Shopify, who really have, you know, written a memo. They talk about from a memo to a movement, so we write a really popular memo. But really, it boils down to the following.

105
00:18:02.420 –> 00:18:13.049
Chris O’Neill: not just hand-wavy on AI, but being very specific by function and by role. What are the expectations for the basics of using AI all the way up to mastery?

106
00:18:13.050 –> 00:18:27.460
Chris O’Neill: And then not only setting that expectation, but enforcing it. To say, this is a part of doing your job, and if you’re not comfortable embracing AI to do your job, then, like, that’s a problem, and you might not be working here if you can’t get behind it.

107
00:18:27.460 –> 00:18:34.739
Chris O’Neill: And then the other part is really celebrating that stuff, so really surfacing examples, big and small, and especially the small.

108
00:18:34.740 –> 00:18:39.959
Chris O’Neill: Let’s say, hey, here’s how we’re using AI to drive, you know, more meaningful work.

109
00:18:39.960 –> 00:18:55.140
Chris O’Neill: and more meaningful outcomes. So I’d say starts with the outcomes, right? How can you trace it all the way back? How can you make your CMO become BFFs with your CFO to say, look, this is how and where we’re driving quantifiable impact.

110
00:18:55.240 –> 00:19:07.880
Chris O’Neill: On the top and or bottom line, and how are you investing in your teams to drive AI fluency? And the last area I’d say is really around data, right? You don’t have AI without data.

111
00:19:07.880 –> 00:19:13.949
Chris O’Neill: Right? You do not have a great strategy for AI without robust, semantic layer of data.

112
00:19:13.950 –> 00:19:30.730
Chris O’Neill: Investing in the hygiene, investing in governance and protection, like, that’s the part that is not sexy at all, but it’s fundamental to basically driving, because agents without context are useless, right? They thrive off of context. They need to understand

113
00:19:30.730 –> 00:19:33.730
Chris O’Neill: The problem. They need to understand what’s worked in the past.

114
00:19:34.020 –> 00:19:50.679
Chris O’Neill: It’s similar to any other role. You need to provide clarity. What is that agent being asked to do? How are they doing against that? And how are you holding that agent accountable to the tasks that you’re setting them out to do? And none of that’s possible without robust data, clear, fresh.

115
00:19:50.740 –> 00:20:04.239
Chris O’Neill: comprehensive and accurate data. That’s the least sexy part, but in many ways, foundational. So those, I mean, I could go on, but if I had to boil it down to just three, it would be results orientation, AI fluency, and then a real crisp data strategy to support it all.

116
00:20:06.000 –> 00:20:21.810
Russell Scherwin: I’ll play that back, because if I’m thinking about, you know, some of the companies that I ice the boards of, what I’d be taking away right now is, are you talking to your CMOs and all your C-level executives about how the technology is driving outcomes and results? That’s number one.

117
00:20:22.110 –> 00:20:37.909
Russell Scherwin: Number two is you’re talking, how are you enabling your teams, but… and using it? But the key thing you said there, Chris, is with specificity. What specific… how are you specifically infusing it into your operation, and how are you measuring it? Which is another key piece.

118
00:20:37.910 –> 00:20:50.009
Russell Scherwin: You said data, but you also said something that I want to highlight as well, which is, how are you celebrating? I used to, on my sales teams, have the most embarrassing call of the week award, because I wanted to drive cold calls. In other words.

119
00:20:50.070 –> 00:20:55.210
Russell Scherwin: You celebrate to ensure that you’re driving the behaviors, and by celebrating the victories.

120
00:20:55.350 –> 00:21:11.149
Russell Scherwin: you’re infusing it into the culture, which needs to be top of mind. You also said something, Chris, which is a little intriguing, which is holding agents accountable. Well, isn’t that the job of a manager, to hold people accountable? What, you know, bring those two

121
00:21:11.410 –> 00:21:21.090
Russell Scherwin: You know, perhaps opposing or not opposing thoughts together around agents or just accountability in general in an era where you have humans and robots teaming up.

122
00:21:21.560 –> 00:21:39.590
Chris O’Neill: Well, so I’ll try to thread a couple things together here, or connect a few dots, and, you know, it’s the notion that I believe even individual contributors are now managers, right? If they think about it on the following plane, that you’re going to be managing agents, right? And it’s… it’s really…

123
00:21:39.860 –> 00:21:47.060
Chris O’Neill: Powerful to think about what makes a manager or even a leader successful, and how that might apply to an agent.

124
00:21:47.160 –> 00:22:01.489
Chris O’Neill: At the simplest level, I think a manager is really clear about the following, right? They say, hey, what are you trying to accomplish, right? Remember back to that direction thing, like, what are the tasks which would be decomposed to lead to that direction becoming possible?

125
00:22:01.570 –> 00:22:13.079
Chris O’Neill: And then you’re saying, what’s expected of that person, or in this case an agent, what are you expecting them to do? What’s the task or the project you’re expecting them to do? So you’re setting clear expectations.

126
00:22:13.200 –> 00:22:29.729
Chris O’Neill: And then you don’t just set it and forget it, right? You actually are engaged in the details, and you give feedback to say, how are you doing against that task, right? You know, what things are going well? Let’s celebrate that things are going well, and what things are, you know, off the mark, so you can do less of that and more of the former.

127
00:22:29.730 –> 00:22:43.699
Chris O’Neill: So, and then, and then, yeah, you do hold people to account, if they’re, you know, you celebrate them in the end. So I think there’s a surprising parallel there, between what makes a great manager and what makes a great manager of people or agents. It’s the same stuff.

128
00:22:43.920 –> 00:22:48.249
Chris O’Neill: And the other thing you’re pointing out, like, I’ll tell a quick story, so…

129
00:22:48.310 –> 00:22:58.579
Chris O’Neill: So I was leading a turnaround of Evernote, this is the note-taking application, just to hire a great company, amazing team, had lost our way a little bit, and there was a nervousness to take risk.

130
00:22:58.630 –> 00:23:13.780
Chris O’Neill: And I was working with, Kim Scott, the woman who wrote Radical Candor and many other books. She’s a friend, former colleague, and she was my coach at the time, and I was like, Kim, like, how do we get people to celebrate failures a little bit, or at least to be less worried about making mistakes?

131
00:23:13.780 –> 00:23:22.509
Chris O’Neill: Right? And it’s natural to not want to make mistakes, but, like, how do you normalize it and celebrate it so you just, like, don’t make the same ones, but you’re actually okay making it?

132
00:23:22.510 –> 00:23:30.049
Chris O’Neill: So we came up with this thing, she had this thing called Whoops the Monkey. We adapted it to an elephant for, because it was our brand.

133
00:23:30.130 –> 00:23:39.319
Chris O’Neill: But it was basically, like you said, every week we celebrated people who actually made a whoops, made a mistake, right? Like, took a risk.

134
00:23:39.330 –> 00:23:53.029
Chris O’Neill: And they learn something. And that’s a really powerful thing here, because, look, while we’re all excited about AI’s potential, and rightfully so, right? This is happening at a pace and a slope that I’ve never seen before in my career.

135
00:23:53.130 –> 00:23:59.549
Chris O’Neill: It’s gonna lead to mistakes. The best teams, whether it’s on the product and engineering side, or the go-to-market side.

136
00:23:59.730 –> 00:24:01.980
Chris O’Neill: Are gonna make mistakes about half the time.

137
00:24:02.220 –> 00:24:18.040
Chris O’Neill: You just are. If you’re shooting for something bold, you’re gonna just fall short about half the time, so I take that as a given, but the teams that are gonna separate themselves from the rest are the ones that actually recover from that quickly. Say, take the things that work well.

138
00:24:18.080 –> 00:24:34.839
Chris O’Neill: and then discard the things that don’t. And they compound a whole loop or workflow faster than others. So, to your point about, yes, results, to your point about AI fluency, and your point about investing in data to fuel it all with insights and context.

139
00:24:34.850 –> 00:24:54.749
Chris O’Neill: It’s really about velocity, back to my point. It’s about moving with high velocity so that you learn a little bit. And, you know, it’s not about Big Bang silver bullet stuff. It’s about, did we move a little bit of friction out of the way? Did we improve, you know, conversion by a small little bit in this channel, etc. Those little things.

140
00:24:54.750 –> 00:25:03.750
Chris O’Neill: Compound. So that’s the game, I believe, and that’s the game of, I think, leadership, and then the surprising parallel across humans and agents at the managerial level.

141
00:25:04.630 –> 00:25:22.769
Russell Scherwin: Love that, Chris. And it really brings home the core message for Growth Loop, which, I’ll be honest, the first time I looked at it, I’m like, okay, that’s neat, that’s nice. But hearing the meaning, which is, you know, my first, one of my first CEOs I worked for said, hey, Russell, I need you to waste half of your marketing budget, which is very similar to what you just said, which is.

142
00:25:22.770 –> 00:25:23.110
Chris O’Neill: Yeah.

143
00:25:23.110 –> 00:25:37.159
Russell Scherwin: You gotta be willing to make mistakes, but compound that… compound that word again, with… with a little bit of vulnerability, which is, hey, let’s fail fast, let’s fail forward, and let’s celebrate the wins and the smart losses that we’re learning from.

144
00:25:38.040 –> 00:25:46.709
Russell Scherwin: the velocity’s gonna allow customers to compound the wins and quickly eliminate the losses and get to that competitive edge, and

145
00:25:46.710 –> 00:25:59.150
Russell Scherwin: And I love the way you frame it. The other thing I love that I want to call out for the audience is, when you’re managing and holding AI accountable, it’s… you said the process is set the expectations, engage in the details, give feedback, and hold the account.

146
00:25:59.150 –> 00:26:14.729
Russell Scherwin: and then repeat. You know, it’s… another leader once told me when I was getting into leadership, Russell, it’s not about having the answers, it’s about understanding and being an expert and asking the questions, which is kind of a way of saying be a great prompt engineer. We’ve got 2 minutes left. Can I give you two quick hits?

147
00:26:15.560 –> 00:26:16.439
Chris O’Neill: Go for it.

148
00:26:16.730 –> 00:26:24.239
Russell Scherwin: Alright, the first one. So, one of my side gigs as part of this sabbatical is I’m teaching an MBA class at UGA, and I’m gonna guess we got a couple of my students listening.

149
00:26:24.490 –> 00:26:31.289
Russell Scherwin: What do you say to new folks, or to students, you know, post-grad, undergrad, who are

150
00:26:31.700 –> 00:26:40.279
Russell Scherwin: graduating and are looking for jobs right now and read all the press about, you know, AI’s gonna eat us for lunch and all that stuff. What do you say to folks coming up into the world?

151
00:26:41.740 –> 00:26:44.149
Chris O’Neill: Yeah, be so good they can’t ignore you.

152
00:26:44.990 –> 00:26:49.360
Chris O’Neill: It’s a line from Steve Martin, and I think it’s the best career advice you have, and…

153
00:26:49.510 –> 00:26:56.440
Chris O’Neill: There’s not a company that I’m aware of on this planet right now that isn’t curious about AI,

154
00:26:56.700 –> 00:27:02.919
Chris O’Neill: And or, you know, wants… wants to hire people who have superpowers.

155
00:27:03.070 –> 00:27:12.070
Chris O’Neill: In AI. Your students… I have a 19-year-old son and a 17-year-old daughter, and I’m working with my son on that very question.

156
00:27:12.280 –> 00:27:14.819
Chris O’Neill: And he loves photography, so I said to him, Jack.

157
00:27:15.340 –> 00:27:18.540
Chris O’Neill: Would you like to vibe code a photography website?

158
00:27:18.630 –> 00:27:27.719
Chris O’Neill: Right? And he’s like, Dad, what the… what the hell is that? You, like, tech guy? Like, what does that even mean? So, he’s in the process of vibe coding a website

159
00:27:27.720 –> 00:27:46.660
Chris O’Neill: And it’s really the advice I would give is, like, whether it’s for photography or a website or whatever, experiment with these things, right? There’s never before been so much knowledge and tooling at our disposal, so become insanely, insanely good

160
00:27:46.660 –> 00:27:52.830
Chris O’Neill: at using these tools, in big ways and small ways. So I… in many ways, I know the headlines are.

161
00:27:52.830 –> 00:28:11.419
Chris O’Neill: are in front of our face, and there’s some real… there’s some realness to it, right? There’s, like, you know, companies are going to always try to be as efficient as they can, but I’m of the belief that those who become super fluent and become so good that the people cannot ignore them around this stuff will not only, you know, get hired, but they’ll thrive.

162
00:28:12.660 –> 00:28:28.870
Russell Scherwin: Couldn’t agree more, and I see we’re at time. Greatly, great ending words, make yourself indispensable. That is an economy-proof philosophy. Julia, appreciate you having myself and Chris. Chris, thanks so much for taking a half out of your day to have this great dialogue.

163
00:28:28.870 –> 00:28:29.799
Chris O’Neill: Thank you, Russell.

164
00:28:30.120 –> 00:28:35.279
Julia Nimchinski: Phenomenal panel, thank you so much, Chris and Russell. And lastly, how can our community support you?

165
00:28:36.480 –> 00:28:37.030
Julia Nimchinski: Priceless.

166
00:28:37.030 –> 00:28:38.630
Chris O’Neill: I’m sorry, Julie, can you repeat that?

167
00:28:38.630 –> 00:28:41.530
Julia Nimchinski: community support you? Yeah, where should our people go?

168
00:28:42.160 –> 00:28:55.940
Chris O’Neill: Oh, you can visit our website, you can track me down on LinkedIn. I always love to hear from folks, so, you know, share examples, share challenges. I love to be a small part of this community, and it’s an honor to do so. Thanks for having me today.

169
00:28:56.560 –> 00:28:58.230
Julia Nimchinski: Thanks so much, Russell, how about you?

170
00:28:59.170 –> 00:29:09.180
Russell Scherwin: Oh, gosh, I just, while I’m on a sabbatical trying to be a better dad, I just love the opportunities to engage in conversations that get the… get the cobbs out of my brain, so I appreciate you having Julia.

171
00:29:09.660 –> 00:29:10.560
Julia Nimchinski: Thank you.