00:00:08 Open to buy and its role in the fashion industry.
00:00:23 How open to buy works as an inventory allocation process.
00:02:45 Industries that use open to buy and its prevalence in fashion.
00:04:30 Why open to buy is popular and the role of subjectivity in its implementation.
00:06:58 Limitations of open to buy in forecasting and budget allocation for sales.
00:08:00 Categorization and its impact on business decisions.
00:10:05 Challenges and complexity of implementing a simplistic budget system.
00:12:45 The problem with top-down management and arbitrary decisions.
00:14:48 Why classic supply chain models don’t work in the fashion industry.
00:15:55 The importance of using better mathematical tools for decision-making in fashion.
00:17:20 Simplifying the decision-making process for adding new products to an assortment.
00:18:41 Changing fashion industry’s hierarchical structure to adopt more efficient solutions.
00:20:44 Impact of crisis on the fashion industry and learning from it.
00:23:42 Closing thoughts and end of the interview.


The founder of supply chain optimization software company Lokad, Joannes Vermorel, discussed the “open to buy” budget allocation tool in an interview with Kieran Chandler. Open to buy is a method for managing inventory by setting budgets at a category level, making it particularly useful for fashion items with short product life cycles. While this method has limitations, it remains popular in the fashion industry due to its simplicity and ease of implementation. Vermorel suggested that using product attributes to create probabilistic forecasts could help address issues with traditional supply chain optimization models in novelty-driven industries like fashion. The interview also explored the impact of the coronavirus epidemic on the fashion industry and how it might lead to companies reevaluating their processes.

Extended Summary

In this interview, Kieran Chandler speaks with Joannes Vermorel, the founder of Lokad, a software company specializing in supply chain optimization. They discuss the “open to buy” budget allocation tool, its popularity, especially in the fashion industry, and whether modern technology can offer an alternative approach for the fast fashion generation.

Open to buy is a simple inventory allocation process with several variations. It serves as an alternative to the more basic min-max inventory allocation method. Min-max works well for companies selling the same products over a long period since it assumes repeat business with the same items. However, it is not as efficient for industries like fashion, where product rotation is high.

The open to buy approach tackles this issue by shifting the min-max perspective from the product level to the category level. Instead of considering individual products, categories (e.g., pants) are examined, taking into account new items coming into the assortment and old ones being phased out. This method is typically driven by collections in the fashion industry.

While the open to buy methodology is predominantly used in the fashion industry, it may also be applied to other novelty-driven industries. However, Vermorel notes that in most cases, its usage in other sectors is due to someone with supply chain experience in fashion recycling their knowledge.

The canonical open to buy method differs from the min-max approach in that it is more financially focused. While min-max is expressed in terms of inventory units, open to buy is generally expressed in dollars.

They discussed the concept of “Open to Buy” and its implications on supply chain optimization, specifically in the fashion industry. Open to Buy is a budgeting and inventory management technique used to determine the amount of stock that a business can purchase based on its financial constraints. Typically, companies allocate budgets for various product categories and then further divide these budgets into quarterly, semi-annual, or annual slots. This method is particularly popular among fashion companies.

The main reason for using Open to Buy in the fashion industry is that the traditional min-max method doesn’t work well for fashion items with short product life cycles. Open to Buy offers a more granular approach to managing inventory by setting budgets at a category level. However, there is very little research on this topic in academic literature, as the method is mostly based on top management’s decisions and gut feelings.

Supply chain practitioners are responsible for converting these budgets into actual purchase orders or manufacturing orders. The quality of the decisions made depends heavily on the expertise and intuition of the individual practitioner. Open to Buy sets macro constraints on budgets, which limits the size of potential mistakes that can be made in the process.

When it comes to forecasting future demand, Open to Buy doesn’t involve any real forecasting. Instead, it allocates budgets for purchasing volumes, which in turn become self-fulfilling prophecies. For instance, if a company budgets for purchasing one million dollars’ worth of pants, that’s what they will purchase, and they will sell them, albeit possibly at massive discounts.

The categorization of products in Open to Buy is subjective yet reasonable, as it is usually based on the types of products being sold. Fashion brands typically have a clear picture of their product assortment and can classify items accordingly. The issue arises from leveraging this categorization to drive arbitrary and rigid budget decisions that may not be aligned with the business’s needs.

While Open to Buy is relatively easy to implement due to its simplistic nature, it is not necessarily simple in practice. This is because it is a very manual process, which may require significant expertise and intuition from supply chain practitioners. Despite its limitations and reliance on gut feelings, Open to Buy remains a popular method in the fashion industry due to its simplicity and ease of implementation.

In the interview, Kieran Chandler, the host, discusses supply chain optimization with Joannes Vermorel, the founder of Lokad. They focus on the challenges in managing complex organizations, such as fashion companies, and the constraints placed by traditional budgeting processes.

Vermorel explains that top-down budgeting processes in complex organizations involve multiple layers of management. The first layer makes broad budgetary decisions, while subsequent layers break down the budget into smaller parts. He mentions the time aspect with seasonality, where budgets are allocated into quarterly or monthly buckets, which many companies consider better. However, Vermorel disagrees, as tighter constraints can limit the potential for big mistakes but also restrict the ability of teams to respond effectively to surging demand.

He emphasizes the importance of data analysis in the decision-making process, and the potential for better granularity in understanding demand through daily or even hourly analysis. However, he acknowledges that open-to-buy constraints can lead to a company culture where management decisions go unquestioned, which can hinder improvement within the organization.

Vermorel points out the problems with traditional supply chain optimization models, which rely on time series and assume a level of stationarity in the business that may not exist in fashion or other novelty-driven industries. These models struggle with products that have no sales history, which is a common issue in the fashion industry.

To address this, Lokad uses an alternative approach that leverages product attributes, such as shape, color, price, materials, and style, to create forecasts for new products without sales history. These forecasts are probabilistic in nature, due to the inherent inaccuracies in predicting demand for new products. By using this method, the company can make more informed decisions on new products, taking into account factors such as cannibalization and substitution.

They explore the idea of making marginal decisions in product assortment, focusing on the fashion industry. Vermorel suggests that traditional hierarchical fashion companies would need to change their operating methods to benefit from this approach, which could make processes simpler and more streamlined. The conversation also touches on the impact of the coronavirus epidemic on the fashion industry and whether it might lead to companies reevaluating their processes. Vermorel believes that while some companies may successfully adapt, many will not learn from the crisis, leading to market filtering and survival of more innovative businesses.

Full Transcript

Kieran Chandler: Today we’re going to understand exactly why it’s so popular and discuss whether modern technology can provide an alternative approach which can cater for the fast fashion generation. So Joannes, how does open to buy actually work?

Joannes Vermorel: Open to buy is basically a very simple process that exists under many flavors. It’s not something completely monolithic in terms of how it’s done. It’s a simple inventory allocation process that is an alternative to just min-max. For very basic inventory allocation methods, you have min-max; you have a minimum and a maximum, and once your stock goes down to a certain level, you just replenish to the max. That works for pretty much all the companies. It’s very crude and not super efficient, but it kind of works, although crudely, for all the companies that have been selling the same products for a long period of time. Because basically, with min-max, you’re just assuming that you will do repeat business over and over with the same products.

As soon as you go into fashion, you end up with a very blunt problem, which is that your products rotate. So if you just adopt something as basic as min-max, which is probably the simplest inventory allocation method you can have, you face a problem: you start to replenish something that is supposed to go away, so it doesn’t work. The idea of open to buy is to kind of slightly shift this very simple min-max perspective that you had at the product level. You’re just going to displace the very same idea at a category level or something. So suddenly, instead of looking at one product, you say, “Okay, I’m looking at, let’s say, pants. There will be new pants that come into my assortment all the time; there are old pants that I’m phasing out.” Obviously, all of that is typically driven by collections when we are in fashion, but basically, you have products that come in and products that come out. Open to buy is a simple idea that you’re going to drive with some kind of constraints, some kind of min-max constraints, at a more granular level, some kind of category.

Kieran Chandler: You mentioned the fashion industry. Are there any other industries that use the open to buy methodology as well?

Joannes Vermorel: I believe it’s very much the fashion industry. There are probably a few other industries that are very driven by novelty that also use that, but in my experience, that’s literally the vast majority. I think I’ve seen that a couple of times here and there, but mostly it was because the person in charge of the supply chain had experience in fashion, so they just recycled their recipes, potentially doing something that was actually fairly different from the canonical open to buy. The canonical open to buy typically takes it from a more financial perspective. The min-max is really in terms of inventories, really in terms of the number of units. Open to buy is more about allocating budget for different categories within the assortment.

Kieran Chandler: Something that is expressed in dollars or euros, where you say, “Okay, I have this category. I’ve already committed this amount of dollars because that’s the stock that I already have, and I’m willing to extend myself up to this, so I’m open to buy this amount of extra dollars to reach my budget target.” And obviously, people still usually add some kind of, I would say, quarterly slots. Sometimes it’s per quarter, per semester, sometimes it’s only per year. A lot of companies are dreaming to go towards monthly slots, but usually they don’t achieve that. They remain kind of stuck with quarterly budgets. And it’s certainly something that’s exceedingly popular with the fashion companies that we talk to. So why is it something that is so commonplace?

Joannes Vermorel: I mean, again, the basic thing is that you start with min-max. Whenever you have long-lead products and you want to graduate, the problem is that min-max doesn’t even work for fashion. So they did the simplest thing that was actually making any sense for fashion, which was to do min-max but at a more granular level, and then you end up with open to buy. And if you look at the amount of mathematics or papers that you will find on open to buy in the literature, there is almost none because, actually, once you’ve said that open to buy is basically having the top management deciding budgets category by category split, quarter by quarter, sometimes month by month, but usually it remains at the quarter level. And how do you end up with those budgets? Well, you just look at what you’ve done last year and you nudge the number a little bit up or down based on mostly gut feelings. It’s not insane, but once you’ve said that and then the supply chain practitioner is supposed to turn this budget into actual purchase orders or manufacturing orders, how do they do that? Well, they look at the novelty and then they just have some kind of gut feeling.

So, it’s very empirical, it’s very subjective. Subjective doesn’t necessarily mean that it’s bad; it just means that the quality purely depends on the person doing it manually. So, if you have a great supply chain practitioner who is super good at feeling the market, maybe this person is going to have tremendous results. If you have somebody who is maybe not as experienced, maybe not as interested in their job in the first place, then you’re going to have results that are not as good. But in any case, the fact that you have put some kind of macro constraints on budgets means that you can’t really go, you can’t really get things that are really insane because the open to buy puts constraints on the budget on all the categories. So, there are limits on the size of the mistakes that you can make.

Kieran Chandler: Okay, so then from a forecasting perspective, basically, those little bumps that you’re adding to a budget, that’s how you’re kind of forecasting what you’re expecting future demand to be like?

Joannes Vermorel: But the thing is that with open to buy, there is no real forecast per se. You don’t really forecast that. And the strange thing with fashion is that you typically, with open to buy, your budget…

Kieran Chandler: Thinking in terms of allocating a purchasing budget, you’re not even allocating for a certain volume of sales, you’re allocating for a volume of purchase. And guess what, if you say you can purchase one million dollar worth of pants, that’s pretty much what you’re going to purchase. So, it’s a completely self-fulfilling prophecy. Are you going to sell those pants afterward?

Joannes Vermorel: Yes, potentially with a massive discount, but you will sell them. There is no forecast really involved, and then you have self-fulfilling prophetic effects that are very strong. Basically, what you decide comes to pass.

Kieran Chandler: How do you sort of decide what the different categories are? I mean, how are you actually implementing this?

Joannes Vermorel: The categorization is the easy part, and typically, it’s subjective but also reasonable, just because it’s type of products. Brands, especially fashion brands, have a clear picture of their assortment and how to classify those products. It’s an established practice to know if you should consider knee-high boots as just like any other boots or a different category. The categorization is not the problem. The problem is that you leverage this categorization to drive tons of arbitrary budget decisions that are very rigid and might not necessarily be aligned with the needs of the business.

Kieran Chandler: So, I guess on the surface of it, the key advantage is its simplicity. I mean, because you’re just taking a category that you already know, you’re applying a budget based on last year, and you can kind of do that from gut feel. Is it fairly simple to implement?

Joannes Vermorel: It is easy to implement, but it’s not simple, and that’s the trick. It’s easy because it’s a kind of simplistic idea, so there is nothing fundamentally challenging. However, in my experience, it’s far from being simple, because it’s very manual. You need a fairly complex organization to make it work. First, you will need to have budgets done at the very top level, such as men garments versus women garments versus children garments. You will have a very top-down process to make all those budget decisions. So, you’ll have one layer of management that is going to take the first 100 decisions, but that’s not going to give you the complete granularity that you need. Then, you’ll have a second layer of management that will cut it down to smaller budgets. Finally, you’ll have to deal with the time aspect, with seasonality, where you’ll need to transform those budget decisions into buckets per quarter, ideally.

Kieran Chandler: So, it seems that most companies view going from quarterly to monthly budgeting as better. What are your thoughts on that?

Joannes Vermorel: I disagree with that notion because while it might appear better, budgets are only constraints on what you do. Having tighter constraints prevents big mistakes, which is the upside, but it also frequently prevents your teams from doing the right thing when there’s a surge in demand. The monthly budget puts a lot of constraints on your actions, and you might feel stuck.

Kieran Chandler: So, would you say that moving from yearly to quarterly, monthly, or even weekly budgeting is better in terms of data analysis?

Joannes Vermorel: Yes, in terms of data analysis, having a finer grain perception of the data is almost always better. But when it comes to open-to-buy, it’s not necessarily better because those are constraints. This can create a lot of discussions within the company since it’s very subjective, time-consuming, and it’s hard to determine who is right or wrong. This often leads to a culture where the management’s decisions are considered final, which can be a problem. You want the management to be challenged on the right things, not have a top-down management culture where things are assumed to be correct.

Kieran Chandler: That’s interesting. So, these constraints seem like a fairly crude approach. What would be a better way to introduce a better level of granularity?

Joannes Vermorel: First, we need to understand why we’re doing this. The classic supply chain optimization models were geared around the notion of time series.

Kieran Chandler: In the literature, optimizing something like min-max or safety stocks all of those things assume a certain amount of stationarity in your business or an ongoing stream of novelty. They’re geared around the notion of time series, which is also a very simplistic mathematical model, and in the case of fashion, it just doesn’t work because a lot of products don’t have any history. So the question is, I mean, the first step, if you want to do better, you just need mathematical tools, analytic or statistical tools that let you handle a situation where most products don’t have histories, at least in the sense of taking decisions.

Joannes Vermorel: It doesn’t mean that you don’t have any history; you have all your previous collections, which is a lot of very valuable data that is readily available. So, the way we at Lokad approach this is to say we don’t need to have sales history on the new products; we can just leverage all the attributes that we have on the products, such as the shape, the color, the price point, the materials, the style. Typically, if you have a novelty-driven business, usually for most of your products, you have a lot of things that you know. So you can use that to do a forecast. Obviously, your forecast is going to be very inaccurate, so you need to have a probabilistic forecast, or otherwise, it’s going to be deceptive. But based on that, suddenly if you have an appropriate forecasting toolkit, you can go back to something much simpler, where every time you have a new product, you can do a forecast.

It’s not a forecast in isolation because you have cannibalization and substitution. So whenever you decide that you’re going to add a product to your assortment, it’s going to slightly cannibalize plenty of other products, and you have to start making decisions with that in mind. But the interesting thing is that suddenly we are taking marginal decisions, one product at a time. When I say one product at a time, I don’t mean in isolation. You look at one product, it has cannibalization effects, so you need to take that into account, and you can decide, yes or no, am I going to order more? And if yes, okay, this product enters your assortment with an initial order quantity, and then you can repeat that with other products.

In a way, it’s much simpler because suddenly you don’t need three layers of hierarchy to decide. It’s much more streamlined. You have a new design, you can quantitatively assess what’s the outcome if you add this design to your assortment and in your channels, and you decide if you do anything and then rinse and repeat.

Kieran Chandler: You mentioned that fashion companies are very hierarchical. Would they have to change the way they fundamentally operate to allow for a solution like that? To work, it’s actually much more simple. You know, the thing is that all those fashion companies want to be relatively reactive with respect to the latest trends. But the problem with open-to-buy is that nearly all the processes that I’ve seen in this area are very geared toward waterfalls.

Joannes Vermorel: Yes, you end up with a first stage where it’s going to be range planning and then some other thing about planning, and then it’s going to be purchasing, and so on. You end up with various names, but typically, you end up with a waterfall process with three or maybe sometimes half a dozen steps. That’s how you can end up with literally six months or even worse, more lead times end-to-end, from the new designs being considered to the product being sold in the channels available to the brand. Overall, it’s very slow, although it looks easy. In the end, you end up with a process that is kind of slow. And so, yes, there’s a lot of change, but mostly it’s good changes. It’s things that are just simpler, more direct, and where you just realize that you have layers of processes that are just not needed anymore.

Kieran Chandler: The fashion industry was one that was particularly hit by the coronavirus epidemic recently, and we’re seeing retailers slashing their prices at present to try and come back from that. Do you think that will cause the industry to reevaluate where they’re currently at and be open to new ideas, or do you think it’s one that’s fairly set in its ways?

Joannes Vermorel: As a rule of thumb, I consider that the market is not a great educator; it’s a filter. My experience in business might be seen as a bit pessimistic, but basically, people just never learn. I’m joking, of course. Individually, people learn, but for organizations to learn, it’s very difficult. What I see is that usually, organizations just don’t learn. What happens is that if you have a very

Kieran Chandler: Dysfunctional processes will just remain until some competitor who does not have a dysfunctional process kicks you out of the market altogether. So, it’s not always the case that you can improve. My hope is that some businesses will challenge themselves successfully and adopt better practices. But I would say my intuition is that a crisis is not going to suddenly make people more educated or push people to get more educated. It just means that a lot of businesses will unfortunately go bankrupt, and the ones that survive will be more willing to adopt better stuff.

Joannes Vermorel: The way I see that, you could have examples due to the crisis. For instance, Walmart just decided to shut down Jet.com, an e-commerce platform that was growing fast and even managing to compete against Amazon. Succeeding in growing against a giant like Amazon is very difficult. They got acquired by a company that didn’t have the right culture, and in the end, they failed after acquisition. This is an illustration that Walmart has been failing at competing against Amazon online for a decade. Even when they acquire, they end up injecting their own culture, and they don’t really manage to fix that.

So, I believe that the companies that will come very strongly 10 years from now will be the ones that have challenged these sorts of things the most. But again, that’s just survival bias. I believe that the market will act as a filter, and the very good fashion companies, some of them have already started to diverge quite significantly from the classic or open-to-buy process that I’ve described.

Kieran Chandler: Okay, well, we’ll have to leave it there, but thanks for your time. So that’s everything for this week. Thanks very much for tuning in, and we’ll see you again in the next episode. Bye for now.