Going Beyond Gut Checks

1

00:00:00,000 --> 00:00:05,600

Marketing is one of those horizontals that's going to be changed and transformed the most. Are

2

00:00:05,600 --> 00:00:11,240

we moving beyond simple content generation to foundational marketing transformation? It's not

3

00:00:11,240 --> 00:00:16,120

going to replace sort of the wholesale task, right? It's individual areas that will be affected. And

4

00:00:16,120 --> 00:00:22,080

in product marketing, we need to keep pace with an increasingly fast paced release cycle. In this

5

00:00:22,080 --> 00:00:27,280

episode of AI Cloud Essentials, I explore interesting product marketing use cases for

6

00:00:27,280 --> 00:00:34,040

synthetic research with Suzanne, VP of Product Marketing at CoreWeave. It's not as simple as just

7

00:00:34,080 --> 00:00:38,159

asking for the answer. You need to know what questions to ask and that will never go away. I

8

00:00:38,160 --> 00:00:44,360

don't think. If you lead product marketing, this conversation will transform how you think about

9

00:00:44,360 --> 00:00:47,400

research, messaging and launch velocity.

10

00:00:56,160 --> 00:01:01,430

Hello everyone. Welcome to season two of the AI Cloud Essentials, a podcast series brought to you

11

00:01:01,430 --> 00:01:06,589

by CoreWeave. Today we are talking about one of my favorite areas with one of my favorite

12

00:01:06,589 --> 00:01:12,110

individuals who I've known for a couple of years. Welcome, Susanne. Thanks. Me too. Yeah. We are going

13

00:01:12,110 --> 00:01:18,109

to talk about one of the the favorite topic of everyone in the industry. How are you going to

14

00:01:18,150 --> 00:01:24,989

transform product marketing, how the velocity meets impact and what can be done in that

15

00:01:24,989 --> 00:01:30,870

area. So everyone's talking about for the last couple of years, Susanne, how you can actually use

16

00:01:30,910 --> 00:01:36,429

AI to do content generation. Right. But I personally believe that that's really not meeting

17

00:01:36,430 --> 00:01:41,429

the entire potential that you can actually do. And there's so many interesting use cases where

18

00:01:41,470 --> 00:01:46,989

things fall short. But before we dive into it, let's hear from you. How did you come into this

19

00:01:47,190 --> 00:01:53,110

role? What brought you here, and what are you seeing in terms of when you're setting up an AI

20

00:01:53,150 --> 00:01:58,269

first product marketing group? Thanks, Ritu. Um, it's so great to be here. And you're right, it is the

21

00:01:58,269 --> 00:02:02,979

topic everyone wants to talk about, right? Marketing is one of those horizontals that's

22

00:02:02,980 --> 00:02:08,860

going to be changed and transformed the most. And within that space, product marketing is kind of in

23

00:02:08,860 --> 00:02:13,460

the thick of it, right? Because we're it's almost recursive, right? We're we're launching the AI

24

00:02:13,500 --> 00:02:19,779

products that we have to use AI to launch. So it's a pretty, you know, fortunate moment to be in this

25

00:02:19,779 --> 00:02:25,460

space. And I certainly feel fortunate that way. I did not fall into it in a straight line, I can

26

00:02:25,460 --> 00:02:30,259

assure you. And I think that's indicative of a lot of people's experience right now. Yeah, I came to

27

00:02:30,300 --> 00:02:37,179

marketing from a technology background. I worked in smart cities and IoT and different spaces that

28

00:02:37,179 --> 00:02:43,100

were also very disruptive at the time. So I feel like I'm just following the disruption, if you

29

00:02:43,100 --> 00:02:48,940

will. And marketing is a great kind of focal point where some of these technologies have to be

30

00:02:48,940 --> 00:02:53,179

explained, they have to be demystified. And so that's certainly what brought me to it, this

31

00:02:53,179 --> 00:02:59,539

desire to make it more, um, um, accessible to more people. And that's what I love about, you know,

32

00:02:59,580 --> 00:03:05,179

marketing and product marketing in particular. Our job is to demystify these systems, demystify these

33

00:03:05,179 --> 00:03:10,819

technologies, and show how they bring value. And so that's what brought me into product marketing. And

34

00:03:10,820 --> 00:03:17,779

in this AI moment, you know, there's certainly a lot of need for diversification, right? Yeah. So,

35

00:03:18,020 --> 00:03:23,300

you know, like I have some similar threats in my career as well. I've been a, you know, product

36

00:03:23,300 --> 00:03:28,819

marketer. I have also been a product manager. And I sit back and think, one of my, you know, jobs that I

37

00:03:28,860 --> 00:03:34,740

hated the most was to be, you know, creative about product naming. Right. Even in my current role, when

38

00:03:34,740 --> 00:03:40,059

I have started my own startup, that's been my biggest, biggest challenge, right? Because there's

39

00:03:40,060 --> 00:03:44,699

so many creative minds moving around. But, you know, how do you settle on what was going to resonate?

40

00:03:44,700 --> 00:03:49,299

So when you started looking into this, where did you start with and what are the use cases that

41

00:03:49,340 --> 00:03:55,619

kind of gravitated to you and your team in terms of, you know, kind of making the best use of AI to

42

00:03:55,660 --> 00:04:01,009

kind of accelerate the velocity and impact. I mean, it's great that you're framing it as different

43

00:04:01,009 --> 00:04:06,689

use cases because like, you know, in a lot of spaces, there's it's not going to replace sort of

44

00:04:06,729 --> 00:04:11,530

the wholesale task, right. It's individual areas that will be affected. And in product marketing, we

45

00:04:11,530 --> 00:04:18,170

need to keep pace with an increasingly fast paced release cycle. Right. And so some of the challenges

46

00:04:18,170 --> 00:04:24,209

that come with that are things like naming consistency, naming conventions, messaging

47

00:04:24,529 --> 00:04:30,289

strategies, persona mapping. And so we've we've started to tackle some of these different use

48

00:04:30,289 --> 00:04:36,449

cases and kind of one by one here at CoreWeave. And then with my team and then also with

49

00:04:36,450 --> 00:04:42,529

interesting partners like Evidenza. And so the the use case of naming is a really interesting one

50

00:04:42,529 --> 00:04:48,969

because it's, it's both in terms of the ideation phase. But then it's also in terms of validating

51

00:04:48,970 --> 00:04:54,328

the choices you're making against different audiences. And I think we're increasingly able to

52

00:04:54,369 --> 00:04:59,840

be specific about what's going to resonate. And so certainly some of the products that we've looked

53

00:04:59,840 --> 00:05:06,800

at and have launched at a very rapid pace over the last several months, we've leveraged tools to

54

00:05:06,839 --> 00:05:13,079

figure out which names will resonate the most. So things like zero egress migration are very clear.

55

00:05:13,079 --> 00:05:18,879

And and maybe we would have picked that name on our own as well. But having that kind of

56

00:05:18,920 --> 00:05:24,639

counterbalance and ability to bounce ideas off of real different personas has been really, really

57

00:05:24,640 --> 00:05:30,839

helpful and fruitful. Yeah. You know, uh, you know, it that I've been in an analyst role as well

58

00:05:31,280 --> 00:05:37,679

in, in, in the past couple of years. And one thing that always was a pet peeve for me is that while

59

00:05:37,680 --> 00:05:43,839

it's exciting to conduct, you know, industry research, and it was really time taking and

60

00:05:43,840 --> 00:05:49,279

consuming and costly and bursty, and sometimes we used to joke that by the time you get the survey

61

00:05:49,280 --> 00:05:54,839

results, it might be out. Totally. Yes. Yeah. Does it really make sense? And sometimes it's also very

62

00:05:54,839 --> 00:06:00,509

difficult to get the right set of personas that you're targeting. And synthetic research was

63

00:06:00,510 --> 00:06:05,069

always on top of my mind as to how you could do that. So could you talk a little bit about, you

64

00:06:05,070 --> 00:06:09,310

know, how you'll use some of this synthetic research, just the examples that you just talked

65

00:06:09,310 --> 00:06:14,589

about. But beyond that, what were the lessons learned? Walk me through the process so that the

66

00:06:14,590 --> 00:06:20,549

viewers who are listening to us can benefit from it. It's a great example, right? And especially

67

00:06:20,549 --> 00:06:26,029

because if you think about yeah, like you said, it takes so much time to do it in person, to do it

68

00:06:26,030 --> 00:06:30,829

with real individuals. And some of these individuals might not be that accessible. If

69

00:06:30,829 --> 00:06:35,949

you're looking at, you know, CEOs, you know, of a particular from a particular industry, how do you

70

00:06:35,950 --> 00:06:42,469

find enough of them to have a really indicative sample? And so this has been very interesting for

71

00:06:42,469 --> 00:06:48,909

us. And we've certainly been drawing on this, um, ability to test ideas in a rapid

72

00:06:48,910 --> 00:06:55,349

way, like in days, you know, not months. Right. And it's been super helpful. And it's, it's it's been

73

00:06:55,350 --> 00:07:00,630

really helpful in terms of also challenging our ideas. right? Because on some level, we've also

74

00:07:00,630 --> 00:07:05,630

gotten feedback. A lot of the times that resonates, we're like, okay, we're on the right track. But in

75

00:07:05,630 --> 00:07:10,028

some cases, we've been surprised. We we thought something would resonate and it didn't. And so

76

00:07:10,029 --> 00:07:15,870

that's been really eye opening and also kind of breaking boundaries, because in the past we may

77

00:07:15,870 --> 00:07:20,389

have not used research at all. Right. I mean, a lot of marketers, they're under time pressure. The

78

00:07:20,389 --> 00:07:25,029

launch is happening. You have to make a decision, right? You can't wait around for a six month study.

79

00:07:25,070 --> 00:07:31,389

And so being able to draw on research and to step back and to try to be more objective and to try

80

00:07:31,429 --> 00:07:38,269

to reflect on your own assumptions and biases, right, in a way that's much more, um,

81

00:07:38,309 --> 00:07:45,109

constructive, has been eye opening and empowering for the team. And so I see us just thinking

82

00:07:45,109 --> 00:07:52,029

harder, challenging ourselves more, and, you know, also starting to create a practice of, of

83

00:07:52,030 --> 00:07:57,389

feedback and, and conversation that we might not even have had before because we were too busy

84

00:07:57,390 --> 00:08:03,858

executing and meeting the deadlines. Yeah, because I'm sitting and thinking with you here is that in

85

00:08:03,859 --> 00:08:08,619

the past, I've seen a lot of you reaching out to the analyst for the message testing. Right. But it

86

00:08:08,619 --> 00:08:13,739

still could be a biased opinion or a very small set of opinion surveys. Have there, you know, long

87

00:08:13,780 --> 00:08:19,899

time and, you know, costs and inefficiencies. But there are some gotchas in these systems as well.

88

00:08:19,939 --> 00:08:24,860

Right? So it could create some friction. Could could create some hallucination, could end up into

89

00:08:24,899 --> 00:08:30,739

some kind of a descent. So, uh, did you guys kind of counter it or balance it with some kind of

90

00:08:30,739 --> 00:08:35,699

traditional methods of research? Along with that, what were the best practices that you learned out

91

00:08:35,700 --> 00:08:41,099

of it? It's a great question, and it's super important because we absolutely are coming back

92

00:08:41,099 --> 00:08:48,059

to, you know, validating our assumptions. We are testing against real focus groups. We're also it

93

00:08:48,060 --> 00:08:54,539

depends on what study you're running, right. Are you doing a naming test or a message test or are

94

00:08:54,539 --> 00:08:59,210

you trying to create sort of longitudinal views, at which point you cannot really rely on the

95

00:08:59,210 --> 00:09:06,209

models yet as much, right? So that's where you want to use real panels of people. And then that allows

96

00:09:06,209 --> 00:09:11,609

you to also continuously cross-check your own the data you're getting back from the synthetic

97

00:09:11,610 --> 00:09:16,769

research against what's happening among the, you know, the communities out there. And they're

98

00:09:16,769 --> 00:09:21,968

progressing, but the models are changing, too. Right. And at the same time and in parallel. And so just

99

00:09:21,969 --> 00:09:27,049

keeping these things in balance and having a back and forth so that you're asking the right

100

00:09:27,049 --> 00:09:32,289

question with the right method like that doesn't change. Right. Like even it's always it that will

101

00:09:32,289 --> 00:09:36,849

always be critical. And so that's where the experience and the and the knowledge of the

102

00:09:36,849 --> 00:09:43,009

marketers comes back into play. Right. Yeah. It's it's not as simple as just, you know, asking for

103

00:09:43,049 --> 00:09:47,729

the answer. You need to know what questions to ask. And that will never go away I don't think. Yeah. So

104

00:09:47,729 --> 00:09:52,489

as you kind of very rightfully said, it's not either or it's a complementary. You have to kind

105

00:09:52,489 --> 00:09:57,330

of use the right set of tools in the right place and complement it each other. Not to ask you

106

00:09:57,330 --> 00:10:01,479

something personal at this level, when you're managing this team, you might have been confronted

107

00:10:01,479 --> 00:10:06,879

with your team members who might have not had the experience or knowledge about these tools. Did

108

00:10:06,880 --> 00:10:13,599

everyone embrace it? Seemingly. You know. Everyone always know. I mean, you know, I'll be honest. I

109

00:10:13,600 --> 00:10:18,759

think I'm just maybe I'm just really lucky, too, that we're we're creating communities of folks

110

00:10:18,759 --> 00:10:24,559

who are really, really curious and maybe coming to marketing from a less traditional background. I've

111

00:10:24,559 --> 00:10:31,479

always, um, brought other people along who are also very, very curious. So it's less about

112

00:10:31,479 --> 00:10:38,119

resistance and more about getting past the initial kind of experimentation phase. Like that's

113

00:10:38,119 --> 00:10:43,519

where I've seen the most challenge, right? So we we dabble, we try, but now we're at this new moment

114

00:10:43,520 --> 00:10:50,358

where we're starting to implement new systems that are really propelling us forward. So, you know,

115

00:10:50,400 --> 00:10:57,159

competitive intelligence work or like you said, message testing. Yeah. Um, persona development and

116

00:10:57,159 --> 00:11:04,159

refinement. Um, these are all things that are now part of our common practice. So it's

117

00:11:04,159 --> 00:11:09,560

less about resistance and more about giving everyone the tools they need to get to the right

118

00:11:09,600 --> 00:11:16,159

level of comfort. And so that's where I'm finding the most need to, to to jump in and give support

119

00:11:16,159 --> 00:11:20,359

and then a little bit. It's also about creating space because I don't know how it is for you. But

120

00:11:20,359 --> 00:11:25,399

when I'm under time pressure I, I you know what they say, right? You fall to your level of training.

121

00:11:25,400 --> 00:11:29,799

So you go back to your safe space. You go back to the things you know how to do. And that's not the

122

00:11:29,799 --> 00:11:34,759

time to try something completely different. And so how do you create a space where people can

123

00:11:34,759 --> 00:11:40,079

experiment. So, you know, we're trying to, you know, come back to the old school kind of Fridays are

124

00:11:40,080 --> 00:11:46,960

for innovation. You know, fewer meetings open up longer periods of time. So people have

125

00:11:46,960 --> 00:11:52,359

the breathing room to do something differently. And that's key, I think, in all of. This, I think

126

00:11:52,359 --> 00:11:57,868

that's very well said. And also empowering them to kind of learn and iterate and, you know. Fail. Yeah.

127

00:11:57,909 --> 00:12:02,070

And sometimes it could be bad, right? And then that's okay. And it lessens and figured out

128

00:12:02,229 --> 00:12:08,309

totally. So I would love to understand. Where do you see the industry? Headache. You know, what would

129

00:12:08,310 --> 00:12:14,988

be your kind of insight that you see from product marketing kind of evolving in the era of AI? Uh,

130

00:12:14,989 --> 00:12:20,669

from a roll perspective, from a usage perspective of AI. I mean, it's going to be central to

131

00:12:20,710 --> 00:12:26,188

everything we do. I mean, I think that is is a given. Honestly, it's a given. It's a given because,

132

00:12:26,229 --> 00:12:32,630

you know, we our craft is related to to language, to expression. Right? So regardless of how things

133

00:12:32,630 --> 00:12:38,389

move, we have to embrace this and weave it into how we imagine, you know, products should be

134

00:12:38,390 --> 00:12:43,909

launched, products should be framed. And so we need to harness it for our purposes. Like there's no

135

00:12:43,909 --> 00:12:49,789

question about it. And so that will happen I think where the the interesting components will be is

136

00:12:49,790 --> 00:12:53,989

around the speed and the level of iteration. You know, one of the things we always talk about is,

137

00:12:54,030 --> 00:12:59,299

you know, things like message consistency. You know, message consistency is critical, right? But also

138

00:12:59,340 --> 00:13:05,900

moving with your customers is critical, right? So do you have a way to, um, evolve

139

00:13:05,940 --> 00:13:11,779

quickly enough where you're still being consistent but also keeping pace? Because that is

140

00:13:11,780 --> 00:13:16,739

where I worry the most. You know, like the the customers are inventing new products. I mean,

141

00:13:16,780 --> 00:13:21,019

whether it's, you know, from all kinds of directions, whether it's an enterprise or AI,

142

00:13:21,059 --> 00:13:27,780

native startups or the leaders themselves, they're all innovating so quickly that the

143

00:13:27,780 --> 00:13:32,619

product marketing risk is falling behind. If it's not using the same tools to keep pace. And so

144

00:13:32,619 --> 00:13:38,579

that's where we need to stay focused and and just move very, very quickly so that we can, you know,

145

00:13:38,620 --> 00:13:44,779

provide the insight and the, you know, that the mystification along with the products that are

146

00:13:44,780 --> 00:13:50,339

coming out. So do you and your team members have some wish list that I wish the the technology

147

00:13:50,340 --> 00:13:55,499

suppliers kind of stall this problem, and I wish they did a better job. Um, I don't know if we have

148

00:13:55,499 --> 00:14:01,810

a wish list. I think our wish list is more about how we how. How quickly we can test things, right?

149

00:14:01,849 --> 00:14:08,728

Can we test things more readily? Can we try more things in different places? Can we create more, um,

150

00:14:09,609 --> 00:14:16,569

persona based journeys in rapid succession? Can we then weave that back in? A lot

151

00:14:16,570 --> 00:14:22,128

of conversations are happening now around loops, right? And marketing. And so it's less about kind

152

00:14:22,129 --> 00:14:26,969

of a linear funnel, and it's much more about this iterative conversation you're having with

153

00:14:26,970 --> 00:14:31,889

different stakeholders in, in a customer. How do you bring all that together and then make it even

154

00:14:31,890 --> 00:14:37,449

more engaging and interesting and beneficial really for everyone, right. Because at the end of

155

00:14:37,450 --> 00:14:42,329

the day, you're providing real value to customers and you want them to see it, feel it in a way that

156

00:14:42,330 --> 00:14:48,488

makes sense to them. I always kind of joke that, you know, we want less of AI slop and more

157

00:14:48,489 --> 00:14:55,049

maturity and more angles and knobs for us to kind of manage it so that the quality improves. It's a

158

00:14:55,050 --> 00:15:00,408

powerful tool for us to use, and how we get the solutions that give us the power to kind of move

159

00:15:00,409 --> 00:15:06,169

it forward with real business, you know, value. Right. Yeah. Awesome. Awesome. So it seems like

160

00:15:06,209 --> 00:15:11,009

you're doing all the right things with your team members in terms of empowering them, providing

161

00:15:11,009 --> 00:15:15,968

them the flexibility. I think the reason I was asking this question is because that sometimes,

162

00:15:16,009 --> 00:15:20,809

you know, some people kind of, you know, are hunkered down and they have the human paranoia,

163

00:15:20,809 --> 00:15:26,249

which is human nature. Uh, so, Suzanne, as we're kind of wrapping up this particular episode, I'd love

164

00:15:26,249 --> 00:15:31,889

to get your take on your parting advice to the viewers of this particular episode. You know, I

165

00:15:31,889 --> 00:15:36,409

think the most important thing and everyone says this right now, but jump in, right? So don't wait.

166

00:15:36,450 --> 00:15:42,169

Don't hesitate. This isn't going away. Um, and there's so much potential, right? The opportunities

167

00:15:42,170 --> 00:15:49,129

are vast. So I would encourage everyone to jump in. And then I would encourage folks to remember what

168

00:15:49,169 --> 00:15:55,840

makes them great marketers. And I think the core of that is discernment and taste and the ability

169

00:15:55,840 --> 00:16:02,439

to actually figure out what is. You know, you called it AI slop versus what is really

170

00:16:02,439 --> 00:16:07,639

valuable kind of nuggets of insight that will resonate with customers and then use the

171

00:16:07,640 --> 00:16:13,719

technology and the tools to iterate faster, to try more things so that you can actually progress

172

00:16:13,719 --> 00:16:20,279

your own thinking faster. So focus on your own discernment, trust it, and and then keep

173

00:16:20,320 --> 00:16:25,479

iterating quickly so that you can keep pace with the innovations of the stakeholders you're

174

00:16:25,479 --> 00:16:31,639

serving within your organization. Awesome. Awesome. So I'd like to wrap it up by saying that don't

175

00:16:31,640 --> 00:16:37,439

just automate content. Rethink your foundational practices like validation and decision making. And

176

00:16:37,440 --> 00:16:42,639

with that, we have come to the end of this episode. Thank you everyone for joining us, and stay tuned

177

00:16:42,680 --> 00:16:44,759

for more insights from us. Thank you.

‍