Klöckner & Co SE (KCO) Earnings Call Transcript & Summary

Welcome to today's focus call on Klöckner & Co SE Deep Dive into Digitalization conference call. After the speaker presentation, there will be a question-and-answer session. [Operator Instructions] I would now like to hand the conference over to your speaker today, Felix Schmitz, Head of Investor Relations. Please go ahead, sir.

Yes. Thank you. Ladies and gentlemen, I'd like to welcome you to our focus call: Deep Dive into Digitalization. This special occasion gives us the opportunity to inform you about our latest progress and achievements in the field of digitizing the steel distribution business. With me today are Mr. Gisbert Rühl, CEO of Klöckner; Mr. Ricardo De Sousa, CEO of kloeckner.i; and Marek Sacha, CEO of XOM Materials, who will guide you through the presentation. After the presentation, we will be open for discussions. With that, I'd like to hand over to Mr. Rühl.

Yes. Thanks, Felix. A very warm welcome also from my side, and thanks for joining us to our focus call today against the fact that we made in the last couple of months significant progress in our digital transformation on the one side and also with our platform, XOM, on the other side, and also against the fact that the complexity of B2B is somewhat higher than B2C. We thought that it makes sense to give you an update on our current status concerning our digital transformation. And then, in the last part, we also will give you an update on XOM. Yes. Let's directly flip into the presentation on Slide 3, if you like, as a wrap-up, the shortcomings of our industry of our steel distribution and metals distribution business. So on the one side, this is still a very traditional linear supply chain. The suppliers produce steel. The distributors process steel, and then we deliver the material downstream to the manufacturers. Still, there's orders and quotes, mostly made still by phone, fax and e-mail. There are also EDI solutions but mainly only for the contract business. And with this, the delivery times are typically too high and also incorrect. And we all, in our industry, have too high inventory levels. This is also a reason because of the missing transparency in our industry. The flow of information is typically several times interrupted. And with this, there are no effective information on data on our -- on inventories and lead times available, which makes the supply chain or the entire supply chain highly intransparent. With this, we ship steel unnecessarily around, which, on the one side, is not good for our environment, and on the other side, of course, very cost-intensive distributors and also manufacturers need to stock material. We, for instance, have to stock material or tool to balance this missing transparency of demand, supply and prices. Density prices itself, as you know, are very volatile driven on the one side by overcapacity, but on the other side also because of the intransparency. And this is typical for our industry that we're making profits when prices are heading up and then losses when prices are heading down. And then the margins itself are also relatively thin, as you know, and especially for commodities here. And with this, most competitors have the same strategy. They want to increase their higher value-add business, which then, in the end, has a consequence that the prices for higher value-add products and services are coming under pressure as well. So differentiation with this with higher value-add business is finally relatively difficult in our industry. The solution. Next slide is a vertical platform, which connects, in the end, all market participants, suppliers, distributors and buyers and also third parties, like logistic companies, financing, insurance, for instance. And with this -- when you imagine that most of the industry participants were really connected, then there would be, of course, much, much better prediction of demand, and with this, much lower stockholding. Also, transaction costs would go down, and we would have more transparency within the whole, in this case then, supply network. We are, as you can see on the next slide, in principle here, very well connected with our platforms, with our vertical platform. So on the one side, we have our proprietary marketplace. With this marketplace, we're selling our products and complementary products. And then we also launched, as you know, XOM Materials as an industry platform, an industry vertical with a full range of industry-specific and also complementary products and services. Then there are competitors, who having mostly in our industries on the web shops with this, they try at least to sell their own products. Typically, they don't have any additional products on this -- on their onlineshops. So these are all verticals. And then there are also horizontal platforms like Amazon, for instance, but they are not really dedicated to steel. So they make sense here and there when you need a limited amount of steel or when you need a small sheet or so, then it can also make sense to buy this through Amazon, but focus of Amazon is truly not an industry vertical, for instance, for the steel and metals industry also because these are completely different products. So we once had a discussion with Amazon, by the way, and they were saying to us, we are not really interested in anything which is larger than a fridge, and steel is unfortunately typically larger and also more heavy. Yes, next slide. If all this is true, if a platform or platforms make so much sense for the industry, then the question is, of course, how to scale these platforms, especially how to scale vertical platforms and marketplace and the B2B business, which is, as mentioned already, really more challenging than B2C. So typically, when you're thinking about a platform, you think about the Amazon model on Amazon -- onlineshop on Amazon Marketplace. This works very well for all these pack-and-ship products, which Amazon is selling this. Unfortunately, it doesn't work really for the steel and metals industry of our B2B -- for, I think, deep verticals because here -- because this process of an Amazon-type web shop, this covers only a limited part of our business. So when there is, for instance, let's say, a small construction company, and when these construction -- this construction company needs a beam or a pipe or something like this in the morning, then it makes sense to buy this through an -- yes, online web shop or online marketplace and Amazon-type marketplace where you can directly buy this piece of steel. So this is what we call spot business. But most of our business in our industry is not spot business, most of our business is a business where our customers are not buying only one piece of steel, so they're typically buying 10, 100, 200, even 500 items, and they're doing this not immediately. So on spot, they typically request a quote, not only from us, also from other competitors as well, to compare then these quotes to finally make a decision. We have to buy, and this doesn't work on an Amazon-type web shop. And we imagine when, for instance, the machinery company when they come to part list out of their ERP program with, let's say, 50 items, then it wouldn't make any sense to go to a web shop or a platform to put in these 50 items to get prices. So when you do it on a marketplace, you would, at least, get prices from different suppliers. But on an online shop, it would be even less sense because then we would have to do it, let's say, on 5 different online shops, which doesn't make any sense. So the needs here are very different because we have a wide range of different customers, larger, smaller, contract customers, as mentioned, spot customers, customers which only buy when getting a quote first. Automotive -- the automotive business has completely different requirements than the construction companies and so on. That means, in the end, there is no one-fits-all solution, which makes B2B, of course, more complex than B2C. Then there is also the change resistance. So typically, all these companies or manufacturers, they have a typical sales process, mostly still an off-line process in the end, and the resistance to change this process is very high. They're doing it only when there is clearly a benefit, a huge benefit. So when we clearly hit a pain point -- when you clearly hit a pain point when a -- with a platform, then it works. But when you don't hit a pain point by only saying, "Look, sending a fax or an e-mail, this is old-fashioned. The future is digital, and you should better buy digital," this, as an argument, doesn't work in our industry. And then, of course, and this is, I would say, also a special problem of the steel and metals distribution business, this hen-and-egg problem in a way that we have multi-local markets. So even if you imagine, let's say, we have -- we would have a platform in Germany, and when we would have 30 or 50 suppliers, steel -- let's say, steel distributors on the platform. And when these suppliers would then spread it throughout Germany, then the local offering would not be attractive when there is only one. When there are only 1 or 2 local distributors on the platform, then there is no attractive platform locally. And it's a local -- finally, most of our business, when it's not higher-grade steel, is a local business. So our customers here, for instance, in Duisburg, they're buying around Duisburg. They're buying not from Munich their steel. So that means that if you want to -- you have to scale your business also on a local basis. And so only on a national basis doesn't work. You need this local offering. It's a bit -- if you compare this, it's a bit like, for instance, mobility platforms. When you are in Berlin and you want to -- you have done work in Berlin. So let's say, when you only have -- when you want a Uber car, then you want to have it in Berlin, so then it has to come from Berlin, and it doesn't come from Hamburg. So it's -- also these markets, these are -- also these businesses are scaling by multi-local markets. So this is also a special challenge we have in our industry. So you are, in the end, only successful when you meet all these requirements. When you don't meet these requirements, you can have the greatest platform in the world, but it will not scale. So on the next slide, there's an overview how we tackle this problem and which kind of platforms we provide in a marketplace as we provide for our customers. So there's -- on the one side, there is Klöckner itself, and we are developing our business. And Ricardo de Sousa will dig deeper into it a bit later on. So we develop our digital business through kloeckner.i. So on the one side, we developed here a couple of applications and platforms, like our online shops, marketplace, contract platform and our part manager. And with this, we were already able to increase now our online business up to 32% of our sales. And yes, EDI will do -- is missing here, so EDI is also an application which we provide here. But increasing now our sales to -- yes, our target is, as you know, 60% in 2022. What we're really challenging when we would provide only these applications because, as mentioned, most of our business is -- or most of our requests compare for quote -- most of our customers request a quote first before they decide where and what they -- where and what to buy. And here, we develop, and this is new now, the Kloeckner Assistant. We'll come -- and Ricardo will -- I and Ricardo will come back later to this. The Kloeckner Assistant is really game-changing because the Kloeckner Assistant will automate the whole RFQ process. So that means a process which currently could take sometimes days will be finally done in a minute because it will be a completely automated process. So this is Klöckner. And then we have the XOM, the common platform. XOM is, in the end, a marketplace. When there is enough traction on the platform, then it's a marketplace. But also here, we need enablers to scale the platform first. And enablers means, on the one side, we have to scale the platform from the buy side, and then we also have to scale the platform from the sales side. From the buy side, we're doing it with our eProcurement solution, which is also new, which partially based on the Kloeckner Assistant. And with which -- we also here automate the process. And especially with this, we will be able to make out of every customer a digital customer without the need that our customers have to change their process. This is very important. And then we have the eShop. The eShop is currently successful because they're always successful because here, suppliers, also big ones, big producers, are selling material through our eShop to their customers. So they bring the -- with this, they bring the customers on the platform. And here, we scale successfully from the supply side, but this alone would not be enough to scale the platform in a way that it's finally marketplace. We will come back to this and explain this -- or this will be then concerning XOM later explained in more detail by Marek. But let me give you quickly going forward and to realize significant savings also on the -- or efficiency gains on the Klöckner side. So what the Kloeckner Assistant is doing is that when we're getting a request from a customer, the requestability comes, as mentioned, by PDF, fax or even voice. Then this PDF will be converted, first of all, into a digital document. This is not the hardest part because their software for this software is available. But a hard part is to make the interpretation of this document because, typically, these RFQs are not structured documents. It's not like a structured invoice. So we have to find out what does a customer want, how many pieces or how many tons that the customer needs and so on and so on. And this needs a logical interpretation of this RFQ, which only could, in the end, be done by artificial intelligence. Also the discretion, for instance, and this is important then for matching the products, is different because also there is no standard in our industry. And so one customer can say, "I need a construction pipe." The other customer, they're saying, "I need a con pipe." Another customer is saying, "I need a V2A pipe," and this is all the same, but the description is different. So we have to match this descriptions -- different descriptions finally with our product catalog which only also works only through artificial intelligence. And then if you -- when you want to automate it -- this, you have -- you'll need an automatic pricing, which not only gives or provides a standard price. So this is a very specific price -- yield pricing like we're knowing it from the airline industry, for instance. And then the quote goes to the customer, and then the customer can accept the quote simply by a click. This takes, in the end, less than a minute here, the whole process. Today, the average is 30 minutes, but it could also take days until the customer's getting a quote. Sometimes the quotes come even too late. And then we have this eProcurement function from -- of XOM Materials. Here, in the beginning, the process is more or less the same than with the Kloeckner Assistant. So the customer sending not -- this -- the only thing what the customer has to change. So typically, today, the customer is sending, let's say, this quote to 5 suppliers. In the future, the customer sends only one e-mail to us, to XOM. Then here, we have auto conversion, the interpretation, the matching. And then the XOM is transmitting these digitalized requests to the potential suppliers of the customer. They then submitting their quote. And when they want to submit quote, they have to go on XOM, and with this, resolving the hen-and-egg problem. So the customer is saying -- gives an information to its suppliers that, in the future, all requests coming from XOM. And then the advantage here, by the way, for the supplier is that they have also a structured screen to put in their quote. Now today, this looks very different. Sometimes, they do -- have to do it in writing on a PDF. Sometimes, they're getting an Excel sheet where they have to put in prices. But here, they're doing it then on a structured way. And with this, we are also able then really to compare these different offers, which is really a pain point today because, as mentioned, the description could be different. Freight costs are often different and so on and so on. So there are several variables which makes it really very, very painful to compare these different offers, especially when there are 100, 200 or 300 items. And this is then solved by XOM because XOM doing it automatically, and they send -- XOM is sending them to the customer proposal where to buy, and the customer only has then to accept and doing this also via XOM. So with this, we are able then to scale XOM because, here, the customer is bringing the suppliers on the platform. They have a huge benefit for the customers. And the suppliers are, if you like, pushed forth to go on the platform. So next slide, the benefits for Kloeckner itself through the Assistant and other applications, like onlineshop, marketplaces and so on. So what we -- here, we have now, for the first time, specified our targets for the next 3 years. So this year, we want to increase our digital sales from 32% to 40%, with savings -- efficiency gains of around EUR 10 million. Next year, we want to increase the share of digital sales already above 50%. And we are targeting EUR 35 million efficiency gains and then another EUR 55 million in 2022 with digital sales of 60%. So total improvement then finally of EUR 100 million in 2022. We are very confident to reach this target because in our sales organization, 60% of our costs are admin costs currently, which we mostly eliminate through this digital processes, especially through the Kloeckner Assistant. We then have, by the way, also the opportunity to grow also our commodity business again, because with this Kloeckner Assistant, we will be faster than every competitor. So there will be -- where competitors need, yes, typically days or at least hours, we will be able to provide here a quote within seconds. So we expect them to grow our commodity business to increase our market share and of course also to optimize our organization itself. And then next slide, a quick overview how we want to scale XOM with these applications. So on the one side with the eShop as mentioned already, the eShop here, the supplier brings their customers on the platform, 1 seller and 2 end buyers. And then we have the eProcurement function. So the -- again, the customer has a huge benefit, and the customer then pushes or forces their -- its suppliers on the platform. We currently are here in a pilot with 15 customers, and we want to go live in March and April. But Marek will explain this in a minute. And when we then have enough scale from the buy and from the sales side on the platform, then, in the end, we will have also this marketplace function. So with this, I am now handing over to Ricardo. And Ricardo, CEO of kloeckner.i, will show us -- or will go with us through the kloeckner.i slides and especially also explain a bit more in detail the Kloeckner Assistant. Ricardo from Berlin?

Great. Many thanks, Mr. Rühl, for the invitation. So I'm happy to talk to you about how kloeckner.i is supporting the digital transformation of Klöckner Co. The presentation has 2 parts. The first part, to refresh what had been the concept of kloeckner.i within Klöckner group. This means why it was created, our vision, how our relationship with Klöckner Co makes the difference between us and the competitors of corporations, and of course, our successfully implemented applications and KPI. And the second part, I would like to inform you about one of our latest projects, Kloeckner Assistant, that Mr. Rühl already comment and how it can help Klöckner to go through the -- into the world of digitalization in the metal industry. Next slide, please. So kloeckner.i is an internal innovation hub created in 2014 as an independent unit in Berlin with an investment volume of around EUR 50 million. We provide digital products and services for the whole Klöckner organization using the newest methodologies and technology. Currently, we are 90 employees, around 90 employees, organized in different departments. So we have product management department; software developers, yes, we have our internal teams of developers; we have data science and artificial intelligence, where we are responsible to get our data and use it for different purpose, including artificial intelligence; business performance and analytics, including reporting; online marketing; user experience and design; and of course, customer service, where we provide the first-level support for our internal employees in Klöckner and of course the technical support for all our customers and merchants. All employees have the same vision: to use digital transformation as unique innovation to ensure that Klöckner Co is the most client-sensitive player of the market. Next slide. So how does the cooperation between Klöckner & Co and kloeckner.i work? Here is our secret recipe. On one hand, we have Klöckner Co with, I would say, an expertise for more than 100 years in international steel distribution. They are the expert in the business and the processing world. On the other hand, kloeckner.i, that is capable of supporting and providing methodology and technology in order that together with Klöckner group create digital solutions, implement it, and finally, support for the best performance of the global business. So let's give you an example, a real one. Most of -- as most of you know, Klöckner Co has with many customers a specific business model called contract business. Basically, a business where it's designed a dedicated stock with a defined size and a specific time frame. As usual in this industry, there's a prodding of trust clients, already commented by Mr. Rühl. So every day, customer calls the sales rep, asking the status of their contracts, deliveries, documents, stock. That implies too much time for the sales rep dedicated on this and of course not the best service the customer. We at Klöckner came together with the sales reps to get inside this and make interviews to customers because, again, we are customer-oriented, define clearly the problem, ideate the solutions and clear, what we call, minimum viable product to validate together in the market if this is the right product. Finally, we start with the web application, and currently is one of the most successful solution provided for our customers. This, at the end, can be only done if we work closely together to learn from each other and achieve the digital transformation of Klöckner. This is our secret. Next slide. So now let's talk about numbers. So since 2015, we already achieved to around 23,000 customers ordering digital online. From those, around 7,000 are new customers that we already acquired via onlineshop and marketplace. And together with the other solutions or applications that we provide, we already achieved year-to-date, since 2015, EUR 3.1 billion for steel. That implies around 32% of our total digital revenue. What I'm meaning is 32% of the total revenue nowadays we are using with a digital channel. Next slide, please. So the numbers that I already mentioned come from a group of customer that we define as a digital customer. It is relevant to mention that we have a clear customer orientation again. So therefore, we have a clear definition of target group, and this is also is in line with a statement from Mr. Rühl that there's not one-fit solution all. So we did the estimation that roughly 42% of our customer portfolio, we name them as a digital customer. From this 42%, around 20% are a small customer, spot customer that are willing to use solution like online shop and marketplace. From this 42%, around 12% are using the contract solution that I already mentioned, like contract platform, part manager, consignment manager. And on this 42%, 10% is using -- or they're willing to be connected via ERP systems via API, EDI or OCI. But what about the target group that we consider non-digital? In light of customer at this moment in time, we don't know, in the future, doesn't want to change their process and success behavior, maybe because of the complexity of the order or due to internal policies or because they just don't want to feel comfortable with the technology. So after several deals, we conclude that for this important share of customer that is around 50% of our customer portfolio, we have to find a solution where we can make them digital, [ preferably with our nut ], and this is Kloeckner Assistant. So now let's talk about the second part of the presentation, our new innovation project called Kloeckner Assistant that I'm sure that you will enjoy. Please, next slide. So normally, our methodology, let's start talking about the process. And with a quote from our sales representative, "I spend about 60% of my time entering customer call information into the system." So as you can see from the picture, customers send a request, meaning thousands of different formats in play with a lot of information to be processed, most of the times, completely on a total data. So the sales team have to translate information with a lot of minor effort into something that our system can really understand. So this is not only, of course, cost of time of an effort, but also a certain level of knowledge involving to assure that what is requested from the customer is what we deliver, meaning that we fulfill their expectation and also as fast as possible, and as you can see, using a lot of paper. Next slide, please. So let's take a deeper look into the process today. So if you take a look at the graphic on the top, the customer sends an inquiry by e-mail to the sales rep with a PDF attached, which is one of the most common cases. The salespeople have to review his e-mail inbox, identify the inquiry. And in this moment in time, there are several tasks that they have to do. They have to identify the customer, identify the product, match the product, and as Mr. Rühl already commented, there is no standard in the industry. So the customer can describe and also use whatever qualification to identify their own product, and we have to basically translate this to our standard. Of course, we had to tie this cost to the system, calculate the price that implies always the higher possible that the customer is willing to pay, send the quote in the same channel. And those, I would say, they take on an average order of 3 lines around 30 minutes of time and sometimes more, a day or 2. Interesting, right? So now comes the solution. The Kloeckner Assistant will be and is capable of receiving and processing a request and send it back to the customer in less than 1 minute. The customer can also complete this order because we have already a solution where they only, with one click, can say I want the product. This give us, without any doubt, an important competitive advantage since the fast response, of course, is a decisive factor for our customers to buy. The Kloeckner Assistant will be capable of all this process, conversion into digital document, logical content interpretation, matching of product description and automatic pricing, send a quote by the same channel that normally the customer sends a request and also is capable to accept the order. Exciting, right? So we're now imagine that we're using artificial intelligence for this, of course, in the more complex processes, like extracting the information, matching the products, and, of course, apply to the right pricing. So this will enable, first of all, to our sales people to focus on what matters, the customer relationship instead of typing lines manually. Second, this will make us able to process more requests in the incredible, scalable way. So we are capable. We have higher capacity to answer to the request of the customers, and finally, support and contribute announced [ EBITDA ] effect already in this presentation announced by Mr. Rühl. Next slide. So we already started with, first, proof-of-concept in U.S. So as you can see, we started the last quarter starting implementation of the first minimum viable products in U.S., achieving around EUR 7.4 million only with the test. Of course, our road map is already defined in our strategy 2020, so we continue with Kloeckner Metals Corporation in the U.S. And of course, we will do the rollout of Kloeckner Assistant to all the organizations in Klöckner in all the countries involving to be aligned with our strategy 2022. We believe that for this clear, realistic target group that we call non-digital customers, where all the order was implemented doesn't fit with their needs, Kloeckner Assistant will make them, for us, to use a digital channel without any effort from their side, just an incredible fast response. Next slide. So the last message that I really want to provide you is the following: Klöckner Co is already achieving more than 32% of digital revenue is our, let's say, solution for the digital customers. We are starting with implementation of this new solution called Kloeckner Assistant. If we, from now on, not only will continue onboarding more of these digital customers but also start with a one external that are more change-resistant that represent more than 50% of our customer portfolio and even without taking into consideration the acquisition of more customer for the different channels, it is a clear statement that our objective to achieve 60% of the digital revenue in 2022 is more than realistic. So thanks for your attention. I'm now handing over to Marek Sacha, CEO of XOM Materials, to present his part. Go ahead, Marek. Thank you.

Thank you very much, Ricardo. Hello, everyone. My name is Marek Sacha. I'm the Chief Executive Officer of XOM Materials. I'd like to take the next 10 minutes or so to tell you about our company. Please go on the next slide. Let me start with a brief overview. As you may know, XOM Materials was founded in 2017 in Berlin with a mission to create an independent digital ecosystem for the materials industry. Since then, we have reached international presence in Europe and the United States. Currently, XOM Materials employ around 50 people. We operate from Berlin, Duisburg and Atlanta, and we also have sales offices in Prague and Valencia. In 2018, we launched our first product, XOM Marketplace, a B2B platform for steel suppliers, traders and buyers. In 2019, we onboarded many clients from this trade. From 10 clients back in 2018, we grew to almost 60 by the end of 2019. We collected a lot of practical feedback and subsequently launched differentiated products, addressing different customer pain points, as we discussed before. We currently have a much broader set of solutions than a typical Amazon-type marketplace. And I would like to go over our products on the next slide. We currently offer 3 specific solutions: Marketplace, connecting many sellers with many buyers; we also offer eShop, connecting one seller with many buyers and vice versa; eProcurement, connecting one buyers with many sellers. We included strong industrial domain knowledge in our products and features like regional customer-specific prices; individual price requests; customer-specific numbers; multi-branch management; dynamic delivery costs; sale of product batches, coupling services; and many others. These features are very different from traditional e-commerce process. We can, today, address needs of many different companies in the field. Let me give you just a few examples. With the Marketplace, you can reach new markets and geographies without a need to build a new sales team. With eShop, you can keep tight control of your customers while bring them digital, reducing, hopefully, your cost of sales. And finally, with eProcurement, you can make better and faster procurement decisions about hundreds of individual products with complicated logistics and commercial terms. Please follow me on the next slide. Let's now take a closer look at the procurement process. The current purchasing process includes a whole portion of manual work, working with different media, file types and often leading to errors. The complexity is quite big as buyers often have to manually compare different prices from different production plans, including exchange rate effect. Further complexity comes from volume discounts and different delivery options. The decision-making is even more complicated as price is not only -- is not only the decision criteria, but you also need to consider quality and delivery times. Buyers spend a lot of time comparing offers from different data sources with complicated commercial terms. Next slide, please. These pain points are solved by XOM eProcurement. We offer one digital process, making comparison of prices and delivery easy. We include real-time feedback on quality and late delivery. We open a direct line of communication between buyers and sellers, enabling easy inquiries of incoming quotes, negotiations online, contract call-offs or managing different location. The eProcurement service is very new and will be live in the second quarter of this year in April. We are currently developing the product with 15 partners, ranging from SMEs to companies trading billions. And our process increases efficiency, saves time as well as reduces the purchasing cost. This is the most common process in the industry with the largest sales volumes. We hit the specific pain points of our customers, even without the need for any significant change to their internal processes. Our pilot customers already see the benefit and are forcing their local suppliers on the platform, which brings the local scale quickly, which is new to XOM. We, however, also understand the requirements for the spot sales, as you can see on the next slide concerning XOM eShops. We have combined Klöckner's domain knowledge and experience with digitization with results of working with more than 60 of our clients to continuously develop the platform. We enable sellers to go digital in just a few weeks without the need to manage large IT projects on their side. We, in fact, reduce their IT cost and IT risk. With our eShop, sellers get branded version of the platform which shares all the industry-tailored features with our Marketplace. I have already mentioned that there are many special features, including [ cat line ] specific -- customer-specific prices, easy product specifications, inter-leading paper or foil, batch purchases and others. Alternative IT products in the market require heavy customization to make them work to fulfill the industry-specific requirement. Thanks to an interface between the systems of our seller and our platform, we automatically update information such as product availability or prices. Our clients use the eShop to offer real-time data on availability and prices to their own customers. The eShop reduces the cost of sales as it eliminates repetitive human interaction and manual errors. On the next page, you find a summary of our results so far. We grew significantly during the 2019. We have more than 60 vendors under contract, and 36 of them have already gone live. We significantly improved our onboarding process. So far, we've traded over EUR 50 million on the XOM Marketplace in the pilot program with our clients. In 2020, we expect major technology adoptions. The volume of transactions will, in our experience, follow the growth of the client base. We also expect much faster scalings in the launch of eProcurement as this is the first time we scale from the buyer side, as was discussed previously. We significantly grew the product portfolio range to more than 21,000 products, yet we don't have the sufficient offering in many local markets. We are addressing exactly this with the eProcurement product. Despite the growth, we maintained the average order size of over EUR 30,000. To see our outlook, please go on the next slide. In 2018, we launched our initial product. In 2019, we were able to get it to the market, onboard many more clients at different markets, and we have incorporated their feedback. Now we have the scaling phase ahead. In 2020, we expect strong commercial growth after a successful completion of pilot XOM eShops. We are launching the eProcurement platform in April and starting to scale also from the buyer side. Moreover, we are continuously building integration with our customers to build one European and American material network. We offer customer-specific data analysis as per the amount of our clients, and then we extend the product portfolio. And finally, I would say, in 2021 will be about further geographical expansion in regions where we don't have strong presence. I want to sum up our vision and where we currently stand on the next and last page of my part. XOM has created a digital ecosystem involving all parties and players within the materials industry. We are the only vertically integrated platform for the materials industry in Europe and United States. Sellers have the choice to either trade on the Marketplace and gain new business or migrate existing business relationships into an onlineshop solution. Both sellers and distributors can reduce their sales cost and IT risks and go online with just a few weeks and even continue to realize further digitalization benefit. Our eProcurement and Marketplace make purchasing and planning easier, more efficient, and finally, smarter. All our solutions enable the industry impacted by low margins to reduce sales costs, reduce purchasing costs and increase overall efficiency. We discuss daily with our clients their needs for digitization, and we only see the need and urgency to go digital to grow over time. Thank you very much for listening to me.

Marek, so I don't -- I think I don't have to wrap it up. I would then open directly the Q&A session, so please.

[Operator Instructions]

Yes. Maybe then I answer already one question which might came up. You know that we wanted to -- or want to make XOM finally more independent from Klöckner. So it's already independent in a way that the German cartel authorities approved XOM and with a request that we put a Chinese wall between Klöckner and XOM because on XOM, there are, of course, competitive data, which we did. And -- but finally, we think also legally, we have to make XOM somewhat more independent. And doing this, we wanted to finance -- we wanted to get external investors for XOM. This failed so far to a large extent, and the reason here is the following that was also learning which we had to make throughout this journey. So we -- in the beginning, we wanted to get here, as investors, venture capital investors because we were positioning XOM as a start-up. The problem here was, and it took some time because no one was really telling us that this is the reason, also because it was, honestly, also new for investors. Now the problem here was that we are not a typical start-up because a typical start-up has 2 or 3 founders. And then when they get money or funds from VCs, then the VC has typically relatively strong power because they have the money or they gave the money. In our case, they were saying in the end that even when they would invest or take a 10% or 20% stake, they would not really have any power because we are the founder, and we also have the money. So we don't need it necessarily, the money. And with this, XOM is more -- and then they were saying, by the way -- the venture capitals were saying, "Look, it might be better if you get some strategic investors first because this would then for us to be some kind of proof of concept." So we got, by the way, a few strategic investors which bought our convertibles from us, but we didn't convert it, these convertibles, because it was getting -- finally being clear to us when we would have, for instance, let's say, 4 or 5 larger strategic investors, then the platform would also not be seen as really independent. And the majority would be by 4 or 5 or whatever major strategic investors. And with this, we decided that we want, in the end, financial investors, in any case, first. But we -- this is then, of course, not a venture capital case. It's a growth capital case. And for a growth capital case, we need more -- we still need more scale. So we need more revenues on the platform because for a growth capital investor, the platform is currently still too small. Therefore, we decided the following, that we keep the platform for the time being as a 100% Klöckner company up to a point where scale is big enough to get growth capital investors on the platform, and then, of course, also to a higher valuation. So that is maybe one question you might have had. So now we have, I think, also the first question from David Varga, right?

Yes. That's correct. I have 2. Firstly, this seems to stem for the success of the Kloeckner Assistant. Could you give us an idea on the current overall rate? How difficult is it? Or how is the success rate in terms of digitizing all these different kind of order types? Do you have a double check or -- yes, some more color on that one would be helpful. And secondly, if you talk about the growth story and about gaining market share, particularly in the commodity business, what's your base case assumption when you talk about the savings until 2022 in terms of volume growth? And what's your idea, how much market share might probably be gained with this strategy?

Okay. Let me start with the first question. Yes, interesting question, interesting point. We started now, first of all, with a template for our larger and midsized customers before. So the -- with this, the AI starts to learn. So it will take some time until the AI has learned, for instance, also the different kind of description that works in principle in a way that the request comes in and then the request is shown on the screen. So the request will be digitalized and then shown on the screen to a sales person, and the sales guy then checks the request and makes manually -- correct manually mistakes, if necessary. And with this, we are already currently at an 80% ratio. 80% is already transformed correctly. And for 20%, we have to make manual changes and -- but -- okay, this will now then change over time. So we expect that we will then have in the next couple of months, we will then reach maybe 95%. I don't know when we need 100%, that can take even a year or so but especially when they are on XOM, when new customers with new unstructured documents are coming in, but we will -- we are already at 80%. So we reached already a relatively high number. So then concerning the volumes and some more details concerning the numbers, I will hand over to [ Falka Asa ]. [ Falka Asa ] is our Head of Controlling, or new head, I should say, of Controlling.

Hello, everybody. The volume growth is relating to the efficiency gains, which is a process, and it's going to, let's say, right now, we're having around 5,600 tons -- million tons on our volume round. So that's going to be 6, 6.5 million in the range...

Million tons.

Million tons.

So the -- so why don't -- why are we expecting this growth? We're expecting this growth because -- as mentioned already, because we are much faster. Today, we're losing, by the way, here and there also potential orders because we are too late with our quotes when we're doing these quotes manually. But then these quotes are automated, then we can process much, much more quotes. And with this, we provide not only a better service for our customers, we are also able to process, in the end, more orders and more quotes. And then finally, what we also expect going forward that when we -- with this, when we optimize really our commodity business in a way that we -- really, from the costs side can differentiate us against competitors, then we also can grow also maybe in the EMPA acquisitions the commodity business again. And this is attractive because, as mentioned, everyone wants more in this higher value-add business, and multiples for acquisitions of competitors who have mostly commodity business are very low. And we would here, first, probably start in the U.S. where we have anyhow a higher share of commodity business. And here, it was would be, by the way, anyhow for us, for instance, unrealistic to try to build -- to increase our higher value-add business from currently, let's say, 20% to 50% or so. That would take too long. But then when we focus on the commodity business, then we will be the champion finally in the commodity business. And with this, we make a market attractive for us which no one else see really attractive. Next question, please.

[Operator Instructions]

Yes. So obviously, there are currently no more questions. If you have further questions maybe after you -- so we gave you, of course, a lot of information today -- after you went through the material, then please give Felix a call. We can also make follow-up calls, if necessary, if we have to dig here and there a bit deeper or give a bit more explanation concerning our digital strategy. But okay. In the end, I think we, clearly -- things are moving clearly now in the right direction. It took some time. This is also clear, maybe more time as we thought in the beginning. But I most recently read a quote from Jeff Bezos, and he was saying in this quote that further it takes for Amazon 5 to 7 years to build up a new business model, and we think we're really fast but probably not faster than Amazon. Yes. Thanks, everybody, for listening and then talk to you later in our Q1 call. Bye-bye. Thank you.

That does conclude our conference for today. Thank you for participating. You may all disconnect. Have a nice day.