The individualisation of products is driving the digitalisation of business processes and the necessity for ever greater transparency of the whole supply chain. If complexity is not mastered comprehensively, you have to pay a high price. Franz Hero, Senior Vice-President of Supply Chain and Logistics Development with SAP SE (Walldorf, Germany), on trends in supply chain networks.

Mr Hero, what drives the logistics and supplier industries?
The need to transport goods from A to B is as old as humanity, but transport is now taking place on a worldwide scale. As a result, supply chain management is in the middle of globalisation under the impact of networks.
The sharing economy, well known from the private sector, is now also meeting us in the business world: With the sharing of resources and capacities, the division of labour between companies is becoming more and more demanding. The exchange of information is increasing, because only in this way can the supply chain be further optimised. The IT industry is required to provide platform solutions, for example, to enable capacity matching within the supply chain in logistics networks, with the goal of establishing visibility through all instances. Questions arise, such as: Where are my transports and with what kind of freight? What effects do I have to expect if certain delivery dates can’t be met? What effects on production? On customer satisfaction?

So, let’s wrap up: The first trends you mentioned are globalisation, sharing economy, and networking…
…right. Another big thing is the individualisation of the products. We are always moving towards a lot size of one. If you go to the discount store, you can’t fail to notice the ever wider selection of consumer goods such as shampoos, deodorants. The products are increasingly tailored to specific target group segments — but note, no more sales space is created to accommodate that! The replacement cycles must therefore be much faster. In the meantime, shelves are refilled on a daily basis, sometimes even several times during an opening period. The consequence: The delivery of the entire product range takes place several times a day. Generally speaking, the supply planning cycles are getting shorter and shorter.
We can also see the effects of this in the warehouses of the distributors. Meanwhile, it has got so far that different product ranges are packed onto one pallet. We have just completed a project with a consumer goods company to automatically optimise stacking of goods on a pallet.

I suspect that looking ahead to demand is a real challenge with such a diverse offering.
Definitely. Especially when you consider that service levels of 97, 98 or 99 percent have to be achieved. This means, for example, that in 98 percent of cases where a retailer orders a product, you have to be able to deliver on time.
By the way, the whole thing goes upstream from the distribution centres of the retailers to the distribution centres of the consumer goods manufacturers. From there to the production facilities and suppliers of consumer goods manufacturers. Previously, planning was done quarterly, then monthly, followed by weekly. You didn’t need to go any more detailed than that – but now planning is done on a daily basis!
We have taken these short cycles into account in our product portfolio. We’ve included algorithms that allow the use of pattern recognition to understand how demands change seasonally under certain conditions, for example. Clearly, the more point-of-sale data is available, the higher the predictive quality.
In addition: Company boundaries are becoming increasingly blurred, which is another trend...

…what do you mean by that?
Take a consumer goods manufacturer who also becomes a retailer, such as Nike or Adidas – this is a consequence of e-commerce. Just a few years ago, the traditional distribution channel was dominated by traditional retailers, but more and more so-called flagship stores are being opened, which are closely interlinked with online shopping.
Another very prominent example is Amazon — traders become logistics firms. For the implementation of an e-commerce strategy the decisive factor is logistics: Can I organise goods delivery to the customer at a reasonable price?

All this means that SCM has gained in importance enormously, because despite all product diversification, the costs should not be allowed to get out of hand.
Yes, we are noticing this very clearly in our business. If a consumer goods company steers towards retail or e-commerce, logistics costs will increase many times over — logistics will turn into a considerable costs block! High-performance software solutions are required to truly and comprehensively optimise logistics.
It can also be observed that logistics companies are turning into production facilities, at least in part. Let’s take 3D printing as just one example: With this technology, logisticians in their warehouses can offer in-house production services. For us, this means that we have to ensure the close link between the warehouse and production. We are currently working on such solutions.
And of course: Manufacturing companies are becoming service providers. A well-known example is Kaeser Kompressoren, which now offers not only compressor products, but also compressed air as a service. Company boundaries are increasingly dissolving and for us as a software vendor, this means that our solutions must be flexibly adaptable. A platform must be provided to support business processes of consumer goods manufacturers, logistics service providers, as well as retailers.
Some days ago, I was talking to a major automotive supplier in Germany. This company now has its own unit that acts as an internal logistics service provider for its various business units. They use our logistics system as a logistics company and not as a manufacturing company. So, we bring flexible, tailor-made software products onto the market.

Does this affect the software architecture?
Yes, sure. As an example, let’s take our SAP Integrated Business Planning (SAP IBP) for on-demand foresight. We deliberately chose a cloud architecture because of flexibility. The client defines the relevant key figures: What sizes do I want to plan with which granularity? And at which aggregation or product group level? Or: Do I even want to do this at the product level? It is a very flexible system that breathes both in product hierarchies and in the time horizon.

Keyword communication: What importance do you attach to connectivity?
Very high importance. It’s not just about the connection in the sense of business-to-business systems, i.e. the exchange of information between companies, but also of business-to-thing.
But what’s it about? In trucks, on-board units are installed. We access this data and feed it into our business systems. Let’s assume, a truck is planned from Munich to Hamburg and we observe via the recorded geodata that it is stuck in traffic and cannot keep to the delivery time. It is thus possible to decide immediately whether the customer needs to be informed or if there is still enough buffer to postpone a reaction.
It is also possible to monitor the prevailing temperatures or other measurements in the truck, which are important for the transport of goods. In case of later complaints, this might be of interest. The same applies to the transport of wafers for chip production, which are very sensitive to shock, or of pharmaceutical products with their moisture sensitivity.
A colleague from Shanghai told me the other day that now over 80 percent of daily life needs are ordered there via smartphone or tablet PCs: The Chinese order in the subway and two hours later their order is delivered. For this purpose, an infrastructure in the sense of the so-called last-mile delivery is necessary. At the same time, a great deal of dynamism must be mastered, because the customer may change his mind in the short term. We have worked out a solution that allows rebooking as long as the goods have not yet left the warehouse. You can imagine what a massive intervention this means in traditional warehouse logistics and what effort must be made to increase the flexibility of the control!
The entire picking process within the warehouse is changing as a result of mobile e-commerce. In the past, whole pallets were loaded with the same product, today small packages are packed. The orders have to be summarised intelligently in order to optimise the warehouse processes.

At the end of the day, the question of sustainability remains: Does that matter at all in these considerations?
Of course, because supply chain management has a relatively large impact on it – an effective supply chain is environment-friendly one. This means that if I succeed in consolidating transports so that there is only little idling, I make a contribution to CO2 reduction.

Makes sense. What is your personal favourite in SAP’s portfolio?
Currently in great demand is our supply chain planning solution: SAP Integrated Business Planning (IBP). All sales figures are consolidated in a database and based on this, smart algorithms generate the demand forecast, even on a daily basis. This is an enormously important topic for the entire consumer goods industry!
A new concept in demand planning is Demand-Driven Material Requirements Planning (DDMRP), which radically simplifies demand forecasting by decoupling the supply chain. Decoupling points can be found in the inventory. The trick is determining where and in what quantity the warehousing takes place, at which points in the network the stock should be — and that has to be adapted dynamically. We have now implemented this concept in SAP IBP. We work with the Demand Driven Institute in the US, leading consumer goods companies, and other partners. With the DDMRP approach, inventories can be reduced by up to 20 percent.

That’s an impressive number! SAP sees itself primarily as a standard software fabricator. You talked a lot about customisation, also in the context of software. How does this fit together?
Our software is customisable and extendable. Adjustments are made via parameter settings, as you know from MS Office applications. Extensions to our cloud platform are possible through so-called User Exits, code-based upgrades, in which the client can store their own programmes to supplement one of our software products. We provide the platform, but the maintenance of these special codes is done by the client.
Not a small number of companies rely on the last 10 percent of optimisation in order to gain a competitive advantage and therefore want to add capabilities to our standard software.

This answer brings us to the buzzword ‘smart enterprise’. What do you associate with this term?
Machine learning. The baseline of our entire portfolio is machine learning. Gradient Boosting, for example, is a new algorithm that can dramatically improve demand forecasting. Other areas we are currently working on are the automatic segmentation of X, Y and Z products with respect to A, B and C customers. X-products are fast-movers that are sold a lot, and Z-products are the store-keepers, if you like. Depending on the individual segments, different algorithms are used. Our goal is to fully automate this segmentation.
From a transportation perspective: How did the dispatchers handle the transports in the past? With which means of transport? The system learns from these historians and automatically makes suggestions on how to combine means of transport and orders, for example to achieve maximum packing density. With each release, new machine learning algorithms are added.

Thank you for talking to us!
Interview: Bernhard D. Valnion