00:00:08 Discussing rogue pricing strategies in the supply chain industry.
00:00:26 Complexity of pricing in supply chain software due to infrequent purchases.
00:02:05 Single license pricing strategy, its drawbacks, and the “take the money and run” problem.
00:05:38 The unrealistic nature of trying out supply chain software for free.
00:07:00 The psychological trap of committing to gratis supply chain software and potential hidden costs.
00:08:00 Public pricing and negotiation in enterprise vendor space.
00:09:33 Taboo surrounding discussing vendor pricing.
00:11:54 Pay-as-you-go pricing models and their challenges.
00:14:19 Cloud vendor pricing issues and lack of competitiveness over time.
00:15:16 Misaligned incentives when charging for man-days in integrations.
00:18:57 The asymmetric relationship between vendor and company in success fees.
00:21:06 Charging per user for supply chain software: low fees vs high-end professional software.
00:23:01 Companies optimizing the number of users and vendors’ expectations.
00:25:00 Challenges of per-user pricing in supply chain software.
00:25:46 Discussing various pricing strategies and their shortcomings.
00:26:27 The process of elimination that led to the flat monthly fee.
00:28:02 Benefits of the flat monthly fee and its impact on incentives.
00:32:07 The importance of trust and performance in maintaining client relationships.


In an interview, Joannes Vermorel, the founder of supply chain optimization software company Lokad, discusses the challenges of pricing strategies in the supply chain industry. Vermorel explains that the “take the money and run” mentality from vendors is a problem, as it leads to a lack of incentive to improve products. He also notes that free trials are not feasible for supply chain software due to its complexity. The conversation covers pay-as-you-go pricing models, customization, and the flaws of success fees. Vermorel highlights Lokad’s pricing model, a flat monthly fee that aligns the company’s incentives with the client’s, as the “zen of supply chain.”

Extended Summary

In this interview, Kieran Chandler, the host, and Joannes Vermorel, founder of Lokad, a software company specializing in supply chain optimization, discuss pricing strategies in the supply chain industry. The conversation focuses specifically on pricing strategies related to supply chain software. Vermorel notes that these operations are done very infrequently, and the chances are that whoever is doing them is actually doing it for the first time in their career. Thus, they may not have a good understanding of pricing strategies. The software vendor has a lot of experience in selling software, closing something like one client a week, making it routine for them. Chandler then asks about single-license pricing, where a company buys the code, and everything is owned by them. Vermorel notes that this approach was popular in the 80s and early 90s because of accounting tricks and simplicity. The company can put the license as an asset in their book. However, the money has already been spent, and reselling the license is challenging due to contract restrictions. Vermorel suggests that single-license pricing is not an ideal pricing strategy in the supply chain industry.

Vermorel believes that the biggest problem in the software industry is the license problem, which leads to a “take the money and run” mentality from vendors. Once a company pays for a license, there is little incentive for the vendor to keep improving the product. The host notes that while companies are willing to pay a one-time fee for purchases like buildings, software licenses are different because they require ongoing improvement.

The conversation then turns to the common practice of companies trying out software for free before deciding to purchase it. Vermorel believes that this is not realistic for supply chain software because it is a complex distributed system that is difficult to test. Even testing a new warehouse management system for one warehouse can take months and require significant effort to retrain operators. Some vendors offer free trials, but even then, companies must invest significant time and effort to deploy the software. Vermorel prefers the term “gratis” instead of “free” to emphasize that it is not actually free.

The interview highlights the challenges of software licensing in the supply chain industry, where it is not feasible to continually deploy new software every few years. Companies must carefully evaluate the costs and benefits of investing in software and must be aware of the potential drawbacks of free trials.

The founder discusses the challenges that enterprise software vendors face when it comes to pricing. Vermorel explains that enterprise software is not like PowerPoint, where there is a fixed license price. Instead, enterprise software vendors often engage in shenanigans, such as having an absurdly high public price that is negotiable. This puts the customer in a weak position to negotiate when they are committed to the software and have invested resources and time in it. Vermorel also notes that vendors often have clauses in their contracts that prohibit customers from publicly stating the price they paid for the solution. This is because it can create tension among employees who are paid minimum wage and may question why the company is investing so much in software instead of employee wages. Despite these challenges, Vermorel acknowledges that customers can gain some understanding of pricing through online forums, but ultimately, pricing in the enterprise software space remains taboo.

The discussion began with a question about whether managers would be willing to advertise poor negotiations with vendors. Vermorel explained that managers might not want to do so because it might seem like they are bad negotiators. They may also want to avoid revealing the extravagant cost of a project, which might not align with the company’s revenue. Vermorel suggested that it is easier for a CTO to say that a project was a success, rather than broadcasting the cost of the project.

The conversation then turned to pay-as-you-go pricing models, which Vermorel described as a nice middle ground between free trials and expensive licenses. However, he pointed out that this pricing model is very technical and disconnected from the added value from a supply chain perspective. If clients are charged based on bandwidth, CPU, SSD storage, and HDD storage, they might not understand how these factors relate to their problem of optimizing their supply chain. Vermorel believes that pay-as-you-go is only of interest to tech vendors who know what they are buying is a good price.

They discussed the challenges of pricing and customization in supply chain software. Vermorel explained that while technology has progressed faster than cloud vendor pricing, integrators often charge high fees for customization work. However, these fees can create a misalignment of interests between the integrator and the company, leading to slow productivity and a dysfunctional relationship. Vermorel suggested that success fees could be a solution, but he also acknowledged their flaws in practice. The stress of meeting KPIs, the asymmetry of risk, and the potential for gaming metrics can amplify distrust and create moral hazard. Vermorel cited instances where success fees caused disagreements between vendors and companies, leading to legal disputes. He believes that success fees are broken by design, and that other solutions need to be explored. Vermorel also discussed the value of Microsoft Excel as a tool, praising its ease of use and versatility in supply chain optimization.

They discuss how Lokad came up with its pricing strategy, which is a flat monthly fee that is all-inclusive. The fee is determined by the complexity of the supply chain problem and the scale of the company. Vermorel explains that the incentive for Lokad is to invest in the setup since the client must stay for two years to break even, which means that the client must remain profitable for Lokad. Vermorel states that this pricing model aligns Lokad’s incentives with the client’s because both parties share the pain if something goes wrong. The founder describes this pricing model as the “zen of supply chain” because Lokad must prove its performance every month to maintain the client’s trust, and the client can leave at any time, creating a high level of risk for Lokad. Vermorel notes that this model incentivizes Lokad to improve its software and lower its computing resources, making the maintenance part of the package as cheap as possible. Lokad’s clients trust their performance, which requires a high degree of trust developed over time, as the level of performance is not accidental and challenging to replicate. Chandler expresses concern that he can’t run anywhere.

Full Transcript

Kieran Chandler: Today we’re going to try and shed a little bit of light on some of these practices and discuss what are the particularly bad ones to keep an eye out for. So Joannes, why is it that pricing strategies vary so much in the supply chain industry?

Joannes Vermorel: My interest, and I think it relates a lot to Lokad, is specifically to the pricing of supply chain software. We have something that is much more infrequent for most companies. Companies don’t buy another piece of supply chain software every single week. Those operations are done very infrequently, maybe twice per decade or something. Chances are that whoever is doing them is actually doing it for the first time in their career. If it’s not the first time, the previous iteration was maybe a decade or two ago, and it’s not necessarily super relevant to what they are doing now. They might have completely changed in terms of position inside the company. So, it’s very interesting. You have a highly asymmetric position where the software vendor has a lot of experience at selling because they are hopefully closing something like one client a week, so for them it is literally routine. But for the company who is purchasing, this is a relatively atypical event. To be honest, depending on the type of software that you buy, it can be exceedingly diverse. I would say the world of enterprise software is super diverse, and supply chain is also very large on its own and has a lot of diversity inside it as well.

Kieran Chandler: Okay, so today we’re going to look at some of the more common pricing practices out there. If we start off with single license, this idea that you buy the code and everything is kind of owned by you. How well does that work?

Joannes Vermorel: That approach, I think, peaked during the 80s or early 90s. Selling a license has a benefit for the company, it’s an accounting trick. If you buy a software license, then you can potentially put it as an asset in your books. So, if you pay a million dollars to a software vendor and think of it as a license that you own, then you put this as an asset in your book, and accounting-wise it costs nothing. Obviously, the money has been spent and you’re not going to recover it. If you think that your license that you purchased for one million is worth one million, just try reselling it to get an accurate perspective on its actual market price. It’s possible that, contractually, you might buy a license but it might not be transferable or resellable. Bottom line, I think this approach peaked in popularity between the 80s and 90s partly because of this accounting trick and partly because it was very simple and it was actually the way things were purchased back then.

Kieran Chandler: So, Joannes, let’s talk about the software industry. What are some of the problems that you have seen in the software industry?

Joannes Vermorel: Generally, by companies, if you need to purchase a new building and not rent it, you expect to just pay a one-time fee and be done with it. If most of the things that a company acquires, it’s a one-time payment for something. So, in this respect, software, it was just aligning software with the rest. The problem is that it comes with many, many problems. And I think the biggest problem of all in the software industry with this license problem is that you have a massive take the money and run problem. And they take the money and run is that once you’ve paid a big license to a vendor, what incentive does this vendor have to actually keep improving the product?

Kieran Chandler: Right, I understand.

Joannes Vermorel: If you look at, for example, a company like Microsoft who is selling upgraded versions of, let’s say, Microsoft Excel every two years or three years, they resell a new version, but it’s supposed to be an improvement over the previous one. And people are sometimes, if the improvement is large enough, willing to upgrade. And they have been playing this game for decades. But it means that you are playing a game where you have enough innovation and progress so that every two or three years, you can have a new version. And people see that the new version is sufficiently better so that they buy. But in supply chain where it’s kind of broken is that deploying supply chain software is so expensive, so complex that you’re not going to redeploy an ERP or an MRP or WMS every three years. That would be crazy.

Kieran Chandler: Yeah, that makes sense.

Joannes Vermorel: If we look at maybe another practice that’s fairly common, particularly with the larger vendors, is this idea of maybe trying out a piece of software for free and then maybe a little bit further down the line deciding if it actually fits your company or not. It sounds like a very nice idea. It’s nice to try something out before you buy it. But does it really work in practice? The problem with supply chain software is that supply chain is a complex distributed system, you know, distributed many locations, many people, many processes. So the idea that you can test anything is, I believe, not realistic. As soon as it would be like saying, can you test out a double WMS, you know, warehouse management system? Well, you know, you have only so many warehouses. If you have a very large company, you might have 50 warehouses, but each one of them tends to be big. So even testing a new WMS for one warehouse, it’s going to be a massive effort. It’s going to be literally months of efforts. You will have to retrain probably dozens, if not a hundred or more operators. So, you see, the idea that you can have anything for free is, is, I would say, very odd. And the problem is that some vendors, especially those who are selling licenses, they know that even if they give you a license for free, a free trial, for free as in preview, not open source, you will have to invest a lot of time and effort to just deploy your gratis software. I prefer the term gratis. And then the idea is that once you’ve invested so much, there is it’s very hard for people psychologically to call it a day and say, okay, it’s not.

Kieran Chandler: Working on supply chain costs and not paying the license, you know, companies are committed to the initiatives they’ve started, and there you can end up in a very bad situation where you’re committed to a piece of software. But because it was gratis, you’ve not really negotiated the price yet.

Joannes Vermorel: So, and this is, remember this is not like PowerPoint where there is a license price where it’s like 100 for the license per user, and this is it, where it’s super simple. We’re talking about enterprise software, so you will always find shenanigans where suddenly the vendor will decide that the price can be something different.

And one very simple way to do that, by the way, that’s what some enterprise vendors are doing, is to have a very simple public price that is absurd. So obviously, you know, okay, you have your public pricing that is absurdly high. So the reality is that when you, after one year of free trial with this vendor, you realize that the price is absolutely high. And by the way, the vendor never lied to you; the public price was public, no problem. It was just absurd that it was public, your fault.

So now, the real price that you’re going to pay is going to be a negotiation, and it can be, you know, really a fraction, only a tenth of what the public price would be. But the problem is, at your point where you have, you’re in a very, very weak position to negotiate because you’ve invested so much. The vendor knows that you have invested so much, and they know that you have also not only the amount of resources but also the opportunity cost. You know, you’re one year later, so you’ve kind of lost all this time to do something else, and you’re now very, very late.

And so, changing a vendor because you refuse this deal that is going to be under negotiation, I mean, it’s very difficult. And thus, you can, as a vendor, get away charging fees that would have been almost unthinkable for the company if it had been the starting point of the relationship of discussing that.

Kieran Chandler: But you’d say there’s kind of these shenanigans going on, but at the end of the day, our customers, they talk on forums, and they already get kind of an idea about pricing before they buy. So surely, people already kind of have this understanding that that might happen before they even buy it, right?

Joannes Vermorel: No, it’s very taboo. I mean, it’s very funny; it’s actually there are many vendors who even put in their contract clauses that say that you are not allowed to publicly state the price. So they have an NDA that actually covers the price of the solution. That’s relatively frequent, but even when that is not in place, the incentives are not to communicate broadly about that. I mean, for many reasons.

First, imagine this is supply chain we are talking about. So, the people who are in the operations, those people are not paid a fortune. So it’s very hard if you have people that have minimum wages when you tell them that the software they’re using is worth millions. They may question whether the company is investing its money in the right direction, why not raise the wages a bit instead of pouring so much money on abstract stuff that doesn’t seem very real. But then, it’s not the only thing.

Kieran Chandler: As a manager, if you know that you’ve negotiated a poor deal with a vendor, are you really super willing to widely advertise that you’ve been a poor negotiator?

Joannes Vermorel: It’s not exactly ideal. If you are an IT director and you’ve deployed a new ERP, it’s a great success, but it costs, let’s say, 100 million dollars. It’s extravagant compared to the turnover of the company. The project might have gone well, but it’s still kind of extravagant. It’s much easier as the CTO of the company to just say, “We have done something great with our teams on this new project,” as opposed to broadcasting the fact that it costs a ridiculous amount of money.

Kieran Chandler: And then somewhere in between the middle ground of trying it out for free and paying a huge license fee, is this idea of pay-as-you-go, both in terms of mandates and computational cost. How well does that work? It seems like a nice middle ground.

Joannes Vermorel: The problem I see with this pricing is that it’s very technical and completely disconnected from the added value from a supply chain perspective. If I tell a client I’m going to charge them by bandwidth, CPU, SSD storage, and HDD storage, the client will ask, “How do those things even relate to the problem that I’m currently interested in, which is optimizing my supply chain?” I believe the pay-as-you-go in terms of computing resources is only of interest for tech vendors or people who have an absolute focus on the tech itself.

And the other thing is that, with pay-as-you-go for computing resources, even if you buy them at a fair price right now, what will make you think that this price will be anywhere fair five years from now? Most likely, it won’t. This was the big problem with IBM, where they were selling MIPS at a price point that was the cost of computing power in the early ’80s. Twenty years later, they have lowered their price a bit, but companies realize that they are paying for compute power that is less than a smartphone, at several dollars per second. It doesn’t make any sense.

By the way, this problem is also happening on the cloud, where typically cloud vendors start with a price for pay-as-you-go computing resources that is very aggressive at launch, and then they just don’t lower the price fast enough. So, people who adopt those technologies end up with something that, five years down the road, is not that competitive.

Kieran Chandler: Tech has progressed so much faster than the pricing of cloud vendors. Then you have the other problem when you go for people and that’s typically when you have a setup that’s a typical sort of setup that you have with integrators. So supply chain software is very complex, and thus you typically end up with a lot of customization. That’s what happens, you frequently end up paying licenses to software vendors, but then you have also the integrators who are doing a lot of work and they are typically paid in mandates. The big problem that I see with as soon as you start charging mandates is that you have a massive incentive for major core productivity. Every single thing needs to be done as slow as possible just because if this problem can be solved in one week, well, if you can solve it in one month, you charge one month. So you end up with this massive disalignment of interest between the integrator or the I.T company who is charging for every debt that they spend, and the company who basically earned value added value by like features delivered or improvement brought to the system, okay?

Joannes Vermorel: And one of the great ways to bring alignment between those two parties is through success fees. The idea is that if a piece of software really gives a company real added value, then both companies can profit from that, and that sounds very much to me like a win-win scenario for both parties, but is it in reality so? That’s a very good question.

Kieran Chandler: In theory, it works, but in practice, it doesn’t?

Joannes Vermorel: The problem with success fees is that on paper, it looks very good. You have a commitment with a vendor, we choose KPIs, and we say, “you know what, if you do something and you prove that you’re going to save, let’s say, 100, for example, the company is going to get 80, and the vendor is going to get 20”. That’s super cool. The problem is that I believe it’s completely dysfunctional, completely, completely dysfunctional in practice for a world series of reasons. The first thing is that it’s completely psychological, the sheer amount of stress that it puts in the relationship. You end up if you have a big success, you end up in a situation where potentially you’re going to put huge fees on the vendors, and although in theory, everybody would be rational, when you’re dealing with a large company and you can end up with $100 million of savings, the idea of paying a vendor $20 billion, you know it looks great on paper, but in practice, when you end up having to sign a check, and this $20 million is going to double the budget of the supply chain of the supply chain euro organization, not the expenditure and whatnot. I mean, it’s very, very different. Then you end up with also things where you’re terrorized where if you got the KPI wrong and you will get it wrong, you can end up paying absurd amounts of money to a third party on things that are unwarranted. The problem is that you can game the metrics; you know you choose a metric.

Kieran Chandler: And then you realize that the vendor can gain the metrics, so they can actually generate a very big number according to this KPI of success that you’ve decided. But actually, it’s not a win for the company, but it’s certainly a win for the vendor. So, it’s very difficult. You have a massive amount of distrust. In theory, the alignment the KPI would create, but in practice, it kind of amplifies any kind of distrust that was pre-existing.

Joannes Vermorel: And that’s not the only problem because you also have an asymmetry that is very hard to reconcile. If the vendor’s solution ends up making it worse, are you actually going to charge the vendor? For example, instead of being +100 million, it’s actually -100 million. It made things worse and generated a lot of cost. Is the vendor going to pay you? I’ve never seen that. So, you have something that is highly asymmetric, where you toss a coin, and if it’s heads, we both win; if it’s tails, you lose, and I don’t. There is this kind of moral hazard going on that is not good.

What happens, and I’ve seen it multiple times, is that with competitors that literally went bust with that, was year one, yes, you achieve great success. Everybody is happy. Year two, you end up with a company saying, “No way we are paying again this massive success fee. This is a new standard. So now, if you want to have another success, it should be the improvement compared to what we have now.” And the vendor says, “It’s so hard to even maintain this new level of performance that we don’t get anything. We have exhausted our low-hanging fruits, and now we have the new performance of the supply chain, but we can’t generate those 100 million dollars of extra savings every year.”

And so, you end up having a massive disagreement, and I’ve seen many companies even going to court with that just because it’s irreconcilable. At some point, you end up with a vendor that says, “Just to maintain this level of performance with regard to the KPI, I need to work a lot, but according to the success fee, I’m not paid.” So, in the end, it’s broken by design.

Kieran Chandler: Okay. One of the tools you kind of mentioned earlier, Microsoft Excel. Excel and one of the things that Microsoft does very well with Microsoft Office is this idea of charging per user, and they don’t charge very much, but when you’ve got a whole company of people using it, it soon kind of mounts up. And for me, that sounds like probably the fairest way to do it. How does that work if it was a supply chain piece of software?

Joannes Vermorel: So again, not very well. The reality is that when you charge per user, you have basically two ways of going at it. One is the Microsoft Office way, where you charge a very low fee, something like ten dollars per month, and basically, you kind of expect that the company will equip their entire customer base, and you expect that there will be very little cheating. So, basically, every single employee will.

Kieran Chandler: Get a license out of sheer convenience because actually cheating would be way more expensive and complicated. And so, you charge that, and basically, because it’s very cheap, companies are going to take this license for everybody, no matter whether they are really using the software or not. Then you have, I would say, on the other end of the spectrum, the high-end professional software. So that would be, let’s say, CATIA from Dassault System for computer-aided design, and there you can end up with something where you charge literally five thousand dollars per month per user. Microsoft Visual Studio is something like five thousand dollars per year per user. So, I mean, there is still a wide range of pricing, but it’s very expensive. And here, what do you expect companies will do?

Joannes Vermorel: Companies will only equip the strict minimum number of people, so they will only take people from a single role and equip them with that. Vendors know that to some extent people will cheat, not massive fraud, but let’s say you have a piece of software that is very expensive. If you have a colleague that only needs it one day a month, he will just use the day he needs it on the same computer where this guy is on vacation. He will just swap with a desk and use the machine of somebody who is in vacation at this point of time. But these sorts of things where you expect that people will do something a bit fringe to kind of lower the cost by having multiple users on the same license. It’s not going to be massively abused, but they are going to have, let’s say, 1.3 people per seat or something, and it’s fair. Actually, the vendors know it, and they charge accordingly, so it’s fine.

Kieran Chandler: So now, what’s the problem with supply chain? If we go back to supply chains as we defined it in a previous episode, where you know, mastery of optionality facing variation, the thing is that you have only a few people who are truly involved with supply chain in the sense of supply chain optimization, and you have way more people that are involved in the sense of supply chain execution, and supply chain execution is everything, you know, transport, manufacturing, purchasing, all sorts of things. So, what’s the issue with software pricing in this context?

Joannes Vermorel: The problem is that if you opt for software that gives you a price per user that is really the price for your specialist, you’re going to pay a lot per user. It’s okay because there are only a few specialists, like the people doing the forecasting and these sorts of things, but they need to share the results broadly with almost the entire company. So, you end up with something that is very dysfunctional where if you have a per-user price, due to the nature of supply chain software, it’s typically not for the whole company, so it’s going to be a few seats, very expensive software per user. It’s fine, but then you end up with a tension that actually almost everybody else in the company wants to have access to that, if only to gain access to the results, because they need that to execute the synchronization. And so, you end up with this massive amount of friction, and overall, it’s not good. You need to have a thinking where it’s kind of company-wide, pretty much by design.

Kieran Chandler: Okay, we’ve covered a lot of ground. It certainly sounds like the pricing strategies out there are a bit of a minefield. Maybe to conclude, it’s probably worth just talking about the pricing strategy we chose at Lokad. It’s this idea of a fixed monthly fee, no implementation costs, no kind of tie-in to any kind of contract. When I first joined, I thought that was fairly risky. Maybe you could just sort of conclude by telling us why you thought that was a good approach to take.

Joannes Vermorel: So, I didn’t think it was a good approach. You know, historically, we did everything else and then, by a sheer process of elimination, we ended up with this pricing model. I mean, Lokad started from day one with pay-as-you-go for computing resources. We were charging, believe it or not, by the forecasted number, so we were charging by how many forecasts we were producing. That was literally how Lokad was charging for the first three years. It was completely dysfunctional.

We did explore all the other alternatives, you know, charging per SKU, charging with success fees. It went terribly, absolutely terrible. And the problem is that it went terrible for success fees, although it was a big success. So, in terms of operation, it was very successful. Technologically, it was a big success; it was working. People and users were happy, but everybody was terrified, and the amount of distrust was just insane. That was probably the most toxic relationship we ever got into with a client, just due to the sheer amount of stress generated by the success.

So, we literally tried all those things, charging for licenses, and we realized that when we do that, we can’t maintain such a crappy situation. So, literally, I didn’t think of any pricing model. We just tested literally everything until we decided to try something that was weird but was after exhausting all the rest. And what was exhausting all the rest was, well, we are just going to charge a flat monthly fee. That’s it, no, you know, and it’s going to be all-inclusive.

And why has it worked very well? So, the way we price nowadays is we have a company that comes to us, and we look at pretty much two things: the complexity of the problem and the scale. The complexity is independent from the SKU count. Do you have MOQs? Do you have multi-sourcing? Do you have price breaks? Do you have one ERP, two ERPs, 27 ERPs, or multiple countries? So, what is the complexity of the landscape we are trying to address?

And then what is the scale? Are we talking about a 20-million-dollar company or a 20-billion-dollar company? Because obviously, if you’re much bigger, you want to achieve a much higher level of optimization. Optimizing down to the last percent when you’re a 20-million-dollar company, it’s not worth it. When you’re a 20-billion-dollar company, even 0.1% is a huge amount of money.

So, basically, we look at complexity and scale, and we end up with this kind of flat pricing. And what are the benefits? Well, in terms of incentives, first, we have to invest. It costs more for the setup, you know, there is no workaround. Initially, the cost is higher, but by having a flat fee, we don’t charge for

Kieran Chandler: So, Joannes, I wanted to start by asking you about Lokad’s pricing model. You’ve mentioned before that you charge a flat fee, which is a bit unusual in the software industry. Can you talk a bit more about that?

Joannes Vermorel: Yes, so we charge a flat fee, and it’s quite unusual because most software vendors tend to charge on a per-user or per-transaction basis. But for us, the flat fee has a lot of benefits. One is that it aligns our incentives with those of our clients. If we charged a higher fee at the beginning, we don’t get the incentive of taking the money and running. As a CEO, I realized that it was great because it was a way to make sure that for me, and for all our teams, including you, Kieran, we have nowhere to run. The only way to have something that is profitable at the end of the day is to have a client that stays with us for literally, yes, actually our own internal calculation. We want to have something that breaks even at two years, so we know that the client pays us right from the first month, but for the client to break even, we need to have those two years’ period, so it needs the client to stay with us for two years to even break even. So, we can even think that it’s insane, but the way I see that is that that way, we know that when we are not doing something great, there will be pain, and we will know it. So, it’s not something where you just sell a license, the client pays, and it doesn’t work for the client. The situation is very terrible, but for you as a vendor, it’s great. So, you see, we share the pain. We can’t really align things like success fees, but we can align the fact that we will share the pain if it goes wrong, and that I think, either way, it’s very important. So, first, that gives us this very strong incentive.

Kieran Chandler: That’s really interesting. What other benefits does the flat fee model have?

Joannes Vermorel: The second thing that it does is that people think it gave us, you know, over time, a very strong incentive to keep improving the solution, whether the client asks for it or not. We are fundamentally paid over and over, and we want to basically lower our cost. Our costs are the computing resources. So obviously, because what we’re earning is a flat fee, it’s us who have to basically improve the software so that it’s more performant so that we don’t waste computing resources. That’s the opposite of pay as you go for computing resources, and also we are providing maintenance as part of the package, so we want to make sure that this maintenance is as cheap as possible. Literally, this is part of this insurance model, where I think it’s kind of the zen of supply chain. You have something, look, charge a flat fee, the client can leave any time, so you have no commitment typically, and we have to prove month after month that we’re good, and we are completely at risk. That’s the interesting thing, is that Lokad is at risk any month of the client leaving us, and the only stickiness that exists is just because clients trust the performance that we achieved over several months and then over years. It requires a high degree of trust that this level of performance is not accident, that it’s very hard to replicate. And although the grass is always greener on the other side, it’s not anywhere clear that they would get anywhere close to this degree

Kieran Chandler: Okay, we’ll have to wrap it up there, but thanks very much for tuning in, and we’ll see you again in the next episode. Bye for now.