Business Models in the Agentic Age: Monetization & Disruption

00:00:11.140 –> 00:00:12.840
Julia Nimchinski: And we are live

2
00:00:13.020 –> 00:00:22.629
Julia Nimchinski: welcome back to day. 3 of the agenda AI summit, and we say some of the best executive, roundtable and fireside chats were last.

3
00:00:23.040 –> 00:00:30.279
Julia Nimchinski: We are for a real treat here, so be sure to follow the conversation along on agency slack. It’s on the right

4
00:00:30.390 –> 00:00:34.630
Julia Nimchinski: lower corner of your screen, and our sponsors

5
00:00:35.090 –> 00:00:41.930
Julia Nimchinski: and Alison, welcome to the show. We are going to pray many Medina in a second.

6
00:00:41.930 –> 00:00:46.160
Allison Snow: Okay, excellent. Well, while he is joining us

7
00:00:46.549 –> 00:01:09.200
Allison Snow: very quickly, I have been an attendee for the last few days. Thrilled to be here on day 3 thrilled to be here. With manny. We are, of course, in a in a larger AI summit called the Dawn of AI, native Gtm, which is super exciting. And we’re here talking about business models in the agentic age, monetization and disruption.

8
00:01:09.200 –> 00:01:18.619
Allison Snow: I’m reviewing all this not because you haven’t seen the agenda. Of course you have. But but who better to talk about this topic than the founder of a company called Paid AI.

9
00:01:18.620 –> 00:01:29.640
Allison Snow: And if you go to that website. And I suggest that you do. It’s clever. It’s funny. It’s a good site. The Ctas. The calls to action that that Manny and his team have put together are, get paid, get paid like you mean it.

10
00:01:29.680 –> 00:01:35.300
Allison Snow: And I think that’s what we’re all kind of trying to do here. So I’m absolutely thrilled to welcome Manny. Manny. Are you around yet?

11
00:01:41.670 –> 00:01:43.210
Julia Nimchinski: I’m seeing him.

12
00:01:46.960 –> 00:01:47.520
Allison Snow: Cool.

13
00:01:47.860 –> 00:01:59.910
Allison Snow: Well, then, behind his back, I will do a little introduction. Manny is the founder of Paid AI. He was the founder of outreach one of the reasons I’m really excited to have this conversation with him and spend 30 min.

14
00:02:00.060 –> 00:02:00.970
Allison Snow: Hey, Manny.

15
00:02:01.590 –> 00:02:04.540
Manny Medina: Hi! Sorry I made it work by myself, so proud.

16
00:02:04.540 –> 00:02:05.560
Allison Snow: You are.

17
00:02:07.425 –> 00:02:08.520
Allison Snow: Yes.

18
00:02:09.160 –> 00:02:14.140
Allison Snow: yes, I can hear you. Can. I don’t know that that folks can see you. I can only see me at the moment.

19
00:02:15.480 –> 00:02:16.380
Manny Medina: I can see.

20
00:02:16.380 –> 00:02:18.779
Manny Medina: See it, Julie. Can you see me.

21
00:02:18.780 –> 00:02:19.420
Julia Nimchinski: Yep.

22
00:02:19.950 –> 00:02:25.514
Allison Snow: Oh, great! And that’s just me. Excellent Manny I was. I was introducing you behind your backs. Now that you’re here.

23
00:02:26.280 –> 00:02:42.969
Allison Snow: I’ve only done a brief kind of introduction, but but I’m very excited to be here with you for many reasons. I’m so excited about what you bring to this topic which I’ve reviewed with the crew here, even though they they know the agenda. So we’re kicking off day 3 very exciting. Would you like to introduce yourself for a little bit.

24
00:02:43.909 –> 00:02:50.379
Manny Medina: Sure. My name is Dina. This crowd will probably remember me as the co-founder of of outreach

25
00:02:51.020 –> 00:02:56.630
Manny Medina: but today we’re here to talk about AI and AI agents. I’m also the founder of paid

26
00:02:56.910 –> 00:03:01.889
Manny Medina: and paid is the business engine for AI agent and AI agent builders.

27
00:03:02.080 –> 00:03:24.440
Manny Medina: We identified a gap in the market based on my experience in the past for creating the entirety of the back office that an AI agent needs, including monetization, cost management showing the value. And those are all things that we had that we didn’t have a great solution in Sas and AI agent makes the problem worse. So we’re set out to help build the AI agent. Economy going forward.

28
00:03:25.580 –> 00:03:40.529
Allison Snow: Excellent. I am Allison Snow. I am the Cmo of a firm event tech firm called Streampoint Solutions. I’m here in Boston and Manny, as far as I’m concerned, my role for 30 min is to is to ask you questions that helps people get

29
00:03:40.720 –> 00:03:44.429
Allison Snow: as much knowledge from you as possible. In the short 30 min we have together.

30
00:03:44.770 –> 00:03:49.689
Manny Medina: I forward to it and look forward to the conversation. I’m sure I’m going to ask you questions, too.

31
00:03:49.690 –> 00:04:09.349
Allison Snow: Excellent. Yeah, I’m certainly ready for that. I’m going to start at the beginning, even though we’re on day 3 of the summit a lot of folks because I’ve been monitoring the chat are are here for multiple sessions. So I don’t want to get too basic. But I do want to do a little context setting. So I’m hoping I can ask you. And and you’ll be okay with it. Just as a baseline.

32
00:04:09.500 –> 00:04:20.199
Allison Snow: What is in the what is an agent in the context of what you’re building today? And what I haven’t heard answered so far is, how is it different from automation as we’ve known it in Sas.

33
00:04:20.829 –> 00:04:21.429
Manny Medina: Perfect.

34
00:04:21.629 –> 00:04:34.169
Manny Medina: So an agent has evolved to where we are right now, where it has agentic behavior and agentic behavior means that it can perform tasks without human intervention.

35
00:04:34.609 –> 00:04:53.449
Manny Medina: So what we had in the past coming from Sas and as a Sas builder myself was that Sas was operated by an individual by a human being, and it was the ability for you to operate a workflow or or manipulate data, or to, you know, generate some value

36
00:04:53.549 –> 00:04:57.729
Manny Medina: from the the thinking of a human being aided by a computer.

37
00:04:58.149 –> 00:05:05.909
Manny Medina: Then we moved into copilot mode where the agent was helping the human being, but the human being had to invoke the agent and prompt it to actually get that value.

38
00:05:06.069 –> 00:05:08.039
Manny Medina: Now, we’re up to the point in which

39
00:05:08.189 –> 00:05:13.659
Manny Medina: the agent understands the context in which the value needs to be delivered, and delivers the entirety of the thing.

40
00:05:14.009 –> 00:05:27.989
Manny Medina: So the aging is taking the whole. A whole set of workflows, or or tasks, or even functions, you know, out of the human’s plate, and performing themselves, and then delivering back to the human being for for

41
00:05:28.329 –> 00:05:30.359
Manny Medina: or as an outcome.

42
00:05:30.639 –> 00:05:36.509
Manny Medina: So that’s what agents are. It’s ability for you to take the whole function that was previously done by a person

43
00:05:36.609 –> 00:05:38.839
Manny Medina: and perform it automatically.

44
00:05:40.410 –> 00:05:45.560
Allison Snow: It’s been pretty remarkable to watch this happen, and you just described it really succinctly. But the

45
00:05:46.430 –> 00:05:58.830
Allison Snow: the the road from not so independent, human, driven to very independent agents on the other side of that spectrum. It’s really quite fascinating. I know everyone in the audience is is excited to learn more about it.

46
00:05:58.940 –> 00:06:03.449
Allison Snow: Thanks for under thanks for helping us understand how it’s how it’s different from Sas.

47
00:06:03.610 –> 00:06:15.700
Allison Snow: I think what folks are really excited about. And and when I was introducing you behind your back, I talked about, paid AI and just encourage folks to go to the site, and how much I admire the cta of get paid like you mean it monetize like you mean it.

48
00:06:16.522 –> 00:06:17.609
Allison Snow: And I think that’s.

49
00:06:17.610 –> 00:06:22.279
Manny Medina: I know it will be big on this crowd, because we’re all here, you know. It’s all about getting paid, so.

50
00:06:22.280 –> 00:06:25.020
Allison Snow: I think they’re gonna eat it up. Yeah, think

51
00:06:25.890 –> 00:06:44.229
Allison Snow: so. I’m excited about it. But but we we sort of know a little bit about. Oh, cool. We know a little bit about how traditional Sas models monetize access right where a lot of us are are so steeped in that we are just sas all day long. But this must change right? So so what does monetization look like when the agent

52
00:06:44.930 –> 00:06:46.960
Allison Snow: is the actor, as you just described?

53
00:06:47.630 –> 00:06:53.570
Manny Medina: So that’s that’s what drove me. So that question was a fundamental question that drove me to start this company

54
00:06:53.870 –> 00:06:59.449
Manny Medina: is that when you have an agent. The concept of seeds doesn’t exist anymore.

55
00:06:59.920 –> 00:07:09.919
Manny Medina: and we’ve been selling seeds for the better part of 20 years, and all of a sudden we are replacing the seeds with an agent that is performing the function that the seeds used to do

56
00:07:10.560 –> 00:07:17.629
Manny Medina: so. If I think it. The alarm bell was sounded about 2 to 3 years ago, when agents started becoming into online

57
00:07:17.740 –> 00:07:23.370
Manny Medina: and people started freezing seeds so you wouldn’t see a seed contraction. But you will see a seed

58
00:07:24.760 –> 00:07:32.530
Manny Medina: and seed stabilization. So your Nrr. That was seed expansion. It started going away. And now all you had in front of you was.

59
00:07:32.630 –> 00:07:36.550
Manny Medina: you know selling seats of a different product. But you’re still selling seats.

60
00:07:36.680 –> 00:07:43.210
Manny Medina: and I think the the train is only moving into one direction. In that the agent is replacing so much of the human task

61
00:07:43.420 –> 00:07:47.129
Manny Medina: that you start needing fewer seats to perform the same function.

62
00:07:47.610 –> 00:07:58.330
Manny Medina: So when you monetize the agent, you can’t do it by seats. So seats. Now go away because it’s a it’s a wrong form factor to to capture the value. Okay, so what else do you have?

63
00:07:58.490 –> 00:08:01.480
Manny Medina: So you have credits or transactions.

64
00:08:01.630 –> 00:08:05.059
Manny Medina: but credits and transactions is a is a is, a bit of a.

65
00:08:05.480 –> 00:08:34.019
Manny Medina: It’s also the wrong way to think about it, because your aging is not delivering credits or transactions. The agent is performing all this body of work, number one and number 2 credits become undistinguishable between agent and agent, and what that forces is to say. 2 agent companies, doing roughly the same work will go at each other with a lower credit or a lower per transaction fee, and that just creates a race to the bottom. So I think that agent companies are waking up to the fact that they are delivering way, more value

66
00:08:34.039 –> 00:08:37.990
Manny Medina: that is captured with the current mechanisms. And you need a

67
00:08:38.270 –> 00:08:42.939
Manny Medina: a new technology, a new framework, a new way of thinking.

68
00:08:43.200 –> 00:08:46.360
Manny Medina: to of how you capture value for an agent, and what we are

69
00:08:46.460 –> 00:08:51.539
Manny Medina: seeing in the marketplace is that agents are fundamentally capturing workflows

70
00:08:51.960 –> 00:08:54.500
Manny Medina: and workflows can be workflows that are

71
00:08:55.452 –> 00:09:11.240
Manny Medina: in in in that are sort of like back office workflows or workflows that are not, you know, seen by a human being, or could be outcomes. Right? So, for instance, to to this crowd, sending an email is actually an outcome to some degree right. If you send an email, you have to go research the individual.

72
00:09:11.240 –> 00:09:40.119
Manny Medina: You have to make sure that you have the right name, and and you have to make sure that the information that you put in putting together in that email is relevant and timely. And then you have to compile that email and send it. That took a bunch of different activities for the agent to perform. And then off the email went. And that could be a workflow. An outcome could be the meeting generated by that meeting, by that email or a positive reply, or the retention event that happened because of that meeting of the expansion event that happened because of that meeting. So all those are, you know, outcomes.

73
00:09:40.120 –> 00:09:46.260
Manny Medina: and that is the kind of value that the agent is delivering. And that’s how. And we are arguing that

74
00:09:46.370 –> 00:09:51.419
Manny Medina: the the builders, the the people selling agents should be capturing a percentage of that value.

75
00:09:51.820 –> 00:09:52.150
Allison Snow: Yeah.

76
00:09:52.150 –> 00:09:54.699
Manny Medina: And and we’re building the technology to help doing that.

77
00:09:55.820 –> 00:10:08.349
Allison Snow: I think I think something really interesting, and let me know what you think of this and you’re also allowed to skip any question that you don’t like. But we’re here in this audience, and I’m here as a Cmo. And I’m listening with my marketing ears. My seller ears. My.

78
00:10:08.920 –> 00:10:14.200
Allison Snow: my, you know my function ears, but it’s not often that you hear

79
00:10:14.310 –> 00:10:35.100
Allison Snow: something about the way that we change marketing and sales that also might help us as buyers, right? This sort of other side of my personality, where I’m interested in capturing value as a business buyer as well, and models that are new like this. And and I think that’s just kind of interesting thing to explore a little bit, that this is good news for buyers and sellers.

80
00:10:35.480 –> 00:10:37.720
Manny Medina: I absolutely agree. And I think that

81
00:10:37.870 –> 00:10:58.650
Manny Medina: you know we all sellers. We wanted that, to begin with, so that you know, we I was just talking to a vendor not too long ago, that was telling me. Oh, you know, instead of using paid, I’m going to Forrester, and they are doing a 12 month study to give me an roi that I can just put on my website, say, according to Forrester, you know I am 512 roi over 6 months.

82
00:10:58.820 –> 00:11:07.380
Manny Medina: and I mean, that’s cute and all. But at the end of the day nobody reads us anymore. Like the the the perception of value is in the eye of the buyer.

83
00:11:07.530 –> 00:11:16.349
Manny Medina: and if you, if they don’t perceive the value as being delivered, if they don’t see it, if they don’t understand it. They’re not going to attribute it to you as as a seller.

84
00:11:16.470 –> 00:11:32.999
Manny Medina: so a better way to do value delivery is to agree with the buyer. What kind of outcome are you expecting out of this? Why are you buying the software? So what happens a year from now, when the software has done its thing. What are you expecting to see? And once and that is such a productive conversation

85
00:11:33.120 –> 00:11:42.190
Manny Medina: that allows you to then map out exactly what is the delivery point to labor. And how are you going to report back to the buyer that you actually did deliver the value that you promised? You’re going to do.

86
00:11:42.470 –> 00:11:46.420
Manny Medina: You know, we’re all here are in the business of of promise making.

87
00:11:46.700 –> 00:11:49.899
Manny Medina: Well, now, we’re going to be in the business of promise, delivering too.

88
00:11:50.350 –> 00:11:55.629
Allison Snow: Yeah, yeah, you promise, you make promises and hopefully monetize those promises.

89
00:11:56.000 –> 00:12:15.439
Allison Snow: Really interesting. Yeah, everything that you’re describing, of course, makes the a a lot of times. Now, in this economy, whatever have you? We are so concerned with renewals, right? We just need to be so. Our our functions really have to shift, I think. Let’s just say it was 50 50 new business versus versus renewals.

90
00:12:15.510 –> 00:12:38.010
Allison Snow: Whatever your percentage was, it’s changed, I think, in favor of renewals, at least in the businesses I’ve served for for the last 5 years, and we’re always searching for ways to make those renewal conversations just, you know not only easier but frictionless. Right? We understood what you wanted. In the 1st place, we talked about it, set some expectations, and and here’s some evidence that we kept that promise that we made.

91
00:12:38.230 –> 00:12:41.150
Manny Medina: A 100%. And and I think as a as a seller.

92
00:12:41.320 –> 00:12:47.090
Manny Medina: it’s so much easier to sell value when the way that you price is through value.

93
00:12:47.680 –> 00:12:54.039
Manny Medina: So I was listening to a podcast where the Cro of Sierra, who is, you know, well known player in the space

94
00:12:54.150 –> 00:13:00.949
Manny Medina: that sales outcomes, was talking about how easy, what a dream is as a seller to, or even as a marketer.

95
00:13:01.060 –> 00:13:09.839
Manny Medina: to say that you only buy Sierra, you only get outcomes. There is no setup fee, there is. No, there is no proceed. There’s none of that like you. If you don’t get the outcome. If you don’t pay.

96
00:13:09.980 –> 00:13:10.950
Manny Medina: That is

97
00:13:11.160 –> 00:13:17.880
Manny Medina: such a disruption. And as a marketer I love the fact that now you can tie your story all the way through monetization.

98
00:13:18.160 –> 00:13:25.220
Manny Medina: You know in the past we used to make all these claims around value delivery that were not substantiated. By the way, we Price, because at the end of that you still have to buy a seat

99
00:13:25.520 –> 00:13:34.010
Manny Medina: or a credit. Those 2 are not the outcome, the outcome is the outcome, and if you get to, if you get paid for delivering an outcome. Then that’s that’s a liming all around.

100
00:13:35.430 –> 00:13:38.009
Allison Snow: Yeah, that’s far more compelling.

101
00:13:39.130 –> 00:13:44.019
Allison Snow: far more compelling. I wonder what you think, where you can tell the audience about what this.

102
00:13:44.250 –> 00:13:47.589
Allison Snow: what this changes a little bit as far as our relationship with technology.

103
00:13:47.800 –> 00:13:54.330
Allison Snow: More specifically, it sounds like agents change the role of software which we’ve. You know, we’ve

104
00:13:55.250 –> 00:14:04.110
Allison Snow: considered this a tool. I’m still just working in tools working in software. Increasingly. My my friend, Chat Gpt is a partner, you know. Increasingly

105
00:14:05.050 –> 00:14:14.680
Allison Snow: thought, partner, I’m I’m running ideas. I’m saying, oh, this way, and it’s fabulous. But how do agents specifically change the role of software from?

106
00:14:15.250 –> 00:14:18.630
Allison Snow: And my suspicion is from tool to partner.

107
00:14:19.060 –> 00:14:25.469
Manny Medina: I think that it changes the software from tool to partner when it anticipates your need.

108
00:14:25.750 –> 00:14:26.640
Allison Snow: Hmm,

109
00:14:27.990 –> 00:14:30.019
Manny Medina: And don’t we all live for that?

110
00:14:30.700 –> 00:14:33.229
Manny Medina: For the pleasant surprise of something done

111
00:14:33.900 –> 00:14:38.629
Manny Medina: that you’re thinking about doing? And now it’s just done at the quality that you need.

112
00:14:39.080 –> 00:14:45.650
Manny Medina: So imagine, if that, you know, campaign is just orchestrated and ready to go for your issues to approve when you wake up, and it’s just there

113
00:14:46.445 –> 00:14:51.939
Manny Medina: enough, and I would love your take on this. I haven’t seen as much disruption in marketing agenda.

114
00:14:51.940 –> 00:14:52.260
Allison Snow: More.

115
00:14:52.670 –> 00:14:57.169
Manny Medina: That, as I’ve seen everywhere else, including sales. Marketing for some reason, is just not.

116
00:14:57.600 –> 00:15:02.610
Manny Medina: you know, coming through in the way that is like I see it in Hr. I see it in

117
00:15:02.790 –> 00:15:06.300
Manny Medina: supply chain. I see it in sales. I see it in

118
00:15:06.610 –> 00:15:11.249
Manny Medina: everywhere except for marketing. I was trying to figure out why that is. And then any thoughts on that.

119
00:15:12.180 –> 00:15:17.550
Allison Snow: You know, I I think there is a a level of conservative and market.

120
00:15:17.550 –> 00:15:18.730
Manny Medina: Oh, interesting!

121
00:15:18.730 –> 00:15:20.760
Allison Snow: And and it’s a level of

122
00:15:21.970 –> 00:15:24.949
Allison Snow: going to say conservative and and just being

123
00:15:26.200 –> 00:15:34.799
Allison Snow: more careful, maybe, sometimes than we need. So even when you were talking about promises. I know that as a marketer and so many marketers I’ve worked with.

124
00:15:35.360 –> 00:15:40.338
Allison Snow: you really have to make decisions all the time about the promises that you make right.

125
00:15:40.860 –> 00:15:45.849
Allison Snow: I confess I’ve said words like this isn’t a legal document. It’s a website

126
00:15:47.790 –> 00:16:00.040
Allison Snow: that on a recording manny there it is for the world to consume. I said it 10 years ago, but we do have to make those decisions every day right in terms of messaging in terms of what we shot from the rooftops and and all of those things.

127
00:16:00.150 –> 00:16:02.220
Allison Snow: And I think we

128
00:16:02.720 –> 00:16:19.349
Allison Snow: because we care about those renewals because we care about the integrity of the organization. We have adopted some conservatism. Now, this isn’t true for every company. Of course we’ve got folks out there making claims that are that are dubious, of course, but I think that when we’re thinking about tools.

129
00:16:19.350 –> 00:16:37.409
Allison Snow: my colleagues in in marketing, not at my firm. We’re moving pretty slowly, and we’re exploring. What does this look like for content, and even for content, as the baseline of this works pretty well, I’m still, by the way, owning it. So. So you just described what if a campaign was built and functional?

130
00:16:37.570 –> 00:16:40.659
Allison Snow: And I love that because I’m pressing a lot of buttons.

131
00:16:40.960 –> 00:16:41.380
Manny Medina: Right.

132
00:16:41.380 –> 00:16:51.809
Allison Snow: And I don’t want to press buttons. I want to have a point of view on audience and messaging and asset and Cta, and then I want it to self assemble, and that’s possible.

133
00:16:52.120 –> 00:17:06.549
Allison Snow: But my marketing brain is still going to, you know, until I’m till a few years, when I see it. And I think, Okay, this is good. It knows me all of these things that knows the preferences. It knows the audience. I’m still going through a lot of those things with a fine tooth comb and thinking.

134
00:17:06.550 –> 00:17:07.099
Manny Medina: Thing.

135
00:17:07.109 –> 00:17:09.019
Allison Snow: I know this puts me kind of behind.

136
00:17:09.149 –> 00:17:19.119
Allison Snow: you know, in terms of people who are aggressively pursuing these things. But I’ve just adopted that kind of conservative point of view, and I think, for reasons that have to do with messaging. I hope that’s helpful.

137
00:17:19.750 –> 00:17:22.199
Manny Medina: Interesting. Do you think it’s because of you being

138
00:17:22.520 –> 00:17:38.299
Manny Medina: exposed and and bombarded with so many solutions? Or you know, there’s 20 solutions for 5 problems in marketing. And and now you’re sort of like your level of your your barrier, for bullshit is significantly higher. So when somebody tells you your dream have come through. You’re merely like no.

139
00:17:38.540 –> 00:17:42.460
Manny Medina: somebody told me that yesterday, and it didn’t come true. So I’m not believing you.

140
00:17:42.460 –> 00:17:49.639
Allison Snow: Absolutely we. Yeah. I mean, I’m I’m in a I’m in an industry, you know. Events. Folks who buy event software

141
00:17:49.870 –> 00:17:52.339
Allison Snow: have been lied to for decades.

142
00:17:54.700 –> 00:17:56.149
Allison Snow: I think so. I.

143
00:17:56.150 –> 00:17:56.759
Manny Medina: Thank you.

144
00:17:56.760 –> 00:18:23.319
Allison Snow: I, you know, parentheses, I think I have to really be aware if the promises that we’re making sound really lofty and just sort of. Consider that in terms of what the audience has been, through what they say about all of my competitors on, you know. G. 2, and trust radius and Linkedin and and the slacks for Cmos that I’m on, that this is not an audience that’s ready to believe

145
00:18:23.580 –> 00:18:26.780
Allison Snow: they are an audience that has experienced frustration

146
00:18:26.980 –> 00:18:32.702
Allison Snow: at, you know, when we talk about events, particularly on site events. Not that virtual isn’t

147
00:18:33.410 –> 00:18:40.229
Allison Snow: you know, full of risk as well. But you’ve got that one chance. Right? You’re an event marketer. You’re an event professional.

148
00:18:40.650 –> 00:18:48.269
Manny Medina: One thing goes wrong that day, and it has ruined 6 to 8 months of your good work. Oh, wow! So those grudges!

149
00:18:48.270 –> 00:18:49.420
Manny Medina: Oh, for a long time.

150
00:18:49.920 –> 00:18:51.040
Manny Medina: Ha, ha! Ha! Ha!

151
00:18:51.040 –> 00:19:03.639
Allison Snow: They laughed. So you just have to kind of consider the audience, but yeah, just stay. I think everyone’s walking around with a I’ve been hurt before. Right? I smell the Bs when it’s coming.

152
00:19:03.640 –> 00:19:12.035
Manny Medina: Interesting, because what you know, interestingly enough, in digital marketing, I’m not seeing a lot which is very surprising that I’m not seeing a lot of agentic work. I’m seeing a lot in

153
00:19:12.440 –> 00:19:19.550
Manny Medina: in brand marketing. So, for instance, this was this fascinating agenda company that that’s sort of like

154
00:19:20.500 –> 00:19:25.539
Manny Medina: it. It use AI agents to ensure that the messaging

155
00:19:25.800 –> 00:19:31.689
Manny Medina: and the and the the campaigns and the brand was well represented for the geography which they were running

156
00:19:33.250 –> 00:19:38.680
Manny Medina: because some global brands just don’t have the the human power

157
00:19:38.790 –> 00:19:52.090
Manny Medina: to be hyper tailored to like every micro region in which they’re selling. And some of them made the the advertising made land wrong. So, for instance, they were helping Jack Daniels with a particular campaign, and when he run here in the Uk.

158
00:19:52.210 –> 00:19:55.840
Manny Medina: It turns out that the the actors in the campaign look underage.

159
00:19:56.020 –> 00:20:01.884
Manny Medina: and they caught it before it actually went to to market before it went to to to Media.

160
00:20:02.530 –> 00:20:05.259
Manny Medina: you know. Would a happy night disaster? Who knows but.

161
00:20:05.950 –> 00:20:06.560
Allison Snow: Sure.

162
00:20:06.560 –> 00:20:15.840
Manny Medina: Avoid it. And it was, you know, tons of headaches, you know that that ruining, you know, all the reputation. 6 months of work. That’s exactly what they avoided so.

163
00:20:15.840 –> 00:20:16.620
Manny Medina: and I think

164
00:20:16.620 –> 00:20:20.710
Manny Medina: feels like a better approach to coming at that from the safety side. You know what I mean. I gotta keep you safe, you know.

165
00:20:20.710 –> 00:20:28.820
Allison Snow: Yeah, no, appreciate that. And and look we are May. Maybe it’s because of that conservative mindset. But I’m I’m pretty jazzed about the idea of risk mitigation, you know.

166
00:20:30.400 –> 00:20:35.670
Allison Snow: but that is, that is jazzy for us, I think, as marketers. I hope folks agree.

167
00:20:37.030 –> 00:20:56.700
Allison Snow: it’s funny that the example with Jack Daniels. The reason I like it so much, and I just want to call this out to the audience. It takes something that that you know that you still as a marketer, I think, want to own in the here and now, which is your identity, who you are, what you want to tell an audience how you want to make them feel all of these things that we obsess about, and it

168
00:20:57.480 –> 00:21:04.299
Allison Snow: it it separates the stuff that you don’t know which is local preferences, right market preferences.

169
00:21:04.300 –> 00:21:04.700
Manny Medina: Right.

170
00:21:04.700 –> 00:21:07.990
Allison Snow: To something acting independently.

171
00:21:08.160 –> 00:21:13.170
Allison Snow: that that you can trust for those things, and that you’re not afraid to trust for those things

172
00:21:14.010 –> 00:21:29.029
Allison Snow: we never really, you know, unless you’re a localization expert, which is a whole story, you simply can’t be an expert in all of those things. So it’s no hit to what might be an ego. It’s no hit to anything else. It’s really just saying I need to blend everything that I know and worked really hard on

173
00:21:29.200 –> 00:21:35.449
Allison Snow: to to assemble and package and become this entity, that we are, and combine it with some other expertise

174
00:21:35.840 –> 00:21:43.890
Allison Snow: that simply doesn’t need to be human and and mitigate risk, and, you know, come out with a campaign that is compelling for the audience given.

175
00:21:43.890 –> 00:21:50.430
Manny Medina: Percent. But you just think about like. So the reason I found this fascinating is because it hits in so many levels. So 1st of all

176
00:21:50.870 –> 00:22:12.470
Manny Medina: risk when it comes to branding, is a little bit in the eye of the beholder. Right? So these people may have looked on the rage to the AI and to some people, but not to others. Sure, you know the people who put together a campaign, you know, they were just looking for young happening people that just, you know, once it was printed and shot, and whatever you know, it came out the wrong way, right? Maybe the clothes made it look that way. The makeup who knows?

177
00:22:13.016 –> 00:22:20.549
Manny Medina: It’s still sort of like a probabilistic answer. So there’s not definitive, yes or no, you know, probabilistically, the agent

178
00:22:20.660 –> 00:22:24.790
Manny Medina: told this consumer that there is a high chance that this will be flagged.

179
00:22:25.200 –> 00:22:26.070
Allison Snow: Yep. Yep.

180
00:22:26.070 –> 00:22:34.170
Manny Medina: But the problem is that that high chance has a massive, you know cost. If it happens, you see what I mean, because now you have

181
00:22:34.170 –> 00:22:35.290
Manny Medina: 100%.

182
00:22:35.290 –> 00:22:36.410
Manny Medina: It’s our app.

183
00:22:36.590 –> 00:22:42.320
Allison Snow: It’s not worth the risk. And and there’s again I mentioned people holding grudges before. I said it as a joke, and I meant it as a joke, but

184
00:22:43.550 –> 00:22:44.830
Allison Snow: memory is long.

185
00:22:45.160 –> 00:22:45.840
Manny Medina: A 100%.

186
00:22:45.840 –> 00:22:47.490
Allison Snow: Just mess up and erase it.

187
00:22:47.720 –> 00:22:55.219
Allison Snow: You mess up and and you’ve made an impression. That’s that’s not so positive. So I you know. I will tell you. We run around this this

188
00:22:55.940 –> 00:23:02.170
Allison Snow: virtual office. I’m in Boston and companies in Canada, and and we chat through. Look if there’s anything wrong here

189
00:23:02.280 –> 00:23:22.819
Allison Snow: we, the most important thing is we catch it as we look at each other. I’m not worried. If I’ve made a mistake in a campaign, I’m worried if it hit the market, and it went public. But you throw your tomatoes while we’re here, and that’s what that’s what it sounds like. That agent did so very cool. I’m going to drag you away from this real quick because you have a term.

190
00:23:23.040 –> 00:23:27.459
Allison Snow: and I think people want to learn about it. The term is agentic margin.

191
00:23:28.340 –> 00:23:29.000
Manny Medina: Yes.

192
00:23:29.300 –> 00:23:39.030
Allison Snow: And I could say what it means, because I cause I thought, Oh, I need to know what that means, and I checked it out, but it’s your it’s your term. Why don’t you? Can you tell us a little bit about what that means?

193
00:23:39.180 –> 00:23:42.210
Manny Medina: Yeah. So we found fascinating

194
00:23:42.960 –> 00:23:47.479
Manny Medina: one of the things that I found fascinating when we were building agents at Outreach. Was that

195
00:23:47.680 –> 00:23:56.249
Manny Medina: so? The the cost, the true cost of running an agent. If the agent is truly truly empowered to do its thing.

196
00:23:56.550 –> 00:24:05.649
Manny Medina: it could be undeterministic, and it could actually turn upside down. So, for instance, for you to write a really relevant email or to have a really relevant brief.

197
00:24:05.970 –> 00:24:15.719
Manny Medina: you go look at the historical relationship you have with that person you’re about to engage, and if you have, you know, a year worth of conversations or relationships, you know that’s a fairly straightforward thing.

198
00:24:15.920 –> 00:24:33.250
Manny Medina: But if you have 20 years of conversations and relationship with this person, and you’ve done several companies, and that person has been in several jobs. So you have that big body of history with that person, you know, ragging that whole thing and putting that into a prompt is going to be very, very expensive.

199
00:24:33.500 –> 00:24:39.420
Manny Medina: Then all of a sudden, an email or a brief is costing you instead of costing you 10 cents. It’s costing you $5,

200
00:24:39.600 –> 00:24:52.049
Manny Medina: and you can’t. And pre factor. You just can’t tell. Yes, of course you can put in safeguards so that you, you know you don’t do the body of the work, etc. But just putting safer on cost is not enough. Right? Because let’s say for a minute

201
00:24:52.150 –> 00:25:21.379
Manny Medina: that you’re saying, look, I’m going to engage with Allison because I have a renewal coming up, and I better be sharp on that renewal. So $5 for that email, of course, is worth it. You know what I mean. $5 for that prep is 100% worth it. You see, I mean, there is more context to the, to the problem that is naked to the eye. So the reason I’m saying all this is because the agentic margin is the cost of performing a body of labor that is only relevant to an agent meaning we’re all incur compute in a storage cost, but only the agent

202
00:25:21.550 –> 00:25:30.870
Manny Medina: has is privy to all the input and output tokens, and that agent can run the tab in any direction depending on the task at hand.

203
00:25:31.110 –> 00:25:35.209
Manny Medina: And if it’s truly agentic behavior, the agentic margin may turn upside down.

204
00:25:35.580 –> 00:25:53.000
Manny Medina: And that’s the part where you have to be very careful in that you have to price the task or the outcome of the workflow based on the complexity of the task based on the behavior of the agent, so that you can safeguard your identity. Your agentic margin comes above your gross margin.

205
00:25:53.130 –> 00:26:08.600
Manny Medina: You see, I mean, the agent does a bunch of stuff. Then you have to pay your aws bill, and then you have to pay your support and your success, and like your go to Market Bill, and that constitutes your whole of your margin. But your genetic margin gets paid first.st You know what I mean. Without those tokens there is no agent.

206
00:26:08.820 –> 00:26:16.429
Manny Medina: and that’s why it’s so important to to all of us to be mindful and to make sure that we accommodate in your in our pricing.

207
00:26:16.540 –> 00:26:17.889
Manny Medina: or at the eigentic margin.

208
00:26:19.100 –> 00:26:27.830
Allison Snow: Excellent. Thank you. I thought that was a really sort of provocative term to explore. I am stunned, truly, to see my clock says 30 min past the hour.

209
00:26:28.830 –> 00:26:29.979
Manny Medina: I know. How did that happen.

210
00:26:29.980 –> 00:26:32.629
Allison Snow: I’m sincere. I’m stunned, and I,

211
00:26:32.760 –> 00:26:38.449
Allison Snow: Julia, are you here to advise? If we have time for another question, or if we’ve got to cut it.

212
00:26:39.340 –> 00:26:56.990
Julia Nimchinski: Got a transition, unfortunately, to the next session. But thank you so so much, many and Allison many. What’s the best next step for our audience to learn more about pay pay.ai. I know you have a podcast listening to almost every episode which is amazing.

213
00:26:57.460 –> 00:26:59.049
Julia Nimchinski: So yeah, where should we go.

214
00:26:59.050 –> 00:27:15.779
Manny Medina: Oh, wow! That just made my day right there, Julia, like I have nothing else to say that was pretty amazing. So you know, follow us on Payai, on Linkedin or on Twitter. You can email me directly, Manny, at Payai. Many of you have already. My email is not hard to guess.

215
00:27:15.910 –> 00:27:27.400
Manny Medina: or or just drop us a line on Linkedin. We are. We are becoming prolific in the content we’re putting out there hopefully. That content is helpful to y’all. But yeah, we’re we’re trying to help everybody think about

216
00:27:27.580 –> 00:27:29.970
Manny Medina: pricing your agents. Make sure that you make you get paid.

217
00:27:32.200 –> 00:27:34.719
Julia Nimchinski: Awesome. Thank you so much. And how about you, Alison?

218
00:27:35.570 –> 00:28:00.039
Allison Snow: Cool visit [email protected] check out the blog. Please connect with me on Linkedin. I am always looking for anyone in the events industry, to interview, to feature in a blog post feature in a video. We just want to talk to you and get to know you. Knowing our customers are really kind of the number. One thing. So I’m going to appeal to this group of of sophisticated marketers and and ask you to engage me on that if you are.

219
00:28:00.190 –> 00:28:03.040
Allison Snow: if you have any time. Thank you, Manny, so much.

220
00:28:03.500 –> 00:28:04.210
Manny Medina: Thank you.