David Schweizer 0:00:05.7: 

Hello. I'm David. Welcome to the supply chain track here. I'm happy to show a bit what we did in the past years. I will start with myself. I'm David Schweizer. I'm vice president finance for the European and international region. I have started at Oatly five years ago exactly in March, and I started as a responsible finance person in the DACH organization, so I know all the commercial struggles and the operational things going up to our solution here, that I will show you. I'm based in Berlin, and I was very happy to visit Stockholm now, due to this event. It will be now a few minutes, between ten and thirty minutes. Let's see how long it lasts. How we get from demand to the contribution providence, part of the integrated business planning that we do in Oatly. Our agenda for today, I think most of you know Oatly, I would assume, at least after seeing the logo on the first page. It's called aided brand awareness. Luckily, it's not a marketing presentation today, but a bit operational facts here. 

David Schweizer 0:01:22.8: 

Anyway, I will tell you a bit about Oatly. Secondly, I'll show you the business challenges that we, as a food company in FMCG, feel the tensions every day. Third, we go into the solution, or first why we have chosen Anaplan as the finance solution to our volume translation, and then how the solution looks, and the outlook, our options, and our potential that we still see in the near future with Anaplan as the solution. So, the Oatly story. You know Oatly, but does anyone also know Rickard Öste? Rickard Öste was a professor at the Lund University, and when he was travelling in the late '80s to Asia, he realized that there are more than 90 per cent of the people lactose intolerant, and how they manage, he also realized there, when looking at the shelves and seeing a lot of milk alternatives, mainly soya and almond at that time, and plenty varieties of it, but this was the way how they worked around their lactose intolerance. So, he came back to Sweden and thought about, how can we do a white, watery, liquid drink that is similar to milk ourselves, and what crop could we use? As a Swede, he was very close to oat, and he found a nice way to turn oat into something similar like milk. 

David Schweizer 0:02:51.5: 

Brussels is watching us. We are not allowed to call it milk. It's only an oat drink, but anyway, I will continue probably to call it milk here. So, Rickard Öste found a way to turn the oats into this milk by a completely different process than all the other competitors, and most of the competitors that are still in the market. He found the enzyme process that breaks down the natural fiber into sugar, and makes it into a completely different white, watery, milky thing than our competitors do, by milling it and squeezing it. So, this process that he found made, not only this liquid that was close to milk, it was also super-healthy, and it was very tasty at that time. So, in the end, he created that drink that we still know as the Oatly drink, the best oat milk in the world still, and I think there are no doubts about that. This is only the beginning of the story of Oatly. This was in the very early '90s and how we come from the early '90s when Rickard Öste created that drink to the very right, for our first, again adjusted EBITDA positive year. I will tell you a bit how we came there over that time. 

David Schweizer 0:04:14.5: 

Still in our core and in our values and in our brand is this very political claim that we want to make it easy for people to upgrade their lives without recklessly taxing the planet's resources. This is challenging consumers and politics consciously to their consumption. So, one-third of the world's CO2 emissions are due to the food industry. By changing to oat milk, to Oatly or to oat drink in general, we reduce our carbon footprint on the milk consumption by 70 per cent. So this is still one of our driving values in our brand, but in 2014, when Toni Petersson and John Schoolcraft created that brand that we know similar to this, it's not only the political words, it's also this very attractive, nice, unique brand that we know today as Oatly. So, this is one pillar of our branding.  

David Schweizer 0:05:11.3: 

The second pillar is we upgraded this first milk from Rickard to a milk that performs very well in coffee. So, it's stable in the acid environment of coffee, we can foam it, and we make great latte art. So, we correct a social phenomenon of the 2010s of the third-wave coffee development of the coffee specialties. So, in the middle of this development, we grew and grew and grew. This was the second pillar of our success. The third pillar is sales excellence. We are in the retail business, we have an FMCG company, we have to correct the retail customers to get big. This is what we achieved in all the excellence. That's a German retail store. We already have one-third of the total milk-alternative shelf, and that equals our market share. So, this is the shelf that you typically find in Germany. It looks similar in Sweden, but I don't know exactly how nice it is at Eko.  

David Schweizer 0:06:17.1:  

Anyway, we're already in the future. We are beyond the milk-mimicking. We are no longer perceiving ourselves as milk or milk-alternative. We are a drink company that offers you taste. No matter whether it's in coffee or latte art, or it's in tea and matcha - maybe you have already seen our ready-to-drink matcha on the shelves - or on-the-go, we also offer ready-to-drink opportunities from the shelf. We call it taste strategy, because we want to provide the best taste experience everywhere to our consumers. So now you know about Oatly, and I'll quickly run through what are the challenges that we have in our industry, food in FMCG, and how does this impact our planning process. So, food in FMCG, it's obvious. On the right side we have the highest needs on quality: we are operating with food. We have short BBDs, what can lead to an obsolete risk when we're not forecasting correctly or super-correctly, and even our raw materials have BBDs. So, our supply chain is a really complex setup, to not lose money. In this young segment of the milk alternatives, it's for us, vital to be on top of the consumer trends. We must now already know what next year the next trend is, so matcha and things, we must get ahead of the wave, to stay up-front in this category.  

David Schweizer 0:07:56.5: 

Seasonality, we are highly impacted by seasonality in our business. Not only are the common flavors that are more linked to winter and summer, we also have a high seasonality on the way of the consumption. Where do the customers consume us? In the winter, or in the times when they stay at home, the home consumption is linked to the retail business, but as soon as you go out - I just heard this story - and it's summer and vacation time, you also have to find Oatly in the coffee stores but also in the hotels and all the other [?form 0:08:28:6] of locations where you consume milk. So, we have this strategic pillar of the other form business. Then, food in FMCG and non-dairy, when looking at the portfolio there was always the question about new categories - yes/no - but 80 per cent of the non-dairy categories in the market is drinks, so we focus on drinks at this moment and keep it to that.  

David Schweizer 0:08:53.9: 

Retailer complexity, everyone operating in FMCG business knows how tough it can be. I learned it's a bit less stressful in Sweden than it is when negotiating with [unclear name 0:09:06.8] in Germany, but it's all about price. You're talking the fourth digit of a price, and fighting, and it can be deal or no deal on the fourth digit, so margin is heavy to keep. Promotions is the way to go, but it's a high spend on the promotions, so you also have to be very conscious and able to control your trade investments perfectly. Then, the next thing, retailer discussions, it's about items. Do you go on the secondary placement displays or cartons? There can be easily 25 per cent margin difference between those two types of ISPUs, although it's the same product on it. So, the retailer complexity is putting a lot of pressure on us in the daily business, and then it's competition. The milk segment is shrinking since years, in volume and in value. The milk-alternative segment is still growing, so we also attract the big milk boys with a lot of money in their back to go into the alternative segment, and if not entering the segment, fighting the segment by different interpretation of KPIs on fiber, of protein, or whatever. So, we have a constant discussion with the milk lobby. 

David Schweizer 0:10:25.8: 

Then, it's very political around our brand. We have on our products what milk doesn't have. We have sugar tax in Norway and Poland, and we have a lemonade tax in the Netherlands that we pay on our drinks. What else we have, VAT is mostly higher than milk in our markets, because our drinks are taxed higher. So, there's a lot of politics going on in this segment, but even in the segment itself, the crops, and the fight among the crops is also very big. There, it comes to consumer trends. Right now, almond is rising again after years of decline, because it's perceived more healthy and less sugary, so every crop can tell their story, and we have social media they find and listen to us. It's very important for us to keep the branding up, not only to be within the oat category the distinguished brand, but also for the whole milk alternative. This all comes down a very volatile demand when we talk about promotions and things. The volume input is completely manual. Our sales guys themselves are forecasting and evaluating the sales of the future. Super-manual input and very coordinative. Trade investments are very high, and hard to control, and the worst thing is to get a return on investment calculation on it. It's very difficult and tricky to get there. 

David Schweizer 0:12:07.1: 

Then, raw materials I've talked about, short best-before date, but the financial risk is also when you keep a too-low inventory, to not be able to fulfil the deal that the sales wanted to make. So, the opportunity risk on not being able to supply is on the same side than the obsolete risk that we have. So, it all comes down to forecast accuracy and an agile process in the integrated business planning. So, let's come to our solution. Why did we decide for Anaplan? First, we feel very familiar with numbers. I would like to start with that. We are talking about 62 different markets, 13 currencies, and we get the input on 375 customers. The demand volume is planned on 375 customers, but with our solution translated down to 945 different price customers, because we are selling on different prices for every customer. We have more than 1000 different COGS, the costs of goods sold, to manage, and this all comes down to a combination of nearly 29,000 different item customer combinations that we have to plan, monitor, control, and be on top of it when it comes to mix analysis and everything.  

David Schweizer 0:13:24.7: 

We had a solution in place for our financial planning that was super-manual, had challenges with the user experience, and we have just outgrown this solution. So, our question was, could Anaplan replace the third-party system that we had, and unify the entire commercial planning process, because for demand planning we have been with Anaplan already for years, and it's working well, and we knew that. We also did our supply and inventory planning there. So, we thought, can we integrate our gross-to-net calculation, can we integrate the COGS, and can we come to a gross margin to enable a faster, full P&L integrated business planning? We decided, luckily, yes. The integration and connectivity that we already had in place for our demand planning was huge, so we could make a lot of use of the solution that was already there. Given the hard time that we have in our former solution, we asked our demand team, who was already here, 'How is the user experience in Anaplan?' They were also convinced that it's quite efficient to navigate in Anaplan, and they liked the way how it thinks and works. In the pitch that we did with different vendors and service providers, we also were convinced that Anaplan has a lot of options beyond what we are asking today. So, we are super-sure even today we can integrate AI options or other things, third-party data, that we are not demanding today, but our options are limitless. 

David Schweizer 0:15:02.2: 

Then, important for us from the financial perspective and from the process-efficiency perspective, we needed a solution that is SOX compliant, to get rid of all those manual controls on our accruals and planned bookings. So, from demand volume to integrated business planning, in Anaplan for Oatly means, basically - and we see in the dark orange the fully automated processes - we come from the volume planning that we take from the existing process. We source list price, discounts on invoices, [?ex-post 0:15:39.5] rebates, all from our systems, and let make them the calculation. We have easy logics and a few smart logics in place to control them, to check on them. Then, the only thing that we manually input is our trade spend planning. This is going for penalties, listing fees, promotions, whatever, and even there we have put the whole sales organization directly in the system. They plan on SKU level, on financial, on customer level, so that we have enhanced the granularity by far to the former solution. 

David Schweizer 0:16:22.6: 

Let me check, do I have everything? Yes. So we have more granularity, we have a bigger transparency on the data, we have increased our accuracy by being able to go down to every price customer, and by this, we have reduced the working time about by 500 hours a year, so we got way more efficient in that solution by all those automations and integrations that we incorporated in Anaplan. This was the core project when we wanted just to replace our former solution, but at that time we realized that it's a very short step to go to the contribution profit or gross margin, because we had so much logics and data in place already that it was, basically, the COGS and the warehousing that we also have in our systems, we could integrate it as well. We have a few remaining items that they can easily fill in, like customer freight-based fee, or the sugar tax on themselves, and we are down to the gross margin. All of this is integrated into Power BI, so as soon as the team - and the Nordics team is here, they can prove that I am not telling any bullshit - as soon as they have input in Anaplan, a few hours later we have everything in the sales dataset, and we can look in Power BI and analyze everything that we want. 

David Schweizer 0:17:48.1: 

So that's the solution that we put in place, and what we get out of it was the trade spend controlling. So, we are super-much-more granular in everything what we do, and we have more transparency. We succeeded a monthly IBP cycle. What we did before quarterly, we can now easily do monthly, given the efficiency that we have, and we look at the full P&L for every of our markets every month, and we have put all the accrual management in place with integrations to D365, going out with booking files. Everything is done in approval workflows, very lean management, just following the SOX compliance regulations, so that we get rid of the manual controls and everything. So, what we achieved as from before, that we have a volume integration, that we have the integration of many of the sales data across sales and rebates and stuff. We have automated the accrual bookings, going into D365. We have the audit trails and the approval gates automated, and we have a unified volume to contribution profit calculation, what has been very manual before, and when you wanted to analyze differences or gaps, it could have been easily that one of the finance managers has taken the wrong COGS to decide, messing the whole story. So now we are assured what we have in place, and we can spend more time on the analysis. 

David Schweizer 0:19:17.1: 

So, the impact coming from that, comparing it to more than a year, or to a year ago when we started with it, we were able to improve our forecast accuracy by 121 basis points. We have saved more than 600k per year value of obsoletes. We have saved more than 500 hours a year on the IBP forecasting, and more than 400 hours a year on the trade spend bookings, monthly accruals and settlements, and management of SOX controls. For the future, looking from now, what we want to do, there are a few things that are already triggered with our partner, Polestar, and we will start very soon. Others that are more in the future, it's about a live net sales calculation. Today, we update the net sales after every month with the new volume planning, but we realized that, given the granularity that we have, we see very better what mix impact is coming every month, and we would like to see it already during the month when the volume planning is updated in the demand volume module. So, we plan on a live calculation there. There is already a logic thought done. Market data integration would be a nice thing that will link us to Nielsen, to get a better return on investment calculation on our promotions, because we only have selling data, but in the promotions it's also a valid KPI to look at the outside market data. 

David Schweizer 0:20:52.8: 

To automate the return on investment calculation, there are many AI solutions that could support us. Even it's the checking for completeness and correctness, but also just scenario planning that you give top-down assumptions and the system takes the calculation itself. One of the bigger things would be to integrate even below margin, so take the full sales journal and administrative costs into the Anaplan solution to get the full P&L done in the system, put into the Power BI for analysis. So those are the next options that we have in the system. That's already what I had for today, but I'm happy to answer all questions, if there are any. 

[Applause] 

Unknown 0:21:52.0: 

Any questions? We have a microphone? 

Audience 0:21:53.8: 

Yes, question, how many involved in it, employees? 

David Schweizer 0:21:57.7: 

Full employees? 

Audience 0:21:58.8: 

Yes. 

David Schweizer 0:22:00.7: 

I was prepared for the question on this process, so it's about 100 stakeholder on that process there, and full Oatly, I'm actually not sure. I think it's easier to look it up. Do you know at the moment? Close to 2000 I would have guessed, but I'm not sure.  

Audience 0:22:18.7: 

Fifteen-hundred. 

David Schweizer 0:22:19.2: 

Fifteen-hundred, okay. Thank you. Yes? 

Audience 0:22:24.9: 

What was the reason for selecting Anaplan? 

David Schweizer 0:22:28.0: 

Yes, it was the existing process of the volume planning, and we knew that our integration is working. We had a lot of data already in the - like customer data and item data - in the demand planning module, and that we knew from our colleagues on demand planning that it's performant. So, it's quick enough to take our complexity, and they liked the user experience. They ran me though what they are doing, and Anaplan chose tables and stuff, it was completely good for us as well, and so we decided for Anaplan because of those three points, mainly, and the SOX compliance was a thing. Our former solution wasn't SOX compliant, and this was a big burden on the financial processes. 

Audience 0:23:19.2: 

Any particular reason to use Power BI? Does it make it more static? 

David Schweizer 0:23:25.0: 

You mean instead of reporting in Anaplan? Yes. I mean, there was a limitation to the project budget, to be honest, but also we are using the Power BI and the Power BI Excel-based analysis tools, and we can just do our datasets and tweak them as we want, and we are very much used to it. So, for us in the first place it was important to get a good integration of all the data, and we are creating a lot of - like 29,000 item combinations per month. Now, you can imagine how much this is. Probably for bigger companies it's not, but for us it's a huge transparency that we have. Using sales datasets for it gave us the best opportunity to analyze it, but yes, maybe I missed that in the milestones. It's a good option, as well, to think of analysis that we can do already in Anaplan. The way of working was a bit different, and is still a bit different at that time. Yes? 

Audience 0:24:31.5: 

Can you plan for calculated COGS as well? 

David Schweizer 0:24:36.4: 

No, not yet. They are looking into it from their self-finance, but today we get the calculated COGS integrated from the D365 system, ERP, into the Anaplan loaded. Yes? 

Audience 0:24:55.2: 

Are you using scenario planning in Anaplan? You mentioned AI [unclear words 0:25:00.2]. 

David Schweizer 0:25:00.4: 

No, not yet. That's a wish that we were thinking about, but our challenges - to be honest I think we are not there with Oatly yet to fully utilize AI scenarios, because we have still to fix - like this was our first step to fix our real basic calculation of net sales. Now we are there, and now we can take another step. Also, given the segment that we have, the dynamics that we have, we are still more determined by politics than we are by scenarios anyone could think of. So, I think, whether AI knows what the politics does in Italy next year better than us, I don't know. So, no, we do those risk analyses in the financial analysis on our datasets. We know what the profit is, and we can calculate whatever, impact any decision it would have on our markets. Yes? 

Audience 0:26:02.8: 

Can you mention a few things that went well in the implementation, and maybe something that you would want to implement next time? 

David Schweizer 0:26:09.9: 

Yes. So, there's one clear thing that I think was the key success factor, was a proper sophisticated - no, not sophisticated - it was a proper and very serious kick-off. So, we had the Anaplan partner team from Polestar over four a week. It was four colleagues from them, it was four colleagues from us. We sat together five days, and we really were thinking through all the options that we have, so that even this live gross-to-net-sales calculation is already in place, the logic from that time from the kick-off, or the way to the contribution profit was already thought in the kick-off. Spending enough time there was really helpful, and the communication with the guys and the mutual understanding. They went off and totally understood what we need. What I also realized, and I thought at that time I have pointed it out enough, is when you're hiring a partner that is super-agile, that put a lot of resources on the decoding, the budget, and whatever you have, they are working 100-time on it. On the other side, you have your own IT that is limited by the daily business they have, and although we have highlighted that before, I probably wasn't strong enough in saying, 'I need half FTE for three months straight only for Anaplan.' So, the things that made us trouble were actually Oatly reasons, because our IT wasn't ready with the integration in time and stuff. So, yes, I think those were the two main impacts on the project. 

Audience 0:27:53.5: 

When it comes to the [?gross net figure 0:27:55.9] there are so many ways how rebates [?discount all of them]. Did you have to do workarounds, or tweaks [unclear words 0:28:05.8]? 

David Schweizer 0:28:08.2: 

No, actually not. You can break it down to the three different types. You have direct discounts on the invoice, you have rebates that you mostly put percentage of gross sales and accrue for it, and then they send an invoice every half-a-year, whatever, and you have the projects that are super-volatile. So, what we do for direct discounts and invoice and rebates, we take the actual data from - and we can choose last month or last three months - per item and financial chain, and we load that calculation into our system, and we just prolong that. Then you have the chance to say, 'Okay, increase of one basis points from September on everything.' You have easy logics in place to review it, so if there's any item outstanding because it's more than 10 per cent different to any other item, you get it highlighted. So it wasn't that difficult. The most difficult thing is, and most of the company, even Polestar, always called it TPM, trade promotion management module, but the project management is not that difficult if you educate your sales people enough, or the finance people do it themselves.  

Davis Schweizer 0:29:24.5: 

It's four things: it's promotion, it's listing, it's penalties, and others that we have in place. They are planning for it for forecast purposes, but we also are accruing on those values in the monthly closing. So, it didn't feel that complicated at that time, but maybe you have another example of your business environment that would break that logic. I don't know. 

Audience 0:29:52.6: 

What about the [unclear words 0:29:52.3], do you have any estimates on the volume of the [over speaking] that purely sales? 

David Schweizer 0:30:00.8: 

That's purely sales. So, there is kind of an offline calculation from the sales guys. Mostly it's probably just that the sales guys know what they have paid the last time, but we have solved our organizational at that time, saying, 'When you put in a promotion in the demand planning, and there they plan baseline and promotion, then you're allowed to put in a promotion cost for it,' but this is not even a next project phase, this is probably one of the next things we fix in the next few weeks, that we link the volume and the investment to each other, because that's the first step to do the return-on-investment calculation then. So yes, the uplift planning and the demand planning module, we have our costs, we now have to link them together. 

Audience 0:30:48.7: 

Thank you. 

David Schweizer: 0:30:49.1: 

You're welcome. I think we are good.  

Unknown 0:30:58.7: 

Any other questions, last-minute questions? No. Right, thank you so much, David. Incredible presentation. 

David Schweizer 0:31:05.4: 

You're very welcome. 

[Applause]