[Transcript] Leveraged Supply Chains – Episode 11: EntProc: James Meads

Key Takeaways

Procurement technology has exploded. Yet many of the hardest direct-spend problems still live in email, spreadsheets, and ERP screens built decades ago.

In this episode of Leveraged Supply Chains, Andrew Stroup sits down with James Meads, founder of EntProc, host of The Procurement Software Podcast, and creator of the Procurement Tech Maps for SME, mid-market, and enterprise teams.

After 16 years inside corporate procurement across automotive and fast-moving consumer goods, James set out to give practitioners a vendor-neutral view of the software market. His maps now cover hundreds of tools, but the map also reveals where investment has lagged, especially across direct materials and supplier execution.

They get into:

• Why mid-market procurement teams should not copy an enterprise technology stack
• The white-collar waste hiding inside procurement, from supplier chasing to manual ERP updates
• Where investment has flowed across intake, orchestration, agentic AI, and niche direct-spend tools
• Why the ERP remains a powerful system of record but often fails as the user experience for supplier collaboration
• Why supplier portals struggle outside a supplier's largest customers, and why email continues to win
• How to separate useful AI from unnecessary complexity, including James's warning not to buy a Ferrari after just passing your driving test
• A practical modernization framework for a $100M manufacturer: understand the spend, eliminate waste, delegate lower-value work, automate the rest, and test the ROI

James also gives blunt lightning-round answers on the most overused term in procuretech, the metric procurement teams overvalue, why best-of-breed wins in the mid-market, and why more software founders should sell to the CFO.

If you lead procurement, supply chain, operations, or finance at a mid-market manufacturer or distributor, this conversation offers a practical way to decide what belongs in the ERP, what should sit around it, and where AI can create real leverage today.

Subscribe to Leveraged Supply Chains for more operator-level conversations on AI, supplier performance, procurement technology, and operational excellence.

Full Transcript

Andrew (00:06)
There are hundreds of procurement software tools. Someone has actually looked at all of them, categorized them, mapped them by company size, and made those maps freely available so that practitioners can navigate the market without getting sold to. That person is my guest today. He spent 16 years doing procurement from the inside, automotive and fast-moving consumer goods in the UK and Germany and around the world before he stepped back and built Entproc and the procurement software podcast.

He has looked at more procurement tools than almost anyone alive. And he will tell you the most important gap in the market are still unsolved in many ways, I'm your host, Andrew Stroup. This is LeveragedAn Supply Chains. James, welcome to the show. Before we get into the maps and the landscape, I'd love to share your arc. 16 years inside corporate procurement, and then you step out and build something for the whole ecosystem.

What was that moment or maybe even the slow realization that made you want to do that?

James Meads (01:01)
Yeah, quite an introduction. Thanks, Andrew. Pleasure to be here. So I guess the simple answer to that is that unless you're a large enterprise organization with a paid subscription to something like Gartner, which let's face it, most businesses don't have, there are very few open source resources that one can access that gives unbiased advice that's not sponsored.

on all of the different procurement tools out there. And that's really what I set out to do was to democratize access to that data. And yes, we're not going to be as detailed as an analyst firm that charges five or six figure sums for subscriptions, but fundamentally, the average procurement leader or CFO for that matter, if you're looking at more procure to pay software,

The typical problem they have is that they don't understand the breadth of tools available on the market. And every business is different. So is the advice that Gartner or SpendMatters or IDC or any of these analyst firms, is it applicable to your specific solution? Probably not. You're going to have to do your due diligence on the short list of providers that you come out with before you go out and buy some software.

So the angle I took it from was really people just want to know what's out there and what particular modules and functionalities each of these solutions have. And then the due diligence around integrations, UI, UX, looking at demos, quizzing the solutions architects and the salespeople as part of that sorting process will come later. And that will probably still be done in-house because

you're probably going to want to do that yourself rather than trust a third party consultancy to do it. So that's really in a nutshell what we set out to do. Obviously a lot of background in terms of how I got to that point, but that was really the problem that I saw and what I sought to solve with what we built with the tech maps and the database.

Andrew (02:57)
Love that. And I'm kind of curious, and maybe if we like zoom in one layer deeper during your time in kind of the tactical procurement days, where the specific observations on like the day to day around specific tools or cycles of procurement that led to like just how you're operating and ultimately, where you solve like this, you needed to zoom out. And this was like an obvious entry point for how you thought about sharing those learnings to other practitioners.

James Meads (03:23)
Yeah, it's a great question. this, so my sort of time goes back to 20, at the end of 2018 when I left corporate procurement. So this was really just before the big wave of VC money that went into procurement and supply chain tech roundabout sort of when COVID hit. But if I look back then, the biggest challenge with all of these technology providers, and I assume, and I also include a lot of the sort of invoicing and accounts payable tools in there as well.

on the, on the finance side was just, they weren't particularly user friendly and they were quite complex to implement. And it was very much, the owners was very much on the individual procurement practitioner or the supplier in the case of, know, invoicing platforms to, to learn how to use the tools and to register to use them and everything. and, you know, we'll talk about this a little bit later with, with, with what leverage has built, but one of the fundamental problems with that is that.

Procurement people are very, very busy firefighting because historically their function has not had much investment in technology. And it's typically quite an under-resourced department in every company. And likewise on the supplier side as well, if you've got 50 different customers and 50 different portals, guess what? Unless that customer is a key customer for you and is one of your A customers, probably ain't going to register for 50 different portals.

So that was the biggest challenge that I saw. was just, there were a lot of barriers to adoption and we didn't really have back then, I know it's changed a lot now, but we didn't really have back then the consumer grade, know, the Uber, the Airbnb, the Tinder level of functionality in B2B applications. You had ERP systems, which

then and also now are pretty clunky and difficult to use and are used as a system of record, but are not really used on the front end in most cases to run the business. And then you had a lot of these legacy source to pay suites. And I won't name them. guess your listeners are probably familiar with them. They just aren't particularly user-friendly. They do the job in most cases, but they're very rigid and not particularly user-friendly, both for the buy side and for the sell side.

Andrew (05:26)
I it. I also do think that there's an entire probably other episode that we can talk about the narrative and story arc of how VC money has shaped the landscape and also I think where maybe the industry was and where there are assumptions of where they could be and the time horizon it takes to get there. Separate conversation, I know, but that's something I know we've chatted about as well. There is something I think you've talked about though in the landscape of I think the phrase of an entrepreneurial procurement and what you've thought about what that means and...

where that industry falls short in some capacities and how you've thought about it as well. And, you your practices would love to talk through how you've thought about that phrase and how you think about helping the industry along the journey in your own work.

James Meads (06:10)
Yeah, so it really came from two very distinct observations that I saw back when I was a practitioner. So when you're an entrepreneur or a business owner, there are two things that you have to be pretty adept at to be able to survive, or three things, I guess, if you include sales. the two biggest things that I saw was you need to be very, very careful about how you guard your time and what you spend your time on doing.

And just, you know, time is your most important commodity when you're, especially if you're a solopreneur, but also with any, if you're any sort of small business owner, your time is your most important commodity and you have to get used to saying no to things without feeling guilty. And I think the other facet of that is you have to be really good at marketing or communication, or you need someone on your team that's very, very good at that. you have a small business or a product that doesn't have a well-known brand.

And I'll then add the third part on, which is kind of related to marketing is that you have to be halfway decent at sales to get any customers. And I think when I take all three of those and apply it to procurement, we suck at communication. We're not very good at time management or we don't jealously guard our time as much as maybe business owners do because we're constantly getting sucked into what I call white collar waste or administrative waste.

You know, you've got procurement leaders that are on six figure salaries that spend a third of their time doing stuff that a $20 an hour admin assistant that you could hire on Upwork could do. But procurement just doesn't think like that. that same process with the same amount of waste was in a factory and you had your state of the art machine just bashing out any sort of commodity product rather than making something that's precision made.

a precision made product at the end of it, then you'd have lean consultants crawling all over it, trying to figure out how to get the most machine efficiency out of that factory process. But in procurement, because it's a white collar process and in accounts payable further downstream as well, or even in sort of day to day tactical logistics and supply chain management, I would say the same thing applies. There's an awful lot of administrative waste that just doesn't get tracked.

I think part of entrepreneurial procurement is looking at your annual salary if you're in a corporate job, and then dividing that by let's say 225 working days a year, if we assume, you know, five weeks of vacation, yes, I know I'm European, plus public holidays, and then divide that by an eight hour day. What do you get? You know, if that's 100,000 a year,

then your effective hourly rate is $55.55, if my math is correct. You shouldn't be paying someone $55.55 an hour to be copying and pasting things out of email into an SAP system or to be picking up the phone, chasing up suppliers saying, are my parts? This is all the sort of stuff that a decent offshore business processing, business process outsourcing function or a good

admin assistant or junior buyer could be doing. So when procurement fired all of their admin people in the 2008 financial crisis, and they never brought them back when the economy improved, it's really a false economy because you're getting skilled knowledge workers doing that work instead of what they should be doing. And I think a lot of people in a corporate mindset don't really, their brain doesn't really tick that way. And then the other part, the whole marketing and communication thing,

The example that I will use is HR. HR over the past 20, 25 years went from being sort of the personnel department, very tactical, very administrative function, and they've rebranded themselves as being chief people officers at the upper echelons or as being a much more strategic function. And in a market where you've got...

you know, Steve Jobs, a small team of A players will run rings around a bigger team of B and C players. I don't know the exact words of the quote, but it was something like that. Talent is important. And sometimes you have to pay more to get the best people. And I think HR, at least in tech and high growth companies gets that. And I would apply the same to procurement that if we could rebrand ourselves as being value orchestrators and not just focused on

firefighting operational issues and delivering sort of more tactical savings, you know, three bids and I'll negotiate the best bid down by two or 3%. And that's my saving. If we could brand ourselves away from that and be seen more as as chief value officers or value orchestrators, rather than the spend police or the compliance police, I'm deeply convinced that we can go down the same route as HR in terms of being seen as a more

strategic function. Are there still very tactical HR departments? Are there a lot of jobs worth still in HR? Yeah, of course there are. And that will always happen in procurement as well. You're not going to go from being an operational compliance based function to an entrepreneurial strategic department overnight. But I'm certainly seeing that change happen in more forward thinking businesses, especially in high growth tech companies that need

that need processing governance more than they need savings because they're growing that quick at breakneck speed that, know, procurement's just playing catch up. So it's very much dependent on the context, but I do think that we can make that journey in a large percentage of organizations.

Andrew (11:27)
Absolutely. I love that because I think one of the pieces I've heard very much

stated criticality of that role to the top line growth and also the board objectives, which I think is also just including them in the room, a whole different also thread that we'll get into. you know, it is fascinating because I think that you have spent time on the inside, but also in that process of moving more on the outside boundary layer of taking a macro perspective, you started mapping the entire process itself. And I think that's I would love to dive in on talking and walking me through.

the procurement tech maps and actually how you've approached them. And specifically, think maybe for people who haven't seen them before, what are they and how have you used them in that tool of helping guide the landscape?

James Meads (12:13)
Yes. So in its simplest form, we divided the tech maps into three. So we have one for SMB, SME, one for mid-market and one which we're actually going to discontinue because it's not really my ICP in terms of who I serve for enterprise teams. And the reason why we did that was fundamentally a piece of enterprise procurement technology like

you know, for example, Cooper or SAP Ariba are probably the most well-known examples. They're a perfectly good fit for a large corporate organization that has a center of excellence team and has got the resources to go out and train and educate their stakeholders as to how to use what is quite a complex piece of technology. And likewise, on the more downstream payment side,

something like that can handle a large degree of complexity if you're operating in multiple jurisdictions and you've got invoicing mandates and you've got different sales tax setups that you have to include in there. Brazil, for example, has different sales tax and different ways of invoicing in each federal state. Will a piece of Procure2Pay software designed for small and medium-sized businesses be able to do that?

Probably not unless it's built for the Brazilian market. So I do see where they have a fit, but fundamentally when you look at Magic Quant or Quadrants or 50 to No or all of these other best of lists that analyst firms put out there, they're all built for big corporate procurement teams. And if you're a procurement leader managing a team of 10 procurement staff,

those solutions just aren't fit for the context that you're going to be using that in. If you're a $500 million business, or even more so if you're 100, $200 million business, those types of solutions just aren't fit for function. So that's why we divided it into those three. And then we have different subcategories or primary categories, as I call them. And I often get solution providers complaining to me saying, why can't you put our logo in more than one category?

But to make sure that it's not too busy and too messy on there, we limit the number of times that a logo is shown to one specific primary category. And yes, when you go onto the website or when you go on into the database, we have tags for different modules and functionality that they have. it is very much a very superficial view of the landscape that we project on the tech maps.

But it's a good starting place for someone that literally has no clue who is out there apart from the big brand names to understand, well, okay, who else can I look at for a source to pay solution apart from SAP or Oracle? And that gives the context of what's out there first and foremost. Should you be using that as your only piece of research? Absolutely not. What it really serves...

is that initial sort of very top of funnel RFI request for information stage where you're thinking, how do I find out who's out there without using Google or chat GPT or without just being reliant on them?

Andrew (15:15)
I

absolutely, you and I love that you framed in the sense of like not for really enterprise and the focus, you know, as you know, a lot of the big analyst firms are producing reports for the people who fund them and or are their customers are typically the large enterprises who can afford that level of research and diligence. And so I think having a wedge for the SME, SMBs and mid market is fantastic. And it's also interesting, I think, because

And I'm curious your take on this. how, how do you end up, how do, how does a SaaS company or a solution provider end up on the maps themselves? And I know you segment them by specificity and focus, which I think is exceptional. Curious how, how that gets shaped in your process, but also even in that process, what have you seen as far as categories have been changing the fastest, especially with the wave of AI in the past, you 12 to 18 months.

James Meads (16:07)
Yes, so we have some basic criteria. We don't do deep due diligence on every solution provider. Otherwise, we'd have to charge for the research and that's kind of the whole point that we don't want to do that. the fundamental criteria that we apply is that a company has to, number one, they have to have a fully fledged English language website, because if all of the database is English and the website is only in Japanese or in German or whatever, then

They're obviously not focusing on the market where typically we get our web traffic from. So yes, there are a bunch of perfectly good Chinese, Japanese, Indian solutions that are out there that cases to those huge domestic markets, but we don't list them because the average person in Tokyo is not going to search on an English language website looking for ProcureTech for something that solves a problem primarily in their home market. So that's number one.

Assuming that you've got an English language website, then we look at two other factors. Are they a standalone SaaS product that is marketing towards procurement or at least has a sub page with procurement use cases for the software on there? Because we don't want to list consulting firms that make the use of their software contingent upon the end customer taking out their consulting services.

We have consultancies that have spun off software products and those software products we've got listed on there. But if the use of the software is contingent on taking services, we typically don't list them. And then the third main criteria is that they have to be beyond MVP. So we want to see evidence that they've got paying customers. Pilot customers isn't a paying customer. Pilot customer basically is just validation of an MVP. And they're essentially

Pilot customer is almost a partner for a software firm in that they're getting early access to the product. in an exchange for that, they get to use the software for free or to have an influence over the roadmap. But I wouldn't necessarily call them a customer in the classical sense. So they're the main three. I we will look at other things as well. But typically, if you pass those three criteria, we will give serious consideration and probably list you.

Andrew (18:16)
Makes sense. And I like that there's rigor around this. It's interesting because as I dived in and did my homework ahead of the recording, there's definitely this very fascinating full picture that emerges. But I'm also curious because I think there's a sub layer on the analysis and the market and the map itself is that there are definitely places where there has been clear investment and focus by

the VC world landscape, private equity, people in their time and entrepreneurs, and subsequently where there hasn't been. I think as you have worked on the maps and thought about the different categories, curious like where there's been evidence that there's been very clear investment or the lack thereof. And very specifically, you know, I think if you look at the past, maybe 10 years or so, there's been a heavy amount of investment in indirect spend.

But yet then this material, direct investment has been trying to catch up. And so very curious as you've spent the time talking to other entrepreneurs and looking at tech and mapping it, how you've seen that evolve and any conclusions or hypothesis around that that may be able to inform the why behind.

James Meads (19:24)
Yeah, I would certainly broadly agree with the statement that there's been less investment into procurement tech that focuses on direct spend. Yeah, I think that's broadly true. The biggest area where I've seen investment, if I look specifically over the past sort of two to three years, there are a couple of areas where I've seen a lot of investment going in. The most obvious one is intake and orchestration. And I think that's really just a...

a natural evolution of the fact that a lot of the legacy source to pay technology is pretty rigid and is pretty old hat. And in some cases is written on a 20 year code base. that's, I'm not really surprised that someone like originally it was zip and now there's a bunch of them holding onto their coattails managed to get a lot of funding and exploded onto that market.

hats off to them, they've also got a very talented sales and marketing team as well. The other area, perhaps unsurprisingly, is the whole agentic AI piece. So companies that are coming into the sphere that are AI native and have been built with AI from the ground up. A lot of existing procurement tech companies have had to go in and retrofit their AI. Some have done a better job than others.

But there've been a lot, particularly over the last couple of years that are AI native that have been getting some traction. Probably the most famous one is, is Leo that just got 30 million in funding, which is again, automated, essentially automated or automated source to pay or source to delivery. Again, one where there's a massive need for in a lot of companies that are still doing parts of that manually.

The other area that I'm seeing investment in is more niche specific tools. So tools for specific types of spend. And then I would divide that into two different into two specific buckets, specifically where I'm seeing investment. So one of them is for software. So there are a lot of tools around managing SaaS stack and getting control of software spend and market Intel on renegotiation for renewals.

And then the other area, which is kind of the area that you play in, is the area that focuses on specific types of direct spend. There's a couple of solutions out there that manage different types of raw materials really well. There are solutions out there that are managing large complex projects for industries like oil and gas, perhaps not direct spend, but certainly not the indirect long tail type of three bids and a buy type of spend.

And then there are also tools like what leverage is doing that are getting traction that sort of gives visibility over the end-to-end supply chain as well. And that doesn't surprise me. But it's also, I think, a much harder sell in some cases, because the logical answer from anyone that's not in the know would be, well, yeah, that...

ERPs are for managing raw materials and direct spend and inventory. So why not just use that? Whereas I think on, if you're managing SaaS or if you're, if you're, built, if you're bringing in a tool to manage the long tail of indirect spend, I think it's, it's more obvious to everyone involved that the existing EIP system and tech stack just doesn't serve that space well. Whereas for direct spend, think it's, it's a, it's a slightly more difficult.

sales argument and marketing conversations to be had around, well, why do you need this type of technology? We both get it, but I think the wider community out there perhaps struggles a little bit with that.

Andrew (22:53)
Absolutely. And I appreciate the TIA because I know that's actually the other thing I wanted to chat about during our time is that I think fundamentally or from a design perspective, the original intent for an ERP was to have the direct spend live there and in all capacities. But I think both as you as a practitioner and me as a provider and spending the time with different types of companies at the SME, SMB level and mid-market,

that is not always the case. And there's a definitely a different relationship with the ERPs and how supplier management gets operated and what do the tactical individuals like the buyers do. Very curious on your perspective and experience around that because I know that you've seen that from the inside but also the outside as well as what whether it's coping mechanisms or hacks have come about that maybe are to your point.

less obvious, but very painful and would love just your perspective in that category.

James Meads (23:51)
Yeah, it's a great question. I think let's be fair to ERP systems. mean, they're not the most user-friendly tools in the world. I think it's an understatement, but

Andrew (24:00)
Good.

James Meads (24:00)
they actually pack a pretty powerful punch if you set them up correctly. If you're using all of the features and functionality of something like SAP and you're doing it well, an ERP system manages your inventory. It can manage your quality management. It can manage your production planning.

it can even have some basic data around delivery performance as well. So if you've got all of that sitting in your ERP system, which then communicates with other systems as well around sort of logistics and distribution at end of the production process and the beginning of the consumer supply chain, then

it's going to be very difficult to convince a business to move all of that out of there and put it into another system. I think that's the, maybe uphill battle is the wrong word, but that's the challenge that anyone that's offering a point solution for any sort of direct spend or area of the supply chain is going to have because the ERP is tracking everything holistically. And yes, it might be difficult to access that information if you

If you don't have access or you don't know the magic four digit code to access what you're trying to do and then deal with the screens on there that look like they were straight out of the 1990s because they were. But

Andrew (25:11)
or earlier.

James Meads (25:13)
then how do you then marry that with the situation that all of that information is already, all of that data is already sat in the ERP. So how do you get both systems to speak to one another?

while giving the front end user the type of consumer grade UX that anyone under 35 is going to expect if they've grown up using an iPhone.

Andrew (25:34)
Absolutely. You know, I do think to your point around the credit given for the ERP, but yet maybe the reality of where the boundary layer is, there's like two fundamental phrases that we always think about with our team. It's that one, you know, every everyone has an ERP, but they all hate it for different reasons. And it's very fascinating around like how it kind of serves as

James Meads (25:51)
Yeah.

Andrew (25:53)
a hybrid bike. You know, ultimately, it's like hybrid bike is good at it can do everything. It's not really great at any one of those things. But ultimately, I think it's also what you're kind of describing as the core

elements, maybe call it the data elements and the structure and the connectivity is there, I think the real question is like, what's the boundary layer, right? Of how much do want to stretch it versus the focus of a point solution? And so that's interesting because I think, you know, obviously for us at Leverage, we have a very specific category around where that sits and how we think about it when it comes to supplier portals and why they failed and the perspective of where the delta is on how to close that gap.

between the boundary layers of an ERP that operates the internal business versus the external layers of something like a supplier network. Why do you think something like supplier portals have consistently failed to drive adoption? I know you alluded to a little bit around being on the supplier and customers, but from the buying side and an honest perspective, there definitely is a friction that I think people ignored for a while, and especially analysts and...

know, people who went to business school and made assumptions about the market would love to get your take on that topic and what you've seen work and not work.

James Meads (27:03)
Yeah, it's kind of the reason why, why if I have, I I, I advise and work with, with a lot of procurement, with procurement tech startups as part of what I do. And, and often if I'm working with them on something that'll ask me to join a Slack channel and it's, it's, it's an effort for me to do it because I don't use it day to day. I'll do it because they're a client, you know, and I want to, I want to keep them happy, but, fundamentally I, it's not a, it's not a tool that I use every day because I don't, I don't have a marketing team or an opera or, or a devs team to.

to need to communicate with on a day-to-day basis. So it's kind of the same thing for vendor portals that if you're a supplier, you might use vendor portals for all of your A customers, but for anyone other than that, it's a nuisance and it's much easier to deal in email or depending on the market or maybe in Microsoft Teams or in WhatsApp, depending on what culturally the natural method of communication is.

Yeah, you've got to meet your customer where they're at. And in most cases, especially in direct spend, that's probably going to be the email inbox or secondary. If they allow collaboration with external users, it's going to be Microsoft Teams. They're probably not using WhatsApp or Slack or any of the other ClickUp or any of these project management tools.

If you bear that in mind, then you've got to make it easier for easy for them to just click from their inbox and have that automatically update in whatever portal that you're using.

Andrew (28:29)
Yeah, absolutely. think there, and I know we've talked about this offline, is that there's this entire approach that I think was just inherently flawed that made the assumption that you can bend people to their will. And I think if you look at maybe the investment trends on the indirect side, that probably was a little bit more true when it was more transactional basis and is more componentized. But to your point, outside of either direction, your top five suppliers, your top five customers, the remaining whatever may be 80,

800, 8,000 suppliers that are part of your actual core supply chain for actual P &L growth, it's a very different relationship. And I think that's a very true aspect that gets overlooked when it comes to software as a software, but the execution delivery for the humans matter a lot. And sometimes it's not always just software when it comes to like a platform. And so I know we'll dive into that a little bit later, but obviously the pertinent conversation, at least where we are in April, 2026 is

Everyone's answer to all things right now is AI. And that includes supply chain procurement, especially in an industry that has felt very manual and kind of delayed, I think in the technology adoption curve. I'm curious from your perspective, when a vendor, know, software vendor or consultant says, Hey, you know, here's AI, we've got AI. What do you've been seeing under the hood, the reality match on that one and where it is actually, you know, AI, the actual truth sit.

when you've seen it actually be impactful to a team that's deploying.

James Meads (29:56)
Yeah, there's a whole different number of ways I could approach this, but I think the good place to start would be if you're not using any AI right now, then most of the use cases where you will naturally think, we want to get AI in to do that, you could probably fix by simple robotic process automation. If you just want to have workflow automations,

to be able to eliminate a lot of the manual work, or if you want to be able to get data from one system to populate into another. For the former, you can use RPA, for the latter, you can use a simple AI automation. It's a task-based and you don't even need software these days. You could use Claude to do that as long as you know what you're doing.

Certainly if you're a novice, you're probably not going to figure that out. But if you've got someone in your IT team that's pretty adept with Claude, you could get it to do that. My first piece of advice would be if you've just passed your driving test, you wouldn't go out and buy a Ferrari. And using that same analogy, you don't need to have a complex agentic AI solution to solve simple tasks. So look at where your biggest bottlenecks are and then do the classic

consultants two by two diagram and look at what would go in that top right hand corner of high impact, but low cost and low disruption. And I think you'll find that simple automations will fix a lot of the administrative waste that you're seeing within your business. Where do I see the biggest impact of AI? I think part of it is having visibility where I previously didn't.

So I have a client right now that's developing a solution that will enable procurement to get earlier involvement in projects because it's able to scan and read email across the whole organization. And one of the biggest challenges anyone in procurement will tell you is that especially in big projects, they often get brought in too late when a lot of the key decisions have already been made, especially in anything that involves building a piece of machinery or designing a new part.

If engineering had already done that part and they only bring you in once they've got a scope to go out to the supplier, you've missed a big part of the potential value that you can deliver by not being part of the whole design to cost conversation. So I think visibility is one. And I think the second one is AI can do a lot of the heavy lifting and a lot of the grunt work. And I don't mean by just simple tactical automations here. I'm thinking.

research for a category strategy or being able to write 80 % of what you'd need to have in an RFP if you can train it well enough. And again, if you know how to use Claude skills, you could pretty easily set up an RFP writer that is cognizant of the context which you want to write the RFP for and your brand voice and the key clauses that you have to come in regardless of what it is that you're buying.

So that would historically have been probably taken the best part of a day to write an RFP. If you're, if you're doing it from scratch. Now, if you've got, if you, if you've got an agent that can write it for you, you can get, you can get something written in 10 minutes and then it will maybe take you another hour to sort of firm that up and correct anything that it might've hallucinated or not, not quite described as well as you would have wanted it to. So I think a lot of the.

not just tactical work, but a lot of the grunt work that went into a lot of strategic procurement activity can now be not fully automated, but you can get AI to do a lot of the hard work that was previously done manually.

Andrew (33:31)
Absolutely. I think the phrase that I've heard recently has been the wave of AI collar, you know, from the blue collar to the white collar and now thinking about what does it look like to work with AI and the capacity, but highly led, especially I think in supply chain and procurement, I have yet to see a world where it is, although, you know, idealically talked about, just set it and forget it. And I don't think that that is even possible because I think one of the reasons why

technology has been slower to adopt in some categories in this industry is that you need humans because there are humans working with other humans to move physical things in the physical world. And then ultimately is a boundary layer that I don't think always exists in other industries, which is fascinating to think about, you know, the understanding of the why. But very much appreciate that perspective too, because I think there is, I've seen massive acceleration in some categories, but also massive trip balls.

where AI is being applied that you can't just assume that'll solve the problem or the equation for you. Now, go ahead.

James Meads (34:28)
I I think there are areas

where you can set it and forget it almost. I I think if

Andrew (34:33)
Mm.

James Meads (34:34)
you're in an organization where you had a policy that, let's just say anything less than $5,000 in spend, procurement didn't touch it because there was no capacity. If you can bring in an AI negotiation bot to take those three bids and just email or email the supplier or chat with them using an AI agent or even use a voice agent to call them.

and just try and squeeze one or 2 % out of the salesperson for spend that would otherwise have not been touched. I think you can pretty much set and forget that because it was spend that procurement wasn't touching anyway. it's kind of, you're starting from a pretty low barrier in that anything is better than nothing. So, you know, as long as it's not

Andrew (35:10)
Mm.

James Meads (35:11)
saying anything, anything racist or offensive or anything to your suppliers, the way that you've trained it, then, you know, I think you can pretty much...

set it and forget it in that sense, as long as you've got some sort of human in the loop that's checking what it's doing from time to time. But yeah, other than that, yeah, I do think for anything more complex than that where procurement is involved and you're using AI just to speed up the process, then yes, some sort of human involvement, certainly for now is still going to be necessary.

Andrew (35:38)
Yeah, absolutely. whether it is before you take the action or even the observability and triage after, it's very interesting to see where humans do get inserted in the entire workflow and the rack structures that get developed in that capacity. Well, I think on the backside of this conversation, I'd love to get a little more, I guess, tactical or practical, especially for the audience who's typically a supply chain director at a $100 million manufacturer distributor.

If they came to you and said, like, I know I need to modernize. have a budget. have 90 days. Where would you start and how would you think about, you know, investing, maybe over-investing or under-investing some categories and where do you think best of breed ultimately sits and walk through that logic as like a teaser? Cause I know this is ultimately what you do on a day-to-day basis.

James Meads (36:22)
If you'd have asked me this two or three years ago, I would have said the first thing that you need to do is to have some basic spend analytics software in there. So, 100 million is probably about the cusp of that because you could make an argument that you could feasibly still do that in a spreadsheet. Anything much above that and you're going to need spend analytics. Nowadays, you can...

And I've seen someone that's done it actually, you can write a prompt and if you know what you're doing and if you've got someone that can maybe code a little bit of Python to help you with it, you can get some decent spend analytics using off the shelf LLMs and AI tools now. So I think that's kind of going away for the lower end of the market. If you just need a very high level synopsis of what you're spending with whom and where there are some low hanging fruits, then

I don't necessarily think at the lower end of the market now that you need spend analytics software, but you certainly need to get a grasp of what am I spending with who and do I have any single or sole source instances that could be a high risk, high impact on the supply chain. As well as, know, do I have on the other side of that, do I have categories where I've got seven or eight suppliers and I could consolidate that down to one or two and make some quick savings.

The next thing that I would do would be just to automate as much as possible of all of my transactional and administrative procurement processes. So I'd follow the eliminate, delegate, and automate model if it's something that's just a completely nonsensical process like...

the head of procurement has to sign off every travel request. I would just get rid of it and treat my employees like adults and give them each a travel budget that they can manage themselves. Now that's just managerialism, pure and simple. It's just waste. It's justifying someone's existence. Then if I've got team members that are more senior or more capable than others, I would look at what are they doing that doesn't justify their salary that could be delegated to...

to a recent graduate or to a junior buyer or to a more admin focused person. And then what doesn't even need a person at all, what can be automated? And once you get down into the nitty gritty of that, then it's just a simple cost versus benefit analysis of what's this process costing us in labor or in waste in the case of, if it's causing supply chain issues that cause factory stoppages or a

or late deliveries to customers? How is it impacting our factory efficiency or our profits if it's causing shutdowns versus what's the cost of the software and what's the estimated ROI? It sounds simple and it is simple. And you can get into the weeds a little bit of how theoretical is that calculation going to be? But the process of how you get there is pretty straightforward.

Andrew (39:05)
Yeah, think those are sometimes the best advice, right? It's a framework that's simple, but requires a lot of rigor to dive into. And I know that's where you spend a lot of time, which is a high impact. Well, I know we've covered a lot of ground. I always do like to wrap up these episodes with a little of a lightning round as well. So if you're ready, we'd love to fire off some questions for you, quick reactions to these and we can go from there. Most overused term in ProcureTech right

James Meads (39:29)
AI.

Andrew (39:30)
Okay, perfect. One metric procurement teams obsess over that doesn't actually tell them much.

James Meads (39:35)
savings.

Andrew (39:36)
Hmm best of breed or sweet if you just had to pick.

James Meads (39:40)
If we're talking mid-market, I would say best to breed.

Andrew (39:42)
And then ultimately, what would you tell a software founder or an operator building procurement today that most of them get wrong?

James Meads (39:50)
Hmm, that one's a really good one. Two things, would have, if not directly on my team, I would have a fractional advisor that understood both procurement and the procurement tech market to some extent. Because a lot of those companies don't have founders that come from this space. And the second one, perhaps a bit more controversial would be, depending on what type of tech it is,

I would say sell to the CFO rather than procurement first of all, because they're the decision maker. Procurement tends to have a lot of grandiose titles in organizations, but the reality is they spend other people's money and they don't have a budget for much beyond just travel and training and a bit of consultancy.

Andrew (40:28)
Yeah, absolutely. Very pertinent on the CFO side. All right, final question. What advice would you give your younger self when it comes to procurement in this industry?

James Meads (40:37)
Yeah, learn sales and marketing and communication. I think that's one thing. That's one thing. I'll hold my hands up and say that I was not particularly good at it in my practitioner days. I think procurement professionals generally are pretty poor marketers and are not particularly good at lucid communication.

you know, simple, clean, easy to understand communication and influence. And if you look at some of the best communicators out there, know, procurement could learn a lot by just being put into a room with sales and marketing for a day and told to talk to each other. You don't need to, you don't need, it doesn't cost much to do it. Just, you know, rent a hotel conference room and give them, give them food and drink. But other than that, it's a pretty cheap way to move the needle.

And I think it's one that often gets forgotten.

Andrew (41:28)
100%. Love that advice. Well, James, this has been great. Thank you so much for walking us through what the maps show and where software investment went and where it didn't and what a mid-market procurement team should actually be doing to close the visibility gap on their direct spend. For the listeners, if this episode made you look at your own procurement setup and wonder whether your ERP actually knows what's happening with your suppliers share it with your ops and supply chain teams. This is

the kind of conversation that reframes how people think about the problem. You can find James at Entproc.com, E-N-T-P-R-O-C.com on the Procurement Software Podcast and on LinkedIn, which we'll put into the notes of the show. The procurement tech maps are free, And then obviously follow Leveraged Supply Chains wherever you listen and leave a quick rating review. It helps other operators and builders find the show. I'm Andrew Stroup. Thanks for listening and I'll see you next time.