Latest / AI Top 5 / Smarter AI, Smaller Teams, Bigger Questions

Transcript

Smarter AI, Smaller Teams, Bigger Questions · AI Top 5

0:00
SPEAKER_00: This episode is brought to you by Elevate 2027.
0:04
SPEAKER_00: Three days, one blueprint to break through.
0:07
SPEAKER_00: Immerse yourself with the latest in retail and CPG.
0:10
SPEAKER_00: Build momentum with cutting-edge tech.
0:13
SPEAKER_00: Then activate those insights and turn them into results with practical, hands-on learning.
0:18
SPEAKER_00: Elevate 2027, the conference that earns the trip.
0:22
SPEAKER_00: Register now.
0:28
SPEAKER_03: Hi everyone.
0:29
SPEAKER_03: Welcome back to AI Top 5 with me and Dr.
0:32
SPEAKER_03: Eve, my partner in crime and the CAI.
0:38
SPEAKER_03: What's your official title?
0:40
SPEAKER_01: Uh yeah, Chief AI Officer at Xerces, but also President at XGS.
0:45
SPEAKER_03: Oh.
0:46
SPEAKER_03: This is fancy.
0:48
SPEAKER_01: Isn't it?
0:49
SPEAKER_03: Yeah.
0:50
SPEAKER_03: I'm really considering getting one of those AI certificates from um MIT, just so I can be like, I went to MIT too.
0:57
SPEAKER_03: Might be worth three grand.
0:58
SPEAKER_03: We'll talk off the show.
0:59
SPEAKER_03: I'm sure.
1:00
SPEAKER_02: I'm sure.
1:01
SPEAKER_03: Well, anyway, there's a lot of things happening, obviously, in the world of AI since we last spoke with you two weeks ago, um, which I feel like we say every time, right?
1:10
SPEAKER_03: It seems like we go to sleep and overnight things are changing.
1:14
SPEAKER_03: In fact, list this list that we talk about, Eve pulls together every night, right before he shares it.
1:20
SPEAKER_03: And then we usually get an update that morning.
1:22
SPEAKER_03: So things are moving so quickly.
1:24
SPEAKER_03: We know that you don't have time to stay on top of what's going on.
1:28
SPEAKER_03: So we do.
1:29
SPEAKER_03: So that being said, let's get kicked off with my favorite part of this program, which we're calling Unprompted, where I just shoot three random questions at Eve he has not seen yet, and he has to answer accordingly.
1:42
SPEAKER_03: Eve, are you ready?
1:44
SPEAKER_01: Well, yes, I'm ready.
1:45
SPEAKER_01: I am ready.
1:47
SPEAKER_03: So the first one was something that kept me up last night at 3:42 in the morning, rough estimate.
1:52
SPEAKER_03: Are we creating a generation of analysts who actually can't analyze because of AI?
2:00
SPEAKER_01: Whew, that is a very good question.
2:06
SPEAKER_01: Um well, look, I think uh what I would say is that I would challenge the premise of the question in the sense that um I think the uh what an analyst used to do even three years ago is very different from what an analyst is doing today or will be doing, call it three years from now.
2:28
SPEAKER_01: Um and the skill set is going to evolve in the sense that you will not need to know how to do a specific analysis in Excel that uh that has been commoditized, but the uh ability to uh ask the right question and troubleshoot the answer, and does it make sense?
2:48
SPEAKER_01: Like you know, do a bit of that sniff test um that will remain.
2:53
SPEAKER_01: And so I think what we're seeing is just a different kind of analysts that will be emerging if things go well, in the sense that it will be less about your technical skills, but more about your business understanding and the context you bring.
3:13
SPEAKER_03: Yep.
3:13
SPEAKER_01: Sad for the nerds.
3:15
SPEAKER_01: But yeah.
3:17
SPEAKER_03: That's an absolutely great point.
3:19
SPEAKER_03: And I think that feeds into my next unprompted question, which if every manufacturer has access to the same powerful AI models, and let's just say in a perfect world everybody's using them accordingly, right?
3:31
SPEAKER_03: So we've homogenized the playing field when it comes to category management.
3:36
SPEAKER_03: How does someone differentiate their category story when now everyone kind of has the same level of skill set?
3:44
SPEAKER_01: Ah.
3:46
SPEAKER_01: Very, very, very good question.
3:47
SPEAKER_01: So look, I think that the the reality is that the models give you the chance to basically weave any story that you want.
4:00
SPEAKER_01: Which means that you know, uh you can use the data, you can do some analyses, you can exclude this, exclude that, put a specific narrative around an analysis set.
4:13
SPEAKER_01: Um and I think that I think is still going to be differentiated in the sense that you know the model will say these are three things, three stories you could go with.
4:24
SPEAKER_01: This is what I would choose as a model, but then you as a user has to go in and then choose maybe you'll go with option B because the model did not know that X or Y, right?
4:34
SPEAKER_01: Um, so that's one thing that I think will still remain differentiated.
4:39
SPEAKER_01: Like, yes, you may have the same tools, but you know, it's not because you've got a say you everybody's using Excel that you end up with the same uh financial model, right?
4:48
SPEAKER_01: That's that's a good analogy.
4:50
SPEAKER_01: But I think even more interestingly, if you really think about it and you pull the string further, then you can almost go into this game theoretic approach where you say, well, okay, if I know that the others are using the models, how can I use the models knowing that they know that I know that they know that they use the model in a bit of that cycle?
5:12
SPEAKER_01: And how can I try to do something that will be more that would be less expected, right?
5:17
SPEAKER_01: To really stand out.
5:19
SPEAKER_01: And you know, and even stand up and maybe say something like, Well, I'm sure that my competitors are telling you that X and Y, because they're using basic models.
5:27
SPEAKER_01: I'm here to tell you that they're wrong, and here's why, and this is what my thesis is.
5:32
SPEAKER_01: Bam.
5:32
SPEAKER_01: You know.
5:33
SPEAKER_01: So this is where a bit of that game theory can come, you know.
5:37
SPEAKER_01: They know that I know that they know that they know that I know, blah, blah, blah, blah.
5:41
SPEAKER_03: That they know that I know that I know that they know.
5:43
SPEAKER_03: Right.
5:43
SPEAKER_03: Yeah, exactly.
5:44
SPEAKER_03: Speaking of they know, I know, I know, they know.
5:47
SPEAKER_03: Um, if I secretly replaced you with an AI-generated co-host and nobody noticed, would you be more offended that the audience couldn't tell that you speak like AI, or would you be more offended that AI got better reviews than you did?
6:06
SPEAKER_01: Oh boy.
6:07
SPEAKER_01: Um, well, first of all, I wouldn't be I would be impressed actually, first of all.
6:12
SPEAKER_01: Um, you know, I wouldn't be mad at being impressed.
6:15
SPEAKER_01: Um, you know, there's actually a number of companies and and authors and thinkers out there that are actually thinking about a world where that actually happens at scale where you know you as a human being cannot physically attend all the meetings you're supposed to be at.
6:32
unknown: Right.
6:32
SPEAKER_01: So what do you do?
6:33
SPEAKER_01: You send your avatar, and the avatar will speak in your voice, will look like you, speak like you, and then send you back the summary of what was discussed, and may even may be able to take decisions that would be the decision that you would make, right?
6:49
SPEAKER_01: And so we're actually not that far away from that.
6:52
SPEAKER_01: And I see it as a as a potentially you know very good force multiplier, right?
6:58
SPEAKER_01: And and interestingly, a lot of the people who think about those topics also then expand it to what about artists, right?
7:06
SPEAKER_01: What if all of a sudden you can just have whatever fancy popular artist today, Justin Bieber, whatever, uh send an avatar to you, and you can say, Hey Justin, please play this song for me, and but just replace every third word with cat.
7:22
SPEAKER_01: It may be able to do it, right?
7:24
SPEAKER_01: I don't know.
7:24
SPEAKER_01: Yeah.
7:25
SPEAKER_01: But so what does that mean to artistry?
7:27
SPEAKER_01: What does it mean for IP, right?
7:28
SPEAKER_01: Um, so there's there's a lot of very interesting topics that go with that question.
7:33
SPEAKER_01: We're not far away from that.
7:34
SPEAKER_01: We're not far away from that.
7:35
SPEAKER_01: And actually, I would be pleased if that were to happen.
7:39
SPEAKER_01: Yeah, I think I think it'd be a great experiment.
7:42
SPEAKER_03: Right.
7:42
SPEAKER_03: I mean, maybe this show will be hosted by AI Chelsea and AI Eve here very soon.
7:48
SPEAKER_01: I'd be a less a lot less anxious about your unprompted question if I had an AI able to answer for me.
7:55
SPEAKER_03: And I would be a lot less anxious about what my hair looks like on this podcast.
7:59
SPEAKER_03: So I mean, maybe that's where we're moving to.
8:02
SPEAKER_03: Who knows?
8:02
SPEAKER_03: Maybe next week, you guys, it'll be AI co-hosted.
8:06
SPEAKER_03: So that's right.
8:07
SPEAKER_01: That's right.
8:08
SPEAKER_03: I had a fourth one that cracked me up.
8:10
SPEAKER_01: Okay.
8:10
SPEAKER_03: So usually I thought it was only three.
8:12
SPEAKER_01: I thought you were only about two weeks.
8:17
SPEAKER_03: If AI can do the work of a McKinsey consultant in 30 seconds, who's gonna spend the other six weeks making the PowerPoint deck and charging the client$400,000?
8:28
SPEAKER_01: Wow.
8:29
SPEAKER_01: Wow.
8:29
SPEAKER_01: A lot of a lot of things to unpack.
8:31
SPEAKER_03: Um I think moreover, it's you know, there's a lot of question about consultancy and how AI is reshaping consultancy and how the big four are going to really respond to that.
8:44
SPEAKER_03: Or how have they already started to?
8:46
unknown: Yeah.
8:46
SPEAKER_01: I mean, look, first of all, you know, your 400k figure, I mean, that's probably not even two weeks of the team of four.
8:54
SPEAKER_01: So, you know, it's you're not getting a lot of bandwidth for that for that money.
8:59
SPEAKER_01: Um, you know, you gotta think bigger, you're gonna increase those are rookie numbers, right?
9:03
SPEAKER_01: Exactly.
9:03
SPEAKER_03: We need another zero on that for sure.
9:05
SPEAKER_01: Um But look, it is very true, though, that um service providers, people who generate knowledge or advice or advisory are will be seeing a big disruption, right?
9:20
SPEAKER_01: Because to a large extent, a lot of the work that, especially the, you know, the the the more commoditized consultancies or the more volume-based consultancies, uh, the topics that they serve can probably be answered largely by a model today, right?
9:39
SPEAKER_01: Um now where I think the what it what this means for the industry, I think, is that the top 0.1% of consultants are going to do extremely extremely well because they know how to use the tools and know how to do to get 100x what a normal human being could from the tools, and they can inject their expertise, their context, things that the model does not have yet, and really get you something highly differentiated.
10:06
SPEAKER_01: Or, you know, uh sometimes you just use a McKinsey or a Bain or a BCG because the board wants you to, and that's a bit of your you know, cover my butt sort of exercise, right?
10:16
SPEAKER_01: But it is true that otherwise, like for a lot of the you know, again, more community consultants, it the value at stake becomes uh more disputable.
10:28
SPEAKER_01: Now, that's my that's one way to frame that.
10:31
SPEAKER_01: The other frame, which is the opposite, could be yes, but a lot of companies out there today still don't really have the maturity to be able to get to that level, right?
10:42
SPEAKER_01: So you you still need will need consultants or advisors to come in and set up agents for you, build them for you, curate them for you.
10:50
SPEAKER_01: Um and you know, I st I I think that there is still a ton of work and revenue in the pipeline for for consultants, and but I think the fundamentally the sort of projects that is being taken on is going to change.
11:07
SPEAKER_01: Like McKinsey is not going to start doing, you know, the sort of mass market study that they maybe would be doing, or you know, uh it will be a lot more pointed and a lot a lot sharper.
11:21
SPEAKER_03: Absolutely.
11:23
SPEAKER_01: And yes, it will not spend, you know.
11:25
SPEAKER_01: I mean, back in my days, you would you you would do the you would do the work, do the analysis, and then send the pack to the um to the visual graphics team in India for like you know polishing.
11:35
SPEAKER_01: So you would have all the all of the all of the stuff on the slide and then hey, make it pretty, and it would come back and be very nice.
11:43
SPEAKER_01: Um now that as that's fully automated, you don't need to do this.
11:46
SPEAKER_01: I mean, even the entire slide deck can be mostly automated.
11:48
SPEAKER_01: So you know, it's a different uh it's a very different world that the consultants are in compared to when I was in.
11:56
SPEAKER_03: Yeah, it sounds like it.
11:58
SPEAKER_03: But the bottom line for Eve is still keep paying the consultants.
12:01
SPEAKER_03: So for now, for now, it's just for now.
12:04
SPEAKER_01: Yes.
12:05
SPEAKER_03: For now, exactly.
12:06
SPEAKER_03: And speaking of for now, let's get into the meat and cheese between the AI croissants.
12:11
SPEAKER_03: Did I say that right?
12:13
SPEAKER_01: Close enough, yes.
12:14
SPEAKER_03: Okay, it's close, right?
12:15
SPEAKER_03: I've been working on my French just for you.
12:17
SPEAKER_03: So maybe one day we'll do one sentence in French.
12:21
SPEAKER_03: Unlikely.
12:22
SPEAKER_01: Perfect.
12:22
SPEAKER_03: A little bit.
12:23
SPEAKER_03: So what really stood out to me this week were a couple of things.
12:27
SPEAKER_03: The first one, Amazon, right?
12:28
SPEAKER_03: Stop measuring AI usage and start measuring business outcome.
12:32
SPEAKER_03: So basically, what you shared with us is that Amazon is pushing back on using AI adoption metrics like token consumption, code generation as measures of productivity, effectively.
12:44
SPEAKER_03: What matters is the business value AI actually delivers.
12:47
SPEAKER_03: So, my question to you on this one is what metrics should cheap CPG organizations be tracking to determine whether AI is actually improving category management performance or hindering it?
12:59
SPEAKER_01: Oh, good question.
13:00
SPEAKER_01: So, look, first of all, let's go back in time, right?
13:03
SPEAKER_01: Um the context of this is that maybe like a year ago, a lot of companies in the valley were all about showcasing how cool they were, and your cool nest level was directly correlated to the how many tokens you would consume.
13:20
SPEAKER_01: Right.
13:20
SPEAKER_01: Um, there was a term called token maxing, and I think maxing is the word of the year, if you've not seen that.
13:26
SPEAKER_01: And so to that extent, managers at tech firms, like even the biggest tech firms, would actually look at people's AI consumption because it was the sense was you need to use AI to be more productive.
13:39
SPEAKER_01: And the only way I can track your usage of AI is simply by how many tokens you've burnt, right?
13:44
SPEAKER_01: Uh without thinking about whether the tokens were burnt effectively to answer the right questions and whatnot, right?
13:52
SPEAKER_01: And actually that went on for quite a while, right?
13:55
SPEAKER_01: People are being so happy to say, hey, look, I have I now have like 50 agents doing the like X and Y things for me.
14:03
SPEAKER_01: And then you know, every day I'm spending, you know, five thousand dollars in tokens.
14:08
SPEAKER_01: It's like great, but what's the real value?
14:10
SPEAKER_01: And I think just the fact that Amazon had to come out and say, hey guys, this is wrong, is kind of shocking because it is so obviously wrong that you know it's it's mind it's a bit mind-boggling to me that you know we had to go this far to come back where we are to this to the to sanity.
14:25
SPEAKER_01: Now, to come back to your question, what are the metrics that a CPG uh category manager should be looking at?
14:32
SPEAKER_01: Well, I mean, first of all, you know, it's how much time you spend doing manual stuff, right?
14:38
SPEAKER_01: And typically my my my very own rule of thumb is if I've got if I have to do something manually more than three times, it should be automated.
14:45
SPEAKER_02: Absolutely.
14:46
SPEAKER_01: It's my yeah, uh by the third time I have to cut and paste something, it just drives me insane.
14:51
SPEAKER_01: I will work on the way to build a tool to do it for me.
14:55
SPEAKER_01: Um and but then you've got to measure, you know, how much time does that really save you?
15:02
SPEAKER_01: Because you know, maybe if that if it's really literally three cut and paste that takes you like three seconds, then maybe it doesn't make sense for you to invest, you know, two days building the tool to replace that.
15:11
SPEAKER_01: Maybe not.
15:12
SPEAKER_01: Right?
15:13
SPEAKER_01: So I mean I think time saved is an obvious one.
15:16
SPEAKER_01: Um the the uh the uh speed at which you can react to new events, of course, is a is a uh is a very important one.
15:27
SPEAKER_01: Um the quality of the recommendations you provide.
15:31
SPEAKER_01: This one's a bit softer, it's hard to measure exactly.
15:34
SPEAKER_01: Um but at the end of the day, it's all gonna go back to like, you know, are we seeing an increase in revenue?
15:40
SPEAKER_01: Am I able to eat the market share of others?
15:43
SPEAKER_01: Right.
15:44
SPEAKER_01: Right.
15:44
SPEAKER_01: You know, that's what that's the whole game.
15:46
SPEAKER_01: And so I think everything has to be anchored there, first and foremost.
15:50
SPEAKER_01: Is are we because it even you know, even if you save time, great, but what do you do with the rest of the time that you've just saved, right?
15:56
SPEAKER_01: You've got to use it for something, and that something should be how do you make your product win?
16:00
SPEAKER_01: How do you make your brand win?
16:03
SPEAKER_03: Absolutely.
16:05
SPEAKER_03: And I think that you hit on something is you know, being in category management, I've I've been in a situation before where I'm doing something and like there has to be an easier way to do this, but then you just keep judging along, right?
16:17
SPEAKER_03: And the challenge is is to really hone in your AI skills, stop, ask AI, is there an easier way to do this?
16:24
SPEAKER_03: More than likely, there 100% is.
16:26
SPEAKER_03: And to Eve's point, if it's a three copy, three cut copy and paste, obviously that's not worth it, right?
16:34
SPEAKER_03: But if it is a situation where you're spending hours doing something, if you're cleaning up, for example, if your team really worked on a really cool tool that would help clean up kind of like, you know, the product masters that we work with that can contain a lot of attribution files, right?
16:50
SPEAKER_03: So from that perspective, you know, you've gotten that down to what minutes, I want to say.
16:56
SPEAKER_02: Yeah.
16:56
SPEAKER_03: You know, where it used to take us days.
16:58
SPEAKER_03: You know, every time there would be giant product launches, which you can kind of time throughout the year, you know that you're spending the next couple of weeks, you know, attributing files out.
17:06
SPEAKER_03: So it's things like that that really excite me because that's what I like to call soul crushing work when you when you're manually inning thing entering things or copying pasting, and you're like, this is how I'm spending my life force.
17:19
SPEAKER_03: That is the only finite resource I have in my life, right?
17:22
SPEAKER_03: So that's really where I think AI can help you and and really come into handy from fraying you up to get out of those tactical weeds.
17:30
SPEAKER_03: And companies like XGS are working on some really awesome tools to help you stream length some of those things.
17:36
SPEAKER_03: So really excited to see where that goes moving forward.
17:39
SPEAKER_03: And and another moving forward piece is what you mentioned about open AI, which really open AI feels like they're about to disrupt an org chart, is really what it feels like.
17:48
SPEAKER_03: So the story as you shared it was that open AI is exploring how AI agents can operate beneath, alongside, and between employees, potentially restructuring traditional organizational hierarchies.
18:00
SPEAKER_03: So, my question for you on this piece is if AI agents are doing the analytical heavy lifting, what becomes the most valuable skill that a category manager can bring to the table?
18:11
SPEAKER_01: Again, I I think I'll separate your question or my answer into two parts.
18:15
SPEAKER_01: Part one is gonna be about how agents, what is it, what does it mean for agents to start working into a business?
18:22
SPEAKER_01: And then I'll and I'll address the other the other one.
18:24
SPEAKER_01: So I think right now it's it's very common for us to see now that you know we have agents working for us.
18:31
SPEAKER_01: So I'm gonna send a prompt and I have like three, five, ten, hundred agents being spawned and then going out there doing things for me, then coming back with an answer.
18:40
SPEAKER_01: That's very straightforward.
18:42
SPEAKER_01: That's model number one.
18:43
SPEAKER_01: Model number two is to say, well, I am just going to completely automate one team, right?
18:48
SPEAKER_01: So it's not gonna be me interacting with one agent, it's gonna be my entire team interacting with one agent to get that one piece of information.
18:56
SPEAKER_01: The the people who used to do this, that doesn't exist.
18:59
SPEAKER_01: They it's been replaced, right?
19:01
SPEAKER_01: Um and then the last one, which I think is even more interesting, is when you start putting agents in between people, so between layers.
19:10
SPEAKER_01: So whereby if you are the head of a category, for example, you may want to know what's happening in the subcategories, right?
19:18
SPEAKER_01: But you may not want to have to go and ask, hey, what's up?
19:22
SPEAKER_01: You know, you may not want to have to read the transcripts of emails and calls and meetings that you don't have time to go to, and maybe that's where you should send your avatar to go back to the first question we had.
19:33
SPEAKER_01: Um, but what if you had uh an agent that could sit and basically read all of those current transcripts for you and then be a bit of your second brain, right?
19:47
SPEAKER_01: Your your spy to some extent, able to tell you, hey, look, this is what people talked about, this is where there may be a concern, this is what you this is what you should know and you should be aware of, and this is what you should pay attention to as a as a manager, right?
20:02
SPEAKER_01: Um not just a manager of people, not just a manager of agents, but a manager of part of a business.
20:08
SPEAKER_01: And I think that's a that's an interesting angle.
20:12
SPEAKER_01: Now to your other the other part of the question, uh, which was you know what skills uh are still required for for category managers?
20:26
SPEAKER_01: Well, look, you know, broadly I would say those are all you the skills you need to hone are the skills that are complementary to AI, right?
20:34
SPEAKER_01: So um that's basic econ 101, right?
20:38
SPEAKER_01: Um the value of something increases when it becomes scarce, right?
20:44
SPEAKER_01: So um or to the point the value of something becomes lower if it becomes commoditized.
20:51
SPEAKER_01: So, like again, doing a lot of the analysis, a lot of the math fully commoditized, there's no value there whatsoever.
20:57
SPEAKER_01: So, but the complement, what what do you need to pair up that analysis with to have a final end product and human judgment, the ability to create relationships, you know, the ability to pitch, connect with people, ask the right questions, right?
21:13
SPEAKER_01: As we talked about before, like validate the answers.
21:17
SPEAKER_01: You know, if an agent gives you 12 options, which one do you know should be implemented?
21:22
SPEAKER_01: Because you know the organization, because you know, hey, look, yes, those 12 options are all good, but we've tr we've tried all of those before and none of them work because guess what?
21:33
SPEAKER_01: This guy higher up in the organization hates those ideas, so it's just not gonna happen, right?
21:38
SPEAKER_01: So those are the sort of uh skills and information that you need to bring.
21:44
SPEAKER_01: And of course, you need to know how to use AI.
21:47
SPEAKER_01: Uh you just need to.
21:49
SPEAKER_01: It's being able to master the tools that you have at at your fingertips and then bring in what is uh that level of differentiation, that complement.
22:00
SPEAKER_01: Is I think what we all need to do.
22:02
SPEAKER_01: Otherwise, we we all we lose our jobs to AI.
22:04
SPEAKER_03: Right.
22:05
SPEAKER_03: And that's a great segue, speaking of losing our jobs to AI.
22:09
SPEAKER_03: So Microsoft, Microsoft effectively has been working on AI productivity and the future of their workforce.
22:17
SPEAKER_03: And the story that you shared was that this report really describes Microsoft offering voluntary exits to 60,000 employees.
22:27
SPEAKER_03: And from what I've seen, it's all over the media, right?
22:30
SPEAKER_03: Of some of the people who have gotten exit checks.
22:33
SPEAKER_03: They're they're being very generous.
22:34
SPEAKER_03: Let's just say that from some of the checks that are going out in early retirements and things like that.
22:40
SPEAKER_03: While we have, you know, teams like American Express who are also reportedly reducing workforces in mid-AI adoption.
22:48
SPEAKER_03: So the question to you is really probably what's on everybody's mind.
22:52
SPEAKER_03: And I feel like we address it almost every podcast that we do.
22:55
SPEAKER_03: But do you think that AI will ultimately reduce the number of category managers that a company needs, or will it simply raise expectations for what a category manager is?
23:07
SPEAKER_01: Hmm.
23:08
SPEAKER_01: Um but that's that's a very, very, very good question.
23:13
SPEAKER_01: The answer is I'm not sure where it's gonna go.
23:15
SPEAKER_01: Um because um we have seen the same thing in accounting to some extent, whereby you know yes, you can automate a lot of what accountants do.
23:27
SPEAKER_01: Cool.
23:28
SPEAKER_01: But it also means that they can do a lot more.
23:30
SPEAKER_01: Right?
23:30
SPEAKER_01: The same team of accountants can just cover a lot more, can keep track of a lot more things.
23:35
SPEAKER_01: And guess what?
23:36
SPEAKER_01: In a company that's AI enabled, where everybody, everybody is doing a lot more with what they have because they just have those tools at their fingertips, you just need to be able to keep track of all of those things, right?
23:48
SPEAKER_01: So you need to be able to keep track of your AI spending by type of model, by you know, by by capability, by by business, by person, whatever, whatever, right?
23:58
SPEAKER_01: Um and so at the end of the day, it's not entirely crazy to think that you may end up with having as many people doing the job, but the job just expands, right?
24:08
SPEAKER_01: You just do a lot more.
24:10
SPEAKER_01: Again, what you will do a lot more of though is not gonna be the manual stuff that we've talked about.
24:15
SPEAKER_01: It's gonna be a lot less about cutting and pasting nonsense into Excel 5 or creating PowerPoint.
24:21
SPEAKER_01: That is that is gone.
24:22
SPEAKER_01: And I think we should all be very happy that it's gone, right?
24:24
SPEAKER_01: That's all gonna be about you know, the hey, let's go meet clients, let's go talk to people, let's create connections, let's understand the customers, right?
24:34
SPEAKER_01: And I think that becomes uh you may end up having more people doing it because you can just do more of it.
24:42
SPEAKER_03: Right.
24:43
SPEAKER_03: Absolutely.
24:43
SPEAKER_03: And I think for me, I get asked this question a lot anytime I talk to someone in the industry, and I kind of answer it the same way you did.
24:50
SPEAKER_03: It's twofold for me.
24:52
SPEAKER_03: That number one, everything that we've hated about category management, right?
24:57
SPEAKER_03: We've always described it as an art in a science, and that has been a downfall.
25:01
SPEAKER_03: But in the world of AI, I feel like that's actually a superpower.
25:06
SPEAKER_03: AI can get the science.
25:07
SPEAKER_03: Science isn't difficult.
25:09
SPEAKER_03: It's the art piece, it's talking to people, it's the human behavior element, right?
25:15
SPEAKER_03: It's, you know, AI is getting very good at pattern recognition, but stringing some of those things, thoughts together as far as you know humans better than AI does right now, caveat right now.
25:25
SPEAKER_03: Uh, but you know, it's things like that where you can still use your brain, the art piece, and really add to that.
25:31
SPEAKER_03: The second piece that I'm kind of seeing, and I I think this is gonna be the way that the the world goes in CPG, is more kind of getting analytical people out of silos.
25:44
SPEAKER_03: So you're gonna see people come out of business insights, category management.
25:47
SPEAKER_03: And I'm assuming there's gonna be like a center of excellence that's just going to be business analytics and we'll cover category management, internal analytics, maybe financial, supply chain analytics, uh retail media.
25:59
SPEAKER_03: So there's a lot of different things that I think category managers are gonna be expanded to do.
26:05
SPEAKER_03: And I think it's gonna be more of an analytical bucket than siloed, to be a hundred percent honest, because really it's just going to be do you have the most analytical brain and how can we deploy you in different areas of the business?
26:16
SPEAKER_03: So that's kind of where my thought is.
26:18
SPEAKER_03: You're gonna be stretched to do more.
26:21
SPEAKER_03: Um, because to your point, you know, copying and pasting is out the door.
26:24
SPEAKER_03: So a lot of that tactical work is out.
26:26
SPEAKER_03: It's gonna be more straight, it's gonna be more strategy, is really what you're gonna be called to do.
26:30
SPEAKER_03: And if those are things that you're concerned about, once again, a little shameless plug, please get into um art campus and take some CMA courses, get certified in category management up to strategic advisor if you're qualified to do so.
26:43
SPEAKER_03: And there's some really awesome classes in there, including Eve's AI courses.
26:47
SPEAKER_03: He's working on a fourth one.
26:50
SPEAKER_01: Still working on it, yes.
26:51
SPEAKER_03: I think still working on the fourth one.
26:52
SPEAKER_03: So, and obviously, we're also putting out articles on how to use different fixtures of AI, including scheduling, and eventually we'll get into building your own bot.
27:01
SPEAKER_03: So, very much still follow what we're doing here at the CMA and ARC.
27:06
SPEAKER_03: Um, and you'll hopefully help elevate your AI skills with that.
27:11
SPEAKER_03: But speaking about elevating skills, let's talk about anthropic.
27:16
SPEAKER_03: And so hunger beats tenure in the age of AI, right?
27:20
SPEAKER_03: That's kind of what we're calling this.
27:22
SPEAKER_03: So uh anthropic's Felix Reisenberg, I'm hoping I did Risenberg or Reasy, yes, Mr.
27:29
SPEAKER_03: Felix argues that traditional technical interviews are becoming less relevant as AI performs tasks previously used to evaluate candidates.
27:40
SPEAKER_03: Adaptability, ambition, and demonstrated business impact may matter more than years of experience.
27:46
SPEAKER_03: So, in your thought, Eve, if technical analytical skills become increasingly accessible through AI, are we going to have to completely rethink how we hire and evaluate analytical minds in the CPG world?
28:02
SPEAKER_01: Yes.
28:03
SPEAKER_01: Next question.
28:05
SPEAKER_01: No, no, absolutely.
28:07
SPEAKER_01: Uh and by the way, this is this is also true for like education retarded, right?
28:12
SPEAKER_01: Like the way we we educate students today and evaluate students is going to change.
28:16
SPEAKER_01: It just has to change.
28:18
SPEAKER_01: Um But look, in in the world of um in the world of tech, it's you used to have to go through those grueling uh coding sessions, like either with a human being or with an interface to demonstrate that you knew how to code.
28:36
SPEAKER_01: That was what you were hired for, your technical skills.
28:39
SPEAKER_03: Yeah.
28:40
SPEAKER_01: That is gone.
28:41
SPEAKER_01: That is gone.
28:41
SPEAKER_01: It just makes no sense.
28:42
SPEAKER_01: It's completely I mean, it uh it's good to know that someone knows about the technology, but to a large extent it's irrelevant.
28:49
SPEAKER_01: What we really care about now a lot more is not so much do you know how to build something, it's a lot more about do you know what should be built.
28:59
SPEAKER_01: And what we're now seeing a lot more of is uh companies uh filtering for a very specific trait or a specific set of traits, which are not, you know, are you the smartest, most technically apt person coming out of MIT?
29:19
SPEAKER_01: Right.
29:20
SPEAKER_01: Or is it more like you are you're hungry, you're ambitious, you're curious, you're willing to learn, you've got the grit to actually go through and test things that may or may not work, right?
29:32
SPEAKER_01: Do you understand the business?
29:33
SPEAKER_01: Do you understand what will make m what will make the business successful?
29:37
SPEAKER_01: Right.
29:38
SPEAKER_01: Um and if you know that, then you'll be able to ask the right questions, and then you'll be able to know what to build, right?
29:44
SPEAKER_01: That's a bit of that that sequence.
29:46
SPEAKER_01: Definitely.
29:46
SPEAKER_01: Um and I think today that's those are the traits that people will be looking for more and more.
29:52
SPEAKER_01: And again, those are the traits that will become in higher and higher demand.
29:57
SPEAKER_01: And it turns out it's not not that many people, I believe, have all of those traits together.
30:04
SPEAKER_01: Right.
30:05
SPEAKER_01: And but it's never too late to to build those muscles, right?
30:10
SPEAKER_03: That's very true.
30:11
SPEAKER_03: And you bring up a philosophical question that sometimes I battle by myself in my mind.
30:16
SPEAKER_03: I need some hobbies.
30:18
SPEAKER_03: I really need some hobbies, Eve, but can you and actually I'm excited to ask you this one.
30:25
SPEAKER_03: Can you teach someone to be curious?
30:28
SPEAKER_03: Can you teach someone that intrinsic why?
30:32
SPEAKER_03: Or are you already just born with that or your brain is wired that way?
30:36
SPEAKER_03: I've got a really strong opinion on it, so I wanted to see what you think.
30:41
SPEAKER_01: That is a tough, that is a very tough question.
30:43
SPEAKER_03: Uh there's no easy questions, haven't you realized that yet?
30:47
SPEAKER_01: Right.
30:48
SPEAKER_01: Well, look, I think I think everybody's born with a range.
30:53
SPEAKER_01: Right.
30:53
SPEAKER_01: So you are everybody it's not that you're born with a specific amount of horsepower and that's all you get, right?
31:03
SPEAKER_01: It's right.
31:04
SPEAKER_01: I think from a curiosity standpoint, you probably, you know, some people just will just not have will just not be wired to ask questions in the way that you or I would ask that it's just a fact.
31:16
SPEAKER_01: Um and you can train them as much as you want, it's just not gonna happen, right?
31:20
SPEAKER_01: That would be one end of the spectrum.
31:22
SPEAKER_01: But I think for the vast majority of people, it's you know you're somewhere there's a range.
31:27
SPEAKER_01: And if you happen to be in a place where you just have not had a chance to really ask yourself those questions, then you would you should also ask yourself, why is it that I don't think that way?
31:38
SPEAKER_01: Is it because I don't have time?
31:39
SPEAKER_01: Is it because you know I just don't care?
31:44
SPEAKER_01: Is it just because I've never been asked to do to do this, right?
31:49
SPEAKER_01: And then maybe the the then then the consequence should be well should I should I change that?
31:56
SPEAKER_01: Should I make more time for myself?
31:57
SPEAKER_01: Well guess what?
31:57
SPEAKER_01: I'm using AI, I've just fed up, you know, 12 hours of my 12 hour 12 hours in my week.
32:03
SPEAKER_01: What am I gonna do with it?
32:04
SPEAKER_01: Well, maybe I should ask myself those those questions.
32:07
SPEAKER_01: Um No one's been asking me this.
32:10
SPEAKER_01: Well, I'm gonna do it anyway because it's interesting, because I'm interested.
32:13
SPEAKER_01: Just you know, take ownership, right?
32:15
SPEAKER_01: I'm I'm an optimist by nature, but I'm also I'm also very aware that you know not everybody will be able to do it, right?
32:22
SPEAKER_01: And it's just a fact.
32:23
SPEAKER_03: Yeah.
32:24
SPEAKER_03: And I think in the age of AI, know thyself is more important than ever.
32:31
SPEAKER_03: It's time to look at, take a look in the mirror and take a ruthless inventory of your skills, what you're good at and what you're not good at.
32:38
SPEAKER_03: And if you're weak or you have a material weakness in something that really needs to be stronger because of the age of AI, that's something that you're really gonna have to focus from a perspective of personal growth.
32:52
SPEAKER_03: Or you've regular you kind of say, okay, this is where I'm strongest.
32:56
SPEAKER_03: These are my skill sets, and I need to figure out how to exploit those in the world of AI if I don't feel like I can be analytical or ask that those analytical questions, right?
33:07
SPEAKER_03: So it's really interesting because we talk a lot about business perspective.
33:12
SPEAKER_03: Like, okay, if this goes away, what can I do from a business perspective?
33:15
SPEAKER_03: But then there's the personal elevation that AI is also requiring people to do.
33:21
SPEAKER_03: And that is taking that ruthless moral inventory of what you are, what you can do, what you can't do, and figuring out if that's worth building on.
33:30
SPEAKER_03: So I think we don't talk about that enough.
33:33
SPEAKER_03: Um, and I love that we kind of got into that.
33:35
unknown: Yeah.
33:36
SPEAKER_01: And look, I think the reality is the world is changing.
33:39
SPEAKER_01: So people have to change.
33:40
SPEAKER_01: And change is difficult, right?
33:42
SPEAKER_01: So no, no one is saying that this is gonna be a completely painless and easy process.
33:48
SPEAKER_01: Right.
33:48
SPEAKER_01: It's not that's just not the fact that, but you know, evolve or die.
33:54
SPEAKER_03: Evolve or die for sure.
33:55
SPEAKER_01: That's that's the reality.
33:57
SPEAKER_03: Yep, exactly.
33:59
SPEAKER_03: So, with that, our last question that we're talking about, or our last little news piece here, is that AI companies are shifting from selling software to selling outcomes.
34:09
SPEAKER_03: Very, very interesting.
34:10
SPEAKER_03: So, you shared that AI startups are moving beyond selling software licenses towards delivering completed services and business outcomes, which will potentially disrupt traditional technology and consulting models, kind of what we talked about a little bit with the$400,000 McKinsey bargain basement project that we spoke out earlier.
34:30
SPEAKER_03: That's right.
34:30
SPEAKER_03: So, my question to you is could we eventually see category management evolve into an insights as a service model, which would be really interesting, where manufacturers effectively purchase finished recommendations instead of purchasing platforms to create them?
34:46
SPEAKER_01: Ah, that's a very interesting question.
34:48
SPEAKER_01: Um Well, I think it's a it's a possible, absolutely a possible outcome, right?
34:54
SPEAKER_01: Um I think that so if you look at software, pure like the typical SaaS product, right?
35:01
SPEAKER_01: Um what the industry is seeing is that you know, it's great that you buy a piece of software, it's great that you buy a SaaS, it's great that you buy Salesforce or SAP or whatnot.
35:13
SPEAKER_01: Fantastic.
35:14
SPEAKER_01: But no one, no one, no one has built or implemented Salesforce or SAP by themselves.
35:19
SPEAKER_01: There's always a level of service to customize it for you.
35:23
SPEAKER_01: And sometimes, guess what?
35:24
SPEAKER_01: You want to change your CRM or your your ERP.
35:28
SPEAKER_01: Guess what?
35:28
SPEAKER_01: You can't do it yourself, you've got to call those guys, right?
35:31
SPEAKER_01: Um historically, a lot of startups would only care about one thing and one thing only, and it was recurring revenue.
35:38
SPEAKER_01: That's what VCs would look at, right?
35:41
SPEAKER_01: And they would say, if you've got service revenue, it's good, but I won't even look at that to create a valuation for your company.
35:48
SPEAKER_01: Right.
35:48
SPEAKER_01: Yeah.
35:48
SPEAKER_01: And I think right now what people are seeing is that a creating software has become super easy.
35:55
SPEAKER_01: You can just build your own thing, you know, super quickly.
35:58
SPEAKER_01: So then the real question becomes now, well, yeah, uh maybe I can build my own thing, but I just don't want to.
36:04
SPEAKER_01: I want someone to do it for me, right?
36:05
SPEAKER_01: I I don't want to cook today, I want to go to a restaurant, or I want to do takeout, or I want to have a chef come to my house and cook for me, right?
36:15
SPEAKER_01: The way I think about it is you've got a couple of different levels.
36:17
SPEAKER_01: You've got like the I want to do it myself, so I want to go and do my own dinner tonight, right?
36:24
SPEAKER_01: There's a do-it with me, so well, I'm gonna have a chef come to me and they'll the prep the ingredients for me, then I'll do the the final product, perhaps.
36:32
SPEAKER_01: Or maybe they teach me how to do something, right?
36:35
SPEAKER_01: And then they're the do-it-for-me.
36:37
SPEAKER_01: I just go to the restaurant.
36:38
SPEAKER_01: Or the chef comes to my house and he cooks for me and then sh disappears, right?
36:42
SPEAKER_01: Wouldn't that be nice?
36:43
SPEAKER_03: Uh so nice.
36:44
SPEAKER_03: I was gonna say, do you have like a in-house chef?
36:47
SPEAKER_03: Like it's me.
36:49
SPEAKER_01: I'm the I'm the in-house chef.
36:52
SPEAKER_01: I would say that.
36:53
SPEAKER_01: But so no, from a from a software perspective, the same thing applies, right?
36:58
SPEAKER_01: Like if you think about you know accounting, category management, well, yeah, maybe you could do it yourself, but maybe you know, you just want to buy the end product.
37:08
SPEAKER_01: And what it actually means is to a large extent, if you go to that model, then you buy less, it's less about the number of hours that you work on a project, but it's a lot more about what you the value of what you created, right?
37:20
SPEAKER_01: And to that extent, it means that it's you know, it's you are then incentivized to use AI to become more efficient.
37:28
SPEAKER_01: Yeah.
37:28
SPEAKER_01: Because then you can create more value faster.
37:30
SPEAKER_03: Absolutely.
37:32
SPEAKER_03: And I think you hit on something that's really interesting that I feel like is a theme this week, which is that AI isn't just changing how category management gets done.
37:41
SPEAKER_03: It's really challenging us to understand what category management expertise is actually worth.
37:47
SPEAKER_03: You know, it it everything has a price, in the words of million-dollar man, um, which I don't think you were a WWF fan back in the day.
37:54
SPEAKER_03: That's another pop culture thing I'm gonna have to send you.
37:57
SPEAKER_03: You're gonna have to learn about wrestling back in the day.
37:59
SPEAKER_03: It's a lot of life lessons there, including taking your vitamins.
38:03
SPEAKER_03: But, you know, it's we used to kind of rest our laurels on the fact that we had this knowledge and this expertise.
38:11
SPEAKER_03: And AI is quickly coming, you know, knocking on the door of being able to replicate a lot of that.
38:16
SPEAKER_03: And how do we elevate our skills?
38:18
SPEAKER_03: And I think that actually rolls into our letter from the audience.
38:22
SPEAKER_03: So we got a letter that I was super excited about, and it's actually not about is AI going to take our jobs?
38:29
SPEAKER_03: Um, which is surprising because I feel like that's been all of our letters.
38:32
SPEAKER_03: But dear Chelsea and Eve, hi.
38:35
SPEAKER_03: I work in category management for a large CPG manufacturer that does home care.
38:40
SPEAKER_03: Okay.
38:40
SPEAKER_03: And our team has started implementing with AI to develop category insights and recommendations.
38:47
SPEAKER_03: My concern is that our retail partners are also adopting AI.
38:51
SPEAKER_03: And I assume that they're going to have access to the many of the same analytical capabilities that we do.
38:56
SPEAKER_03: If retailers can generate their own assortment recommendations, promotional analyses, and category growth strategies using AI, what value will manufacturers bring to the table?
39:07
SPEAKER_03: Ooh, good.
39:08
SPEAKER_03: And are we going to originally, or I'm sorry, are we going to reach a point where retailers no longer need manufacturers to provide category management support?
39:19
SPEAKER_01: That's a tough question.
39:22
SPEAKER_01: I think the premise of the question is 100% correct, right?
39:25
SPEAKER_03: Yeah, definitely.
39:27
SPEAKER_01: If everybody has access to this, has access to the same data and the same tools, or largely the same tools, then the recommendations will end up being the same.
39:37
SPEAKER_01: And you have zero differentiation, right?
39:40
SPEAKER_01: So if you are a category manager, what do you do?
39:43
SPEAKER_01: Well, first of all, you cannot, it's not, it's not a battle you can win by you know just by being smarter.
39:49
SPEAKER_01: So what you really need to do is uh figure out what is it that uh the m the the retailers cannot do.
39:59
SPEAKER_01: So what is the data that you as a manufacturer has access to that no one else has access to?
40:05
SPEAKER_01: Or and you know, one possible way to look at it is well, as a manufacturer I know the category better than anyone else.
40:12
SPEAKER_01: Definitely, and I know that what the customers want, and therefore that's my differentiation, right?
40:17
SPEAKER_01: And that is not a set of data that you can buy it because it is data that I have created and acquired through 20 years of trial and error, right?
40:27
SPEAKER_01: As an example.
40:28
SPEAKER_01: Definitely.
40:28
SPEAKER_01: Um and look, it is a it's a it's a valid, it's a very valid point.
40:36
SPEAKER_03: Absolutely.
40:37
SPEAKER_03: I mean, the days of jump balls are gone where you'd be able to pull one by jump balls.
40:42
SPEAKER_03: So we would call like jump balls if, for example, if you were putting together a planogram and there was some extra space that you could give an extra facing to a product that might not necessarily deserve it, but it's one of your items.
40:54
SPEAKER_03: Taking a jump ball because of a relationship that you have from a retailer.
40:58
SPEAKER_03: You should not be doing that, but it happens all the time.
41:00
SPEAKER_03: But the days of jump balls are gone because you kind of banked on the fact that the retailer had so much going on, which is true, former retailer, 2,500 emails a day, physically impossible to get through all of that, right?
41:12
SPEAKER_03: You would take advantage of the fact that they didn't have the time to dig into this information, right?
41:18
SPEAKER_03: And that's where a little bit of the spin would come.
41:20
SPEAKER_03: Well, now they do.
41:21
SPEAKER_03: That's at the tip of their fingertips.
41:24
SPEAKER_03: Um, you know, as a category captain, you would commonly provide a scoreboard or, you know, a daily flash of what's going on in a category.
41:33
SPEAKER_03: Once again, click of a button, they can schedule something in AI and get that themselves.
41:37
SPEAKER_03: So um, it's not that we're irrelevant.
41:40
SPEAKER_03: It's just that you're going to have to effectively up your game.
41:43
SPEAKER_03: And to your point, there are some really interesting data partners out there that can help bring some really interesting views, whether it be commodity, weather reporting, a lot of social listening, right?
41:55
SPEAKER_03: We've got a lot of social listening partners with the CMA that really brings some cool insight because more people are now going to AI for product recommendations than they are Google or word of mouth, which is traditional.
42:08
SPEAKER_03: So there's a lot of cool things going on that people are tracking.
42:12
SPEAKER_03: And that's really where you're going to shine is bringing those interesting takes and those nuggets that a traditional retailer is not going to be able to get, you know, just from using an AI platform.
42:22
SPEAKER_03: So to Eve's point, your first assessment should really figure out, okay, what can AI do in a flash, right?
42:29
SPEAKER_03: Because we know that the retailer is going to be able to do that for themselves.
42:33
SPEAKER_03: What requires a little bit more in-depth knowledge, in-depth skill that they might not be able to really, you know, conquer that we can bring to our manufacturer or to our retailer to still be, you know, an effective thought partner.
42:48
SPEAKER_03: So I think you were really right about that, Eve.
42:52
SPEAKER_03: To be a hundred percent honest.
42:53
SPEAKER_03: And that is the first and last time I will say that on this podcast.
42:57
SPEAKER_01: Wonderful.
42:58
SPEAKER_01: Awesome.
42:59
SPEAKER_03: Well, Eve, do you have anything else you want to add about this week or looking forward to the next couple of weeks before we meet again?
43:06
SPEAKER_01: Well, look, I think maybe just from a from an AI market perspective, we are obviously at a very interesting time, right at the verge of anthropic going public.
43:15
SPEAKER_01: And I think we are going to see what's going to happen there.
43:20
SPEAKER_01: Like we're we're going to see on the one hand the the emergence of a new class of billionaires, right?
43:27
SPEAKER_01: Um and on the flip side, we're going to see whether anthropic is indeed sustainable at that valuation, and if they are able to maintain the growth that underpins that valuation.
43:39
SPEAKER_01: And whether they can only underpin that growth by you know turning the screws on their customers by increasing prices or lowering what quota they're using and whatnot.
43:52
SPEAKER_01: So this is going to get really interesting.
43:54
SPEAKER_03: It really is.
43:55
SPEAKER_03: I mean, if you just kind of think about what we've talked about today, right?
43:58
SPEAKER_03: We've talked about avatars and IPOs and upping skills and you know, deep philosophical questions about the intrinsic why.
44:07
SPEAKER_03: It's really interesting because I keep seeing articles, you know, AI is going to force us, your, you know, kind of force us to think more strategically, or it's going to take away our knowledge because we don't have to think as much.
44:22
SPEAKER_03: But honestly, since AI has been launching these things, I feel like that's all people are doing is thinking, which is awesome.
44:28
SPEAKER_03: So keep that up.
44:29
SPEAKER_03: We'll be back in two weeks with more exciting AI news bits and how they apply to you.
44:35
SPEAKER_03: I'm Chelsea.
44:36
SPEAKER_03: This is Eve, and we've been super excited to spend the last hour with you.
44:41
SPEAKER_03: Once again, if you have any questions, please write them in.
44:44
SPEAKER_03: We love answering them on our podcast, and we will catch you all on the flip side.
44:48
SPEAKER_03: Thank you.
44:49
SPEAKER_01: Thank you.