Bản chép lời đầy đủ của From Data to Decision: Unpacking AI's Impact on Procurement with Pierre Mitchell (Executive Edge)

00:04:06.320 --> 00:04:08.780 certainly a great one. That's a good place to start. 00:04:09.060 --> 00:04:12.640 There's certainly a lot of impact, you know, on just good old, 00:04:12.740 --> 00:04:15.560 like I was saying before, good old fashioned spend analytics, 00:04:15.880 --> 00:04:20.820 predictive analytics, you know, around, you know, on the forecasting side, supply chain risk, 00:04:21.300 --> 00:04:25.260 anything that's like a big data of problem, you know, as well as anything that 00:04:25.260 --> 00:04:28.440 has a mass number of stakeholders, either on kind of your, 00:04:28.981 --> 00:04:32.041 Business stakeholder side, which is like that area of intake management, 00:04:32.341 --> 00:04:37.081 let's get a common choke point and form for everybody to engage with procurement. 00:04:37.261 --> 00:04:40.801 So it's kind of like the CRM of procurement. How do we support our stakeholders? 00:04:41.081 --> 00:04:47.681 Well, that piece requires can use some AI around policies and guided buying. 00:04:47.881 --> 00:04:51.521 And guided buying is a term I actually invented back in, you know, coined back in 2005. 00:04:52.101 --> 00:04:54.801 And now we're actually starting to have the technology to do it versus, 00:04:55.041 --> 00:04:57.501 you know, smart forms are really dumb forms. 00:04:57.501 --> 00:05:01.741 You know, these kind of, you know, kind of catalogs anyway, but so, 00:05:01.841 --> 00:05:04.401 so I think anything that's kind of like analytically focused, 00:05:04.641 --> 00:05:09.341 anything involving contracts, anything involving risk, which isn't the easiest 00:05:09.341 --> 00:05:12.341 ROI thing, but it's certainly strategic, right. 00:05:12.461 --> 00:05:18.301 As well as just anything that touches mass stakeholders where you can, you know, 00:05:18.741 --> 00:05:22.661 automate the, a lot of the self-service interaction, either with the stakeholders 00:05:22.661 --> 00:05:26.541 on the intake side or with suppliers on the supplier portal side, 00:05:26.561 --> 00:05:31.781 because that's a, you know, the shantytown portals that we erect for the poor suppliers, 00:05:32.021 --> 00:05:35.161 whether it's through a network or, you know, a la carte, you know, 00:05:35.261 --> 00:05:37.341 it's still a bit of a disaster mess right now, right? 00:05:37.421 --> 00:05:40.981 So the more in which we can, you know, make that more, you know, 00:05:41.001 --> 00:05:45.181 have agents help and help the suppliers do more of that, get more self-service. 00:05:45.621 --> 00:05:46.981 That's a good thing right now. 00:05:47.741 --> 00:05:50.581 Yeah, thanks. So there's suppliers that they're applauding you for that because 00:05:50.581 --> 00:05:55.381 they're often overlooked in any technology. I understand where you're going there. 00:05:55.941 --> 00:06:01.001 So it sounds like, okay, so I hear data retrieval, contract summarizations, 00:06:01.681 --> 00:06:04.401 maybe helping to draft RFQs, right? 00:06:04.501 --> 00:06:09.221 So there is, there are, I should say, there are many valid use cases today. 00:06:09.401 --> 00:06:13.821 A lot of it is around efficiency and time savings, but also making sure that 00:06:13.821 --> 00:06:15.801 the policies are enforced and whatnot. 00:06:15.801 --> 00:06:20.081 So it's happening there. Yeah, and statement of work generating, 00:06:20.161 --> 00:06:23.941 you know, anything that's kind of rather than me having to do it myself and 00:06:23.941 --> 00:06:26.561 either kind of reinvent the wheel or find the closest template. 00:06:27.235 --> 00:06:30.755 I can kind of generate it, but then I really, you know, there needs to be that 00:06:30.755 --> 00:06:33.455 human in the loop and whether you're training it or whether you're just kind 00:06:33.455 --> 00:06:37.875 of having it be a starting point, you know, there's nothing wrong with that 00:06:37.875 --> 00:06:39.735 because, you know, I'll just tell you. Yeah. 00:06:39.875 --> 00:06:42.315 So I think that's a lot of that generative stuff. 00:06:42.535 --> 00:06:47.495 And the agents is kind of just a way to do that more at scale, you know, in workflows. 00:06:47.915 --> 00:06:54.215 But yeah, that's it certainly helps productivity around on sort with sourcing folks right in there. 00:06:54.315 --> 00:06:58.435 It's not just, oh, it makes us more efficient and we just need less bodies. 00:06:58.915 --> 00:07:03.055 But it's like, oh, no, that means you get to spend more time driving deals. 00:07:03.235 --> 00:07:07.635 It means you are, you know, your bogey for value creation is going to be higher. 00:07:07.635 --> 00:07:09.415 That means you're going to touch more spend. 00:07:09.635 --> 00:07:12.555 You're going to touch it more deeply. You're going to go after new value levers. 00:07:12.775 --> 00:07:15.815 You're going to, you know, be able to deliver more with your stakeholders. 00:07:16.415 --> 00:07:20.515 It becomes, as I always say, kind of efficiency funds effectiveness, right? 00:07:20.715 --> 00:07:24.355 Because the biggest cost of procurement is the opportunity cost of you wasting 00:07:24.355 --> 00:07:28.915 your time on non-value added, you know, crap on the transactional side when 00:07:28.915 --> 00:07:32.055 you really need to be spending your time on driving more, you know, value. 00:07:32.235 --> 00:07:35.755 So efficiency drives effectiveness to the tune of, you know, 00:07:35.915 --> 00:07:37.995 whatever that procurement ROI metric is, right? 00:07:38.255 --> 00:07:42.555 5X, 10X, 20X, depends on where you are, how good you are, right? 00:07:42.675 --> 00:07:48.975 But that's a pretty good ROI calculation. And if you can't come up with 5X, 00:07:49.215 --> 00:07:50.555 you're doing something wrong. 00:07:50.735 --> 00:07:54.855 And if there's a CFO that doesn't think that 5X is better than their cost of 00:07:54.855 --> 00:07:58.675 capital, then either they don't believe procurement, which is a bad thing, 00:07:58.815 --> 00:08:01.735 or

it's a pretty good place to invest some money, right? 00:08:01.895 --> 00:08:05.775 Look at Goldrat and Theory of Constraints and where you invest your money, 00:08:05.875 --> 00:08:08.675 right? That's a good place to generate profit. 00:08:09.295 --> 00:08:15.815 For sure. Yeah, great. So, yeah, I love the efficiency leads to effectiveness, right? 00:08:16.075 --> 00:08:20.095 And because what are you doing? You're taking away the tasks that really don't 00:08:20.095 --> 00:08:21.915 add a lot of value. They need to be done. 00:08:22.095 --> 00:08:26.355 But you can focus your time and brainpower on more strategic activities. 00:08:26.555 --> 00:08:27.855 I'm going to give a shout out to my... 00:08:28.424 --> 00:08:33.384 Good friend, Omer Abdullah, who liked to say it's a combination of HI, 00:08:33.704 --> 00:08:35.304 human intelligence, plus AI. 00:08:35.664 --> 00:08:38.684 That's really how you need to look at this, right? It's like any tool, right? 00:08:38.844 --> 00:08:42.524 You need somebody that can write the prompts, somebody that can manage it, 00:08:42.624 --> 00:08:45.764 be that human in the loop, as you mentioned. So very important for that. 00:08:46.184 --> 00:08:51.264 So I know that whenever technology comes along, we tend to try to ram it into 00:08:51.264 --> 00:08:53.444 existing processes that we have. 00:08:53.604 --> 00:08:57.564 And sometimes we're our own worst enemy. I mean, we don't really open our minds 00:08:57.564 --> 00:09:01.244 to the full possibilities of technology and what that could bring. 00:09:01.404 --> 00:09:06.684 So when you look out there at procurement teams, are they ready from an organizational 00:09:06.684 --> 00:09:09.744 point of view and also from a technical point of view, maybe. 00:09:10.404 --> 00:09:12.384 To leverage AI effectively? 00:09:12.684 --> 00:09:17.344 Or, you know, what's missing most often when you're out there talking with clients 00:09:17.344 --> 00:09:18.824 about their strategies? 00:09:19.084 --> 00:09:24.884 Yeah, for sure. I mean, there's just a constant theme of doing more with less. 00:09:24.884 --> 00:09:32.684 And that means, you know, being able to drive more value, not just on the savings side, right, 00:09:32.844 --> 00:09:36.944 but also on, you know, all the other things that you can do in terms of, 00:09:37.004 --> 00:09:40.864 you know, risk reduction and top line benefits and, you know, 00:09:40.944 --> 00:09:45.464 things like that, as well as obviously, you know, driving more, you know, 00:09:45.684 --> 00:09:49.324 margin impact and cost impact and, you know, cash, you know. 00:09:49.664 --> 00:09:50.404 Impact and things like that. 00:09:50.404 --> 00:09:57.764 So the challenge is that our business environment is getting ever more complicated, 00:09:57.804 --> 00:10:04.104 especially with tariffs, tariff du jour, the pandemic, just all these things that are happening. 00:10:04.284 --> 00:10:09.304 There's just a lot of – it's a VUCA world that's out

there to coin military 00:10:09.304 --> 00:10:12.764 term, the variability, uncertainty, complexity, and ambiguity. 00:10:12.764 --> 00:10:15.824 And so it's like we're constantly reacting and 00:10:15.824 --> 00:10:18.564 we're also reacting to constant stakeholder shifts that's going 00:10:18.564 --> 00:10:21.364 on in the business units or whatever so just trying to keep up 00:10:21.364 --> 00:10:24.104 with that while trying to be 00:10:24.104 --> 00:10:28.844 more proactive to be involved earlier to bring new value to the table new analytics 00:10:28.844 --> 00:10:34.184 now bringing ai so when now this you know the new technology like ai and gen 00:10:34.184 --> 00:10:38.824 ai now genic ai comes on to the onto the scene it's like well is this thing 00:10:38.824 --> 00:10:41.744 a net add or you know or a net subtract. 00:10:42.144 --> 00:10:47.664 And it's so easy to just ignore it and be like, yeah, I'll get to it and learn more about it. 00:10:47.784 --> 00:10:51.164 But it's like that, to quote the old Rush song, you know, and the quote of, 00:10:51.244 --> 00:10:53.904 you know, if you choose not to decide, you still have made a choice. 00:10:54.264 --> 00:10:59.044 So if you are not, you know, sharpening your saw and really understanding just 00:10:59.044 --> 00:11:05.144 what the LLM chatbots can do, you're going to be at risk on how to prompt them 00:11:05.144 --> 00:11:07.764 and how to use them and just make yourself personally more, 00:11:08.364 --> 00:11:13.024 effective and productive and just to understand what the technology does and 00:11:13.024 --> 00:11:15.864 then learn how to build your own agents and just really start to get, 00:11:16.316 --> 00:11:20.116 into the technology a bit more, you're going to fall behind. 00:11:20.296 --> 00:11:24.196 And this is not a space where you want to kind of fall behind. And it's also not a fad. 00:11:24.356 --> 00:11:28.936 I mean, it is, this is not like, oh, blockchain, great, you know, 00:11:29.116 --> 00:11:33.696 kind of cool transaction registries and, you know, provenance, great. 00:11:33.856 --> 00:11:36.336 You know, that's, that's terrific, but not strategic. 00:11:36.516 --> 00:11:39.356 Well, I think I take away from this, it's not gone away. 00:11:39.756 --> 00:11:43.976 I think about this, like the internet, right? I think I told you about a week 00:11:43.976 --> 00:11:46.336 ago that since the internet, 00:11:46.536 --> 00:11:51.156 this is the most exciting thing I've seen in terms of technology developments 00:11:51.156 --> 00:11:55.256 and the impact I believe it will have on business and people in general. 00:11:56.676 --> 00:12:01.036 I think you're saying here, hey, just start using it, get used to it, 00:12:01.156 --> 00:12:04.236 understand what it delivers. You'll get better at prompt writing. 00:12:04.396 --> 00:12:08.576 You'll figure out how to start applying it to your job because it's not going 00:12:08.576 --> 00:12:13.936 to go away and there's going to

be an evolution of the way people use this technology. 00:12:14.296 --> 00:12:18.796 But embrace it or you can quickly be left behind. 00:12:18.916 --> 00:12:24.516 Because the biggest difference I see between the internet and this generative AI. 00:12:24.876 --> 00:12:29.256 Especially not just AI, but generative AI, is that it's really, 00:12:29.416 --> 00:12:34.076 really faster this time than it was in, say, late 90s, right, 00:12:34.156 --> 00:12:38.216 when the internet was really starting to come into focus in terms of how do 00:12:38.216 --> 00:12:39.756 you use it for business applications. 00:12:39.756 --> 00:12:44.736 So I think this is really this is a fast moving train and you got to hop on 00:12:44.736 --> 00:12:49.376 or really you'll you'll fall far behind fast, I believe. Yeah. Oh, absolutely. 00:12:49.516 --> 00:12:55.036 And that's the whole thing around agentic and agents is basically every gorilla 00:12:55.036 --> 00:12:58.776 in the tech space and all the startups and all the old STP players, 00:12:59.016 --> 00:13:03.616 everybody wants to own that kind of agentic orchestration layer, 00:13:03.936 --> 00:13:09.616 taking the LLMs and baking that into your business processes and into your analytics 00:13:09.616 --> 00:13:12.956 because they don't want to be below the line. and they want to be above the line, right? 00:13:13.136 --> 00:13:18.016 And so, you know, that just, so that basically means that every single tech 00:13:18.016 --> 00:13:20.396 vendor is focusing on this. 00:13:20.636 --> 00:13:26.756 So just, and the supply market is fragmented, but it's huge and growing. 00:13:26.976 --> 00:13:32.996 And it's, you know, all about how do you drive the value of these LLMs and of 00:13:32.996 --> 00:13:36.876 Gen AI into the business processes to add more value. 00:13:37.016 --> 00:13:41.496 And primarily that is happening through this agentic framework that is emerging. 00:13:41.736 --> 00:13:44.836 And we don't get too much into the rat hole inside the black box there, 00:13:44.916 --> 00:13:49.676 but absolutely, it is foundational. It is the biggest change that I've seen 00:13:49.676 --> 00:13:52.956 in my 25 years kind of covering the ProcureTech space. 00:13:53.576 --> 00:13:57.456 Great. Okay. So now I want to move into the vendor landscape, 00:13:57.656 --> 00:13:58.576 sort of where we are today. 00:13:59.536 --> 00:14:06.076 One, there was a lot of money went into the supply chain and procurement space, I would say. 00:14:06.196 --> 00:14:09.236 And I used to say, I remember when it was hard to raise money, 00:14:09.396 --> 00:14:12.576 right, trying to run a software company and get funding. 00:14:12.916 --> 00:14:16.956 And it seemed like it was overfunded, right? There were almost too many choices 00:14:16.956 --> 00:14:19.756 out there. And now you've got all those choices. 00:14:20.336 --> 00:14:23.276 Many of them are still around. Some have gone out of business. 00:14:23.476 --> 00:14:27.056 Some have consolidated, right? Things that naturally happen in that world. 00:14:27.496 --> 00:14:30.696 But now you've got this AI element coming in. 00:14:30.816 --> 00:14:35.376 And so if you're out there buying technology, if you're part of the procurement 00:14:35.376 --> 00:14:39.636 team or the CPO and you're out buying technology, how do you cut through the hype? 00:14:39.756 --> 00:14:45.356 How do you really find out where the value is and maybe pick the right horses 00:14:45.356 --> 00:14:46.816 at this point in the game? 00:14:47.596 --> 00:14:50.736 Yeah, no, it's a great question. So one is, I mean, you almost, 00:14:50.916 --> 00:14:54.756 there's kind of like starting with your existing ProcureTech stack. 00:14:54.956 --> 00:14:57.836 And our old approach is just throwing apps at the problem, right? 00:14:57.876 --> 00:15:02.256 This is like Monty Python in the castles, right? We'd throw this castle and 00:15:02.256 --> 00:15:04.256 it sank into the swamp. And then we'd throw another one. 00:15:04.796 --> 00:15:07.476 You know, we had ERP, right? And then we had S2P suites, right? 00:15:07.576 --> 00:15:11.076 And then we had, you know, best of breed. And now we're hooking them all together. 00:15:11.296 --> 00:15:14.636 And now, right, so we're just bolting on these things. And underneath the covers, 00:15:14.836 --> 00:15:18.916 they're all kind of similar, similar data models. So now we're fragmenting all our data. 00:15:19.036 --> 00:15:22.656 Now we have 17 supplier masters, you know, all over the place. 00:15:22.796 --> 00:15:24.576 And it's going to be great. Now we've got to throw another, you know, 00:15:25.276 --> 00:15:30.536 information, you know, MDM tool that sits on top of, you know, these various tools. 00:15:30.696 --> 00:15:33.236 And that's just within source to pay. That's not even kind of across, 00:15:33.416 --> 00:15:37.076 you know, the landscape when you get into enterprise risk or compliance or whatever. 00:15:37.216 --> 00:15:40.416 Right. So when you start from kind of the traditional apps, right, 00:15:40.576 --> 00:15:45.496 they are all over undergoing the vendors, they're all having to kind of retool 00:15:45.496 --> 00:15:50.336 their architectures and not just bolt, you know, kind of the agents on top and whatnot. 00:15:50.516 --> 00:15:53.076 So a lot of times a lot of this stuff is just going to show up, 00:15:53.136 --> 00:15:54.236 but you have to just be smart. 00:15:54.828 --> 00:15:58.668 In how you buy some of this from the legacy tech vendors. 00:15:59.068 --> 00:16:03.428 Don't pay extra for all the token costs, you know, unless there's really clear ROI. 00:16:03.828 --> 00:16:07.868 Work closely with IT, right, and all that. But on the other side is, 00:16:08.088 --> 00:16:12.268 you know, I was saying before, great, if you have something like Copilot and 00:16:12.268 --> 00:16:15.508 you can like do an agent builder and, you know, just start getting familiar 00:16:15.508 --> 00:16:18.088 with

the tools, it's fine to do that. 00:16:18.188 --> 00:16:23.728 But also just look around the enterprise as you're working with IT and with HR or marketing. 00:16:23.728 --> 00:16:27.148 You know, if there are some or like analytics team or whatever, 00:16:27.408 --> 00:16:31.428 if there's stuff that's kind of happening, right, glom onto that for sure, 00:16:31.668 --> 00:16:34.888 you know, and be able to see, great, how can I now apply that? 00:16:35.248 --> 00:16:38.088 And maybe we can get some folks together and work in a certain area. 00:16:38.268 --> 00:16:41.908 Maybe it's around, you know, supplier compliance, you know, and how do we do 00:16:41.908 --> 00:16:46.148 supplier onboarding and certifications and bring some of the agentic stuff and 00:16:46.148 --> 00:16:49.988 have it work with, you know, you may have heard around, you know, orchestration, right? 00:16:49.988 --> 00:16:53.288 Intake management is certainly a process, right? Common way for stakeholders 00:16:53.288 --> 00:16:54.368 to engage with procurement. 00:16:54.708 --> 00:16:57.468 But orchestration is around kind of this coordination and workflow. 00:16:57.868 --> 00:17:01.808 And now we can agentize it. That's a great place to start. 00:17:01.908 --> 00:17:04.808 There's basically two ways to do transformation. There's this process-centric 00:17:04.808 --> 00:17:07.408 thing, and then there's kind of the analytics, right, around spend analytics. 00:17:07.728 --> 00:17:12.028 And then you start branching off, right, supplier and risk and contract and all these analytics. 00:17:12.368 --> 00:17:15.708 Single pane of glass, look back down into the business, where are the opportunities? 00:17:16.048 --> 00:17:21.008 And those things kind of work together. So it's kind of working with the business, 00:17:21.168 --> 00:17:23.768 doing a little organizational judo, right, or aikido, 00:17:23.968 --> 00:17:28.368 like wherever the, you know, the momentum is going, you try to latch onto that 00:17:28.368 --> 00:17:33.468 and use that to kind of fund some stuff that you're doing in a particular area. 00:17:33.468 --> 00:17:37.108 And if you got a little bit of budget to kind of that you're maybe given to 00:17:37.108 --> 00:17:42.388 do something around agentic AI, great, you know, pilot with some small startup, 00:17:42.608 --> 00:17:44.388 you know, I mean, got to spend matters. 00:17:44.568 --> 00:17:48.388 We have just tons of insights on all the startups as well as what the traditional players are doing. 00:17:48.808 --> 00:17:52.308 I mean, that's kind of like our core business, but, you know, 00:17:52.648 --> 00:17:54.348 just get your hands dirty with it. 00:17:54.868 --> 00:17:57.028 Pilots are great. But before you scale. 00:17:57.727 --> 00:18:02.647 That's where you have to work hand in hand with IT. Do not ignore them, 00:18:02.747 --> 00:18:05.287 right? No shadow procurement, no shadow IT. 00:18:05.967 --> 00:18:09.987 Procurement and IT live like amazingly parallel lives, right? 00:18:10.447 --> 00:18:15.007 IT does digitization. Procurement does externalization. You got to get those, 00:18:15.207 --> 00:18:17.227 you know, you got to be buddy system there, right? 00:18:17.307 --> 00:18:20.067 To help each other and to help the business, right? 00:18:20.167 --> 00:18:23.467 And so I think if you, that's a foundational thing, right? If you have that 00:18:23.467 --> 00:18:29.207 alignment, that helps a lot because now you can kind of get access to more of the goodies, right? 00:18:29.267 --> 00:18:31.867 If you're actually helping IT source some of these tools, right? 00:18:31.927 --> 00:18:34.347 You can kind of see all the stuff that's coming into the candy store. 00:18:34.627 --> 00:18:37.247 So you can kind of see some of those things that you can make use of. 00:18:37.327 --> 00:18:38.547 So there's a lot of on-ramps, right? 00:18:38.627 --> 00:18:43.327 I've just given you some examples here, but definitely keep IT and keep the 00:18:43.327 --> 00:18:46.767 supply market close at hand so you can figure out, you know, 00:18:46.967 --> 00:18:49.627 when you can bring that to bear to deliver some value. 00:18:50.267 --> 00:18:56.427 Yeah. So get some people together, make sure you've got a team looking at this 00:18:56.427 --> 00:19:02.287 and it's on the agenda, right, for tech selection, whether that's out of your COE or whatever. 00:19:02.507 --> 00:19:05.667 Got to have a COE. You've got to have a procurement center of excellence. 00:19:05.767 --> 00:19:09.647 And I don't mean like the administrative group that counts up the savings for finance, right? 00:19:09.787 --> 00:19:13.247 But someone that's really doing the outside in-market intelligence, 00:19:13.587 --> 00:19:17.427 right, that's doing training enablement and the digital COE. 00:19:17.547 --> 00:19:19.967 And that doesn't have to be some big thing if you're not a big company. 00:19:20.267 --> 00:19:24.127 It just means you have a single, accountable, responsible person within procurement, 00:19:24.627 --> 00:19:28.827 responsible for getting the IT infrastructure that you need to do your job. 00:19:28.967 --> 00:19:33.127 And you need to have the same thing over on the IT side as your lead IT procurement people. 00:19:33.247 --> 00:19:36.787 And you better have the buddy system between IT and procurement, 00:19:36.967 --> 00:19:38.587 not just with the CPO and the CIO, 00:19:38.867 --> 00:19:43.287 but also kind of with the centers of excellence to make sure everybody's aligned 00:19:43.287 --> 00:19:46.687 here with the projects that you're doing, your objectives, getting that aligned 00:19:46.687 --> 00:19:49.287 and kind of working jointly on some of these things. 00:19:49.767 --> 00:19:55.347 Great. Yeah, work together with IT. I think it's so important that that partnership 00:19:55.347 --> 00:19:59.407 is there and that alignment is there to make anything like this successful, 00:19:59.587 --> 00:20:02.947 obviously, given

the deep reach into the data. 00:20:03.187 --> 00:20:09.127 So you talked about guided buying and you coined the term. Let's give you mad props for that again. 00:20:09.427 --> 00:20:14.287 And you've also, so far, you've quoted Rush and you've referenced Monty Python. 00:20:14.447 --> 00:20:17.627 So I give you great, great credit for that as well. Thank you. 00:20:19.477 --> 00:20:22.857 Let's talk about autonomous procurement just for a second, because, 00:20:23.197 --> 00:20:25.277 I mean, that's sort of the dream, right? 00:20:25.437 --> 00:20:28.857 And guided buying, it's sort of steered you to the right place to buy. 00:20:29.037 --> 00:20:30.677 It's a phenomenal term, by the way. 00:20:30.857 --> 00:20:34.717 I thought it was, when you hear it, you know exactly what it is if you're in this world, right? 00:20:35.257 --> 00:20:39.917 Guide to the right supplier so you buy on contract and ensure compliance and 00:20:39.917 --> 00:20:43.357 you ensure that the category strategy plays out all the way at the point of 00:20:43.357 --> 00:20:45.777 view. It's a co-pilot class solution, right? Right. 00:20:45.877 --> 00:20:49.797 It's guiding you to the right place with a best practice process and workflow 00:20:49.797 --> 00:20:52.537 and system and all that stuff. 00:20:52.697 --> 00:20:54.737 Right. But it's human in the loop. Yeah. 00:20:55.337 --> 00:20:59.617 Right. So can we take that human out in autonomous procurement? 00:20:59.757 --> 00:21:03.337 And I'm not talking about running MRP. It has been running for years. 00:21:03.577 --> 00:21:09.657 I'm talking about a machine making a decision without a person, 00:21:09.657 --> 00:21:15.077 you know, at the very end there to say, wait a minute, before we say yes to this decision. right? 00:21:15.257 --> 00:21:20.377 I need to review it, right? Do you think it can happen or is it already happening? I don't know. 00:21:20.617 --> 00:21:25.217 For sure. Absolutely. So I think one of the things is the term autonomous is 00:21:25.217 --> 00:21:29.737 an important term because ultimately, if you think about what an agent is and what that means, 00:21:30.157 --> 00:21:34.197 it means you're given a piece of software and these are pieces of software wrapped 00:21:34.197 --> 00:21:38.757 around LLMs as much as we want to, you know, personify them and turn them into 00:21:38.757 --> 00:21:40.197 humans. And I understand that, right? 00:21:40.357 --> 00:21:43.417 It's like They have skills, they're kind of, you know, whatever. 00:21:43.757 --> 00:21:47.117 They can kind of seem, you know, human like you train them, right? 00:21:47.197 --> 00:21:49.177 They're like your little, they're like minions, right? 00:21:49.317 --> 00:21:53.377 You know, it's like, they're kind of human, but, you know, they're kind of like bots. 00:21:53.537 --> 00:21:57.517 But there's a guy named Andrew Ng who's like, you know, a big tech guru. 00:21:57.877 --> 00:22:01.697 And he says, you know what, the terminology has just kind of gone off the rails. 00:22:01.877 --> 00:22:08.897 Let's just say we have agentic AI systems, right, that have varying levels of autonomy. 00:22:09.502 --> 00:22:16.322 And the more autonomy that you have using agents, it just means the agents have more agency. 00:22:16.422 --> 00:22:21.982 They have more of the ability to do things themselves, even if it's just, 00:22:22.082 --> 00:22:24.962 let's say, a final kind of presenting you with some options, 00:22:24.982 --> 00:22:28.282 which you can argue, well, if you get numb to that and you're just constantly, 00:22:28.282 --> 00:22:30.582 you know, like the rat hitting the lever, it's like, 00:22:30.922 --> 00:22:33.922 you know, what good is that, right? 00:22:33.922 --> 00:22:39.502 But the term autonomous has been around for a long time with just with straight up tech automation. 00:22:39.522 --> 00:22:44.642 The problem with the old, you know, approaches is that as processes are not 00:22:44.642 --> 00:22:47.962 just, you know, this is like, you know, ISO, say what you do, 00:22:48.082 --> 00:22:50.522 do what you say. The processes are nice and documented. 00:22:50.882 --> 00:22:56.782 I mean, stuff's changing all the time. Right. So you have to be able to shapeshift, 00:22:56.782 --> 00:22:59.962 you know, the integrations and, you know, things like that, you know, 00:23:00.142 --> 00:23:04.942 with this workflow layer, just independent of, you know, agentic AI. 00:23:05.162 --> 00:23:08.562 So to do it, but the problem is, is every time things change, 00:23:08.762 --> 00:23:12.402 then you got to manage all the integration and all this stuff around orchestration 00:23:12.402 --> 00:23:18.402 really requires a lot of tight integration, you know, as well as more and more, 00:23:18.542 --> 00:23:22.162 it's becoming more agentic to make it more autonomous. Great. 00:23:22.922 --> 00:23:28.602 Okay. Shift gears once more. Let's look to the future. So we have a CPO. It's 2025. 00:23:29.382 --> 00:23:33.062 Still got six months to go. I think we just hit the halfway point in the year 00:23:33.062 --> 00:23:34.542 a couple of days ago, right? 00:23:35.022 --> 00:23:41.542 So, you know, what should the CPO be thinking about in 2025 and beyond? 00:23:41.542 --> 00:23:44.342 You've touched on having a COE to look at this. 00:23:44.482 --> 00:23:49.482 I think that's a super smart idea. there's got to be some central place for it, right? 00:23:49.822 --> 00:23:55.522 To come up with the strategy for any technology serving a function in the business. 00:23:55.722 --> 00:24:00.582 But what should they be thinking about in terms of the AI strategy? 00:24:00.622 --> 00:24:04.542 If it was you and you just got dropped in as a CPO somewhere, 00:24:05.524 --> 00:24:08.584 Where do you begin? No, that's a great question. 00:24:09.124 --> 00:24:13.144 I mean, to net it out, I would say focus on the I. 00:24:13.484 --> 00:24:18.224 So focus on the I in IT, namely the information versus just the T

in the tech 00:24:18.224 --> 00:24:19.904 apps and the process automation, right? 00:24:20.024 --> 00:24:25.784 Think about become the master of your own destiny around data, 00:24:25.964 --> 00:24:30.384 data management, and the analytics that use that data and, you know, 00:24:30.484 --> 00:24:35.004 that use it and also, you know, create it, right? Also, the I in AI. 00:24:36.304 --> 00:24:39.344 So, intelligence, right? And the whole thing around intelligence is, 00:24:39.524 --> 00:24:42.464 and the key to, like, if you're going to think about, oh, the future and AGI, 00:24:42.944 --> 00:24:46.424 you know, and, you know, the key to AGI and where we're going, 00:24:46.544 --> 00:24:50.324 you know, long term with whenever, you know, the singularity comes or whatever, 00:24:50.604 --> 00:24:54.604 it's really around knowledge representation and knowledge modeling and management. 00:24:54.604 --> 00:24:58.804 And in fact, if you go back to 2001 and there was a paper written by a guy named 00:24:58.804 --> 00:25:03.504 Tim Berners-Lee, a smart guy, basically invented the internet and came up with 00:25:03.504 --> 00:25:07.964 things like HTTP and HTML and all these things. 00:25:08.504 --> 00:25:12.824 He wrote this article and basically it was like, first it was around the semantics 00:25:12.824 --> 00:25:16.844 and the data and the information that eventually agents will need to communicate 00:25:16.844 --> 00:25:19.104 with each other. And we've like gone right to the agents, right? 00:25:19.584 --> 00:25:23.804 And at the same time, what you really need is the information that you need 00:25:23.804 --> 00:25:29.024 to make the decisions to, you know, be like, where should I focus my time? 00:25:29.324 --> 00:25:33.684 Or if you just think about, you know, analytics, I would say the one single 00:25:33.684 --> 00:25:37.504 analytic to focus on is opportunity analytics, right? 00:25:37.604 --> 00:25:40.044 And you know this, right, you know, certainly within sourcing, 00:25:40.244 --> 00:25:43.984 but where, you know, what's all the value that's out there hiding, 00:25:43.984 --> 00:25:47.024 you know, that I don't have time to like go through all these dashboards, 00:25:47.264 --> 00:25:50.364 right, and all this stuff, or just bring in a consultant all the time, 00:25:50.704 --> 00:25:53.984 you know, serve up, you know, where are the things where I have the most risks 00:25:53.984 --> 00:25:57.664 or where I have the most opportunities, stick it in that two by two, right? 00:25:57.784 --> 00:26:00.964 And then I can kind of figure out, all right, what can I actually go after? 00:26:01.124 --> 00:26:02.644 And the more in which you have. 00:26:03.246 --> 00:26:06.586 Agenta capabilities or just execution and automation capabilities, 00:26:06.846 --> 00:26:09.906 that ability to execute becomes easier. 00:26:10.066 --> 00:26:12.046 So now things start shifting, right? So you can be like, oh, 00:26:12.086 --> 00:26:14.126 I can push that into

wave one, wave two. 00:26:14.566 --> 00:26:20.046 So focus on the analytics to surface up opportunities. 00:26:20.546 --> 00:26:24.786 And to do those analytics means you need to get your house in order around the 00:26:24.786 --> 00:26:30.786 master data suppliers, contracts, Build out your category, taxonomy and knowledge, 00:26:31.286 --> 00:26:34.746 cost modeling, risk, you know, what are the biggest risks? 00:26:34.986 --> 00:26:38.826 And all of these things interact with the rest of the business around what are 00:26:38.826 --> 00:26:41.886 the biggest risks and compliance requirements that we have, sustainability, 00:26:42.086 --> 00:26:44.166 ESG, all those things, right? 00:26:44.506 --> 00:26:47.806 So, again, there's two design patterns to transformation. 00:26:48.006 --> 00:26:51.286 One is kind of layer the workflow on top, right, and of which, 00:26:51.346 --> 00:26:54.246 you know, intake is a great piece, right, and the supplier portal is great for 00:26:54.246 --> 00:26:56.626 suppliers. The other is data and analytics. 00:26:57.066 --> 00:27:01.106 Data and analytics are going to inform what's your biggest economic cost, 00:27:01.286 --> 00:27:04.946 go-forward cost, the things where you can drive the most value to the business. 00:27:05.326 --> 00:27:12.086 That's a good place to start, right? So I would focus on the I and on the intelligence piece. 00:27:12.246 --> 00:27:17.906 And there's some interesting technology that is emerging in that space around. 00:27:19.326 --> 00:27:22.386 It's a data-centric approach, data mesh, data products. 00:27:22.386 --> 00:27:26.366 I don't want to send you books down the rat hole, but this idea of using things 00:27:26.366 --> 00:27:31.626 called knowledge graphs, this ability to model your data and outside of the 00:27:31.626 --> 00:27:34.446 apps that you're buying to automate different processes, 00:27:34.746 --> 00:27:37.446 it's kind of like liberating your data and bringing that up, 00:27:37.466 --> 00:27:40.406 let's say, into a data lake and then kind of the analytics wrapped around that. 00:27:40.406 --> 00:27:44.286 That whole analytics and intelligence piece is 00:27:44.286 --> 00:27:47.126 absolutely crucial because otherwise you just have these you know 00:27:47.126 --> 00:27:50.486 dumb agents and bots that are just trying to plug on 00:27:50.486 --> 00:27:54.426 to whatever you have and make that more efficient but you're just making your 00:27:54.426 --> 00:27:57.186 exist you know you're just putting lipstick on the little pigs and you're making 00:27:57.186 --> 00:28:00.586 the pigs run faster but you're not saying hey why do i have all these old pigs 00:28:00.586 --> 00:28:03.826 here i'm not beating this analogy to death you know why don't i actually have 00:28:03.826 --> 00:28:08.286 a better you know procure tech stack and a way to do that is independent of 00:28:08.286 --> 00:28:09.226 whatever tools you're using, 00:28:09.406 --> 00:28:14.746 make sure that the data that you have is high fidelity and giving you the insights 00:28:14.746 --> 00:28:17.286 that you need to go after these things. 00:28:17.406 --> 00:28:20.086 And oh, by the way, as you go after them, then you can self-fund, 00:28:20.206 --> 00:28:23.226 replace out some of the old tooling, put in some of the new apps, 00:28:23.526 --> 00:28:24.406 that kind of stuff, right? 00:28:24.546 --> 00:28:30.546 So focus on the I, not the T, focus on the I in IT and focus on the I in AI, 00:28:31.246 --> 00:28:34.746 not just the A part around artificial and agentic. 00:28:35.186 --> 00:28:39.406 Yeah. Coming back to that information, data, right? This is something we haven't 00:28:39.406 --> 00:28:42.306 talked about, but it's pretty obvious that. 00:28:43.109 --> 00:28:48.689 If the underlying fuel, if you will, for agentic AI is the data, 00:28:49.029 --> 00:28:56.649 the veracity and accuracy and timeliness of that data is super critical, right? 00:28:56.829 --> 00:29:01.609 So I see a lot of the data, what I'll call the data vendors out there who are 00:29:01.609 --> 00:29:07.609 cleaning up databases, enriching data, right, are really pushing that agenda. 00:29:08.149 --> 00:29:11.909 But I think they're right. You know, you have to have clean data. 00:29:12.369 --> 00:29:15.569 And how do you keep it clean? How do you keep it from getting polluted again? 00:29:16.009 --> 00:29:20.569 You know, is there any thoughts you have just about how do you make sure your 00:29:20.569 --> 00:29:24.749 data is as good as it could possibly be considering this is the real world and 00:29:24.749 --> 00:29:26.169 it's never going to be perfect? 00:29:26.529 --> 00:29:31.349 Right, for sure. So let's talk about kind of just the basics of you got a bunch 00:29:31.349 --> 00:29:35.449 of apps and those apps are prosumers of data, right? 00:29:35.529 --> 00:29:40.269 They produce data and they consume data and they produce a lot of crappy data. 00:29:40.269 --> 00:29:44.169 And if you haven't fixed the problem of we're going to give end users commodity 00:29:44.169 --> 00:29:46.969 code dropdowns that they can pick whatever the hell they want. 00:29:47.329 --> 00:29:50.689 And then you're like, well, why is my spend data? You know, why am I buying? 00:29:50.769 --> 00:29:52.729 You know, I didn't know I was buying nuclear material. 00:29:52.909 --> 00:29:56.829 You're not. You just that was like your first dropdown code or whatever. 00:29:57.169 --> 00:30:00.849 You know, you had miscellaneous or other or 999. You know, you've been in this 00:30:00.849 --> 00:30:04.969 game for a long time. It's like, you know, you have got to put the controls, 00:30:04.989 --> 00:30:10.369 you know, in place to make sure there is that data quality at the source. 00:30:10.569 --> 00:30:15.509 And this is where AI can help. If you go back to that whole guided buying thing 25 years ago, 00:30:15.769 --> 00:30:20.729 you know, one of the technologies was real time and use cases was

real time 00:30:20.729 --> 00:30:27.329 auto classification of your spend at the time of need and requisitioning and 00:30:27.329 --> 00:30:28.769 POs and, you know, all that. 00:30:28.769 --> 00:30:33.909 So that we can actually guide people down the right paths based on the category 00:30:33.909 --> 00:30:38.709 of stuff that they're looking for, services that they're looking for. Genius, right? 00:30:38.889 --> 00:30:42.729 But that means real-time auto-classification, right? And having that, 00:30:42.889 --> 00:30:45.109 fixing that data right at the source. 00:30:45.349 --> 00:30:48.269 There's a lot of, I mean, there's a lot of myths right now. And like you're 00:30:48.269 --> 00:30:51.289 saying, right, there's a hype cycle that happens. 00:30:51.909 --> 00:30:55.849 And right now it's like you kind of have, and unfortunately it's like we've 00:30:55.849 --> 00:30:57.169 almost been conditioned end. 00:30:57.729 --> 00:31:01.289 To be like, oh, this is hype. It's going to go away, right? 00:31:01.329 --> 00:31:03.689 There's going to be a big hype and then it's going to drop off and then we're 00:31:03.689 --> 00:31:06.569 going to get back to business as usual and maybe a few things are going to change. 00:31:07.449 --> 00:31:11.029 And that's not really going to happen this time because it is fundamentally 00:31:11.029 --> 00:31:16.569 kind of a shift as big as actually kind of the internet itself and then kind 00:31:16.569 --> 00:31:19.209 of, you know, getting apps to be, you know, to SaaS, right? 00:31:19.449 --> 00:31:22.329 And then now to, you know, the agenda gauge. 00:31:22.469 --> 00:31:26.009 So, you know, there's one myth which is like, Like, oh, you know, 00:31:26.429 --> 00:31:29.709 Gen.AI is going to, you know, going to come in and agents are just going to 00:31:29.709 --> 00:31:31.649 replace everybody and replace SaaS. 00:31:32.529 --> 00:31:36.589 And that is not really going to happen because, first of all, agents are SaaS. 00:31:36.969 --> 00:31:42.789 Two, agents help you improve your enterprise SaaS landscape, as we've talked about. 00:31:43.209 --> 00:31:46.389 And also, you just need to kind of get that house in order. And a lot of these 00:31:46.389 --> 00:31:50.069 agents, they don't really, they're not, you know, kind of SaaS apps. 00:31:50.069 --> 00:31:54.969 They're not like big, persistent data model, you know, kind of based applications 00:31:54.969 --> 00:31:58.789 that are your system of record, your system of kind of, you know, 00:31:59.329 --> 00:32:01.189 analytics, you know, your core workflows. 00:32:01.769 --> 00:32:05.529 They're kind of like augmenting that, right, and helping you make those more autonomous. 00:32:05.709 --> 00:32:10.089 So AI is not going to, you know, just replace SaaS. 00:32:10.349 --> 00:32:13.229 It's also not going to just replace people. But certainly that thing about, 00:32:13.429 --> 00:32:18.769 you know, it won't replace you, but the person using AI is certainly going

to 00:32:18.769 --> 00:32:22.029 maybe replace you or at least be more competitive, you know, 00:32:22.089 --> 00:32:23.889 than you are. That is true. 00:32:23.989 --> 00:32:26.249 But at the same time, now there's this myth of kind of like, 00:32:26.369 --> 00:32:30.789 oh, you know, agents are just, you know, it's hype. 00:32:31.069 --> 00:32:35.669 It's going to go away. You know, we don't want to let these things touch our systems. 00:32:36.409 --> 00:32:40.169 And the thing is, is that they are going to evolve. They are going to be more 00:32:40.169 --> 00:32:42.229 than just the easy, the simple use cases. 00:32:42.449 --> 00:32:47.809 They will add the knowledge and the, you know, the ontologies, 00:32:48.109 --> 00:32:51.129 you know, kind of the analytics and the data modeling. They will get better. 00:32:51.669 --> 00:32:55.229 And they're just, and they're already kind of, honestly, we're at the point 00:32:55.229 --> 00:33:00.509 already where the biggest constraint is no longer the technology. 00:33:00.509 --> 00:33:05.649 It's actually our thinking and actually our proclivity to actually use these 00:33:05.649 --> 00:33:10.549 tools. So I would just challenge every listener out there, get out there and 00:33:10.549 --> 00:33:12.089 like take one of the tools. 00:33:12.449 --> 00:33:15.329 And, you know, for example, I just don't, you know, a couple of weeks ago, 00:33:15.409 --> 00:33:16.229 right? It was like, all right. 00:33:17.149 --> 00:33:21.109 I think it was chat TPT or DeepSeek or Claude, I'm sure I remember. 00:33:21.189 --> 00:33:24.289 But it was like, what a best practice for a category strategy. Great. 00:33:24.689 --> 00:33:28.949 Give me a stakeholder interview guide for the things that I need from them to inform that. 00:33:29.750 --> 00:33:33.910 Now do that and go out to a category called facilities management. 00:33:34.490 --> 00:33:36.770 Terrific. Who are the biggest players that are out there? Okay, 00:33:36.950 --> 00:33:40.510 now apply category strategy to that and now give me a set of negotiation tactics 00:33:40.510 --> 00:33:44.570 to engage them based on the strategy that I've, you know, now determined I want 00:33:44.570 --> 00:33:47.630 to go for an integrated systems, you know, supplier. 00:33:47.910 --> 00:33:51.490 Terrific. Now take that, you know, whatever. And then ultimately you can, 00:33:51.610 --> 00:33:55.830 as you're doing those things, you're almost, there's an agentic feature where 00:33:55.830 --> 00:33:59.250 the tools are out there, they're gathering data, they're analyzing things. 00:33:59.250 --> 00:34:02.990 They're like, they're coming up with a strategy to help you come up with this stuff. 00:34:03.290 --> 00:34:06.190 They'll generate forms and inquiries and tables. 00:34:06.370 --> 00:34:10.670 You can come up, you can like come up with a category management application. 00:34:10.710 --> 00:34:14.570 You can just like auto-gen that, you know, vibe coding is the term, 00:34:14.730 --> 00:34:18.110 but, you know, proceed with caution, right? 00:34:18.250 --> 00:34:22.150 But that ability and then say, great, now load this into my sourcing tool, 00:34:22.350 --> 00:34:25.430 you know, to kind of with optimized negotiation strategies. 00:34:26.030 --> 00:34:29.590 Yeah, it's going to hallucinate potentially some stuff. The intelligence won't 00:34:29.590 --> 00:34:35.010 be perfect because it's scraping a lot of mass data, but it is pretty scary 00:34:35.010 --> 00:34:39.070 and sometimes scary good as to how well it can do a lot of these things. 00:34:39.270 --> 00:34:46.450 And it's only continuing to get smarter. So you need to really understand what these things can do. 00:34:46.670 --> 00:34:50.950 So I don't want to go full Yoda here, but, you know, it's the challenges within 00:34:50.950 --> 00:34:56.030 yourself, you know, and change how you think more than not. So that is that 00:34:56.030 --> 00:34:59.570 myth, like, oh, this is just a fad and, you know, the stuff's going to go away. 00:34:59.710 --> 00:35:02.390 It's not going away. So you better know how to master it. 00:35:03.013 --> 00:35:06.113 No, so embrace it. Thank you, Yoda. I appreciate that. 00:35:06.933 --> 00:35:11.273 So, hey, listen, I want to give you a chance to plug maybe some of your research. 00:35:11.473 --> 00:35:16.753 Where can our listeners on the pod, where can they dive deeper into your AI 00:35:16.753 --> 00:35:20.773 research at Spend Matters, you know, which is now part of the Hackett Group? 00:35:21.233 --> 00:35:24.293 Yeah, yeah, for sure. So, I mean, I think the easiest thing is you can just 00:35:24.293 --> 00:35:26.093 go to our site and just sign up for the newsletter. 00:35:26.653 --> 00:35:30.653 That's pretty easy. It's, you know, in my, in a lot of procurement folks. 00:35:30.753 --> 00:35:33.193 And for me, you know, hey, if it's free, it's me, right? So, 00:35:33.413 --> 00:35:34.453 you know, we don't spam you. 00:35:34.773 --> 00:35:39.373 I think, you know, it's like, I think you can set the frequency on the newsletter subscription. 00:35:39.673 --> 00:35:42.793 And that gives you just kind of a roll up of the stuff that we write about. 00:35:42.913 --> 00:35:47.133 I mean, not, and most of it's all just available that you can get access to. 00:35:47.393 --> 00:35:51.273 There is stuff that is proprietary, that is the market intelligence from particular providers. 00:35:51.893 --> 00:35:54.773 So if you're going to be doing an RFI and you want to go out to market and you 00:35:54.773 --> 00:35:58.753 want to see how, you know, whatever the SDP guys are doing versus some best 00:35:58.753 --> 00:36:01.753 of breeds and the AI and like our latest research on AI. 00:36:02.473 --> 00:36:05.633 And some of these topics that we're, you know, delving into, 00:36:05.873 --> 00:36:07.873 I think that's probably the easiest place to go. 00:36:07.973 --> 00:36:11.853 And we also have a bunch of free guides too, which

are basically like kind of 00:36:11.853 --> 00:36:15.333 pick the area that you're interested, you know, whether it's e-sourcing or spend 00:36:15.333 --> 00:36:20.153 analytics or supply chain risk or whatever it is, you can kind of contract management. 00:36:20.313 --> 00:36:23.533 You can dive in there and it'll basically be like, here's what it is. 00:36:23.653 --> 00:36:26.993 Here's the space. Here's what it's all about. Here are kind of the main features to look for. 00:36:27.253 --> 00:36:30.673 Here are some of the vendors in the space. So we, you know, kind of show a little 00:36:30.673 --> 00:36:35.153 leg, give away, you know, a certain amount of content and just get you educated on the space. 00:36:35.293 --> 00:36:38.513 So I would say that's probably the best place to go is kind of to go to our 00:36:38.513 --> 00:36:42.553 guides section or just do the free newsletter sign on. 00:36:42.673 --> 00:36:44.953 But, you know, obviously on LinkedIn, if you want to, you know. 00:36:45.873 --> 00:36:48.753 Connect with me or subscribe or follow or, you know, any of that stuff, 00:36:48.973 --> 00:36:51.153 either me personally and the company. 00:36:51.513 --> 00:36:53.373 Yeah, that's kind of the easiest thing to start with. 00:36:54.047 --> 00:36:57.467 Awesome. Now, I think you bring such a wealth of knowledge. 00:36:57.687 --> 00:37:01.627 Obviously, it was on display here. Here, you can go super deep on these subjects, 00:37:01.627 --> 00:37:05.727 and it's always great to talk to you and sort of understand what's happening. 00:37:05.847 --> 00:37:10.607 I think there's just a tremendous amount of uncertainty out there because things 00:37:10.607 --> 00:37:11.987 are changing so quickly. 00:37:12.227 --> 00:37:15.187 And I know some people that I talk to, 00:37:15.347 --> 00:37:18.927 they don't want to be left behind, but they also want to make sure they're not the, 00:37:18.927 --> 00:37:21.807 you know the pioneer the the planes are 00:37:21.807 --> 00:37:25.147 littered with a lot of dead pioneers right so 00:37:25.147 --> 00:37:28.327 they want to they want to approach it the right way 00:37:28.327 --> 00:37:32.547 and i think you your guidance can help them approach it maybe the right way 00:37:32.547 --> 00:37:37.787 the optimal way and and arrive at uh at value faster without without making 00:37:37.787 --> 00:37:42.327 too many mistakes but i think like anything i think if you don't make any mistakes 00:37:42.327 --> 00:37:46.207 you're probably not trying hard enough and trying new things. 00:37:46.627 --> 00:37:50.207 And so I think if to the extent that there can be some controlled experiments 00:37:50.207 --> 00:37:54.087 and, you know, guidelines and guardrails around this activity, 00:37:54.087 --> 00:37:56.407 I think people should just go for it, right? 00:37:56.487 --> 00:37:59.047 That's what it was coming through steady in your message. 00:37:59.487 --> 00:38:01.707 Totally, man. Totally. If you want to go deep and have human, 00:38:01.907 --> 00:38:04.467 you know, human in the loop, you know, advisory and intelligence. 00:38:04.667 --> 00:38:06.687 Yeah. You know, feel free to reach out, right? That's. 00:38:08.469 --> 00:38:11.849 For the rest where you're just trying to get educated on all this stuff a lot of this stuff, 00:38:12.369 --> 00:38:15.229 honestly i would just get your personal version of you know chat 00:38:15.229 --> 00:38:18.209 gpt put on the you know the voice voice mode 00:38:18.209 --> 00:38:21.749 and you know you don't have to have the voice mode you can just you know type 00:38:21.749 --> 00:38:26.409 or whatever for a good typer but basically just ask your questions and be like 00:38:26.409 --> 00:38:29.769 i don't get it you know or how does this relate to this or i don't know that 00:38:29.769 --> 00:38:33.869 seems like a little optimistic you know what about bloody blood right and of 00:38:33.869 --> 00:38:37.009 course these chatbots are hilarious they're always anxious to please, 00:38:37.169 --> 00:38:39.329 right? They'll be like, oh, that's a great insight, right? 00:38:39.589 --> 00:38:42.289 So you can have these conversations. And honestly, I've been spending a lot 00:38:42.289 --> 00:38:47.369 of time, perhaps too much time, really having these conversations to just see 00:38:47.369 --> 00:38:48.869 how things are connecting. 00:38:49.049 --> 00:38:52.149 And I've learned a lot, and I'm pretty deep into it, right? So the more, 00:38:52.409 --> 00:38:55.109 I would just say, again, the only limitation is yourself. 00:38:55.309 --> 00:38:58.949 So just be curious and get out there and learn. 00:38:59.169 --> 00:39:02.869 And any questions that you have, don't be like, you know, say, 00:39:02.949 --> 00:39:03.669 well, I can't be answered. 00:39:03.749 --> 00:39:08.289 No, just go talk to the chatbot and ask and ask the hard questions and, 00:39:08.309 --> 00:39:10.269 you know, get into it and just get smarter on it. 00:39:10.349 --> 00:39:13.169 And once, and once you, that's half the battle, right? Once you get kind of 00:39:13.169 --> 00:39:15.989 smarter, then you can kind of figure out what does it mean for me? 00:39:16.089 --> 00:39:19.749 You know, the, what's in it for me and then work with, you know. 00:39:20.289 --> 00:39:23.449 IT and your other colleagues at work to figure out, you know, great. 00:39:23.649 --> 00:39:27.989 How can we maybe do this in a certain, you know, area X, Y, Z in the business? Awesome. 00:39:28.449 --> 00:39:31.849 Well, thank you, Pierre, for your time today. I appreciate it. 00:39:31.929 --> 00:39:34.169 I'm sure our listeners will get a lot out of this session. 00:39:34.609 --> 00:39:39.909 And it's great to have such an open conversation with you about this. 00:39:40.049 --> 00:39:43.649 Again, Pierre Mitchell, he's the chief research officer at Spend Matters, 00:39:43.709 --> 00:39:45.269 which is now part of the Hackett Group. 00:39:45.629 --> 00:39:50.249 He's been covering procurement technology as long as procurement technology has been around. 00:39:50.609 --> 00:39:55.709 I think it's safe to say that. So he's really seen the advances that are out there. 00:39:55.889 --> 00:40:00.209 And, you know, it's great to have you on the show today. And all the best to you, Pierre. 00:40:00.749 --> 00:40:03.609 It's been great. Thank you. Yeah, thanks, Greg. I really enjoyed it. 00:40:03.609 --> 00:40:05.569 I hope the listeners get some value from it. 00:40:05.709 --> 00:40:08.309 And yeah, it was just nice to connect with you again. All right. 00:40:08.469 --> 00:40:09.369 Thank you, man. Take care.