One is demand-driven planning, which allows companies to plan materials based on consumption without a forecast, rather with a preview of future requirements. And the other is the Theory of Constraints, which is particularly suited to situations where accurate forecasting is difficult.

Anyone who has been involved in materials planning in recent years will be familiar with MRP and CBP. MRP uses a demand forecast as an input for planning and, therefore, usually requires access to historical data or to reliable sales order data. CBP, on the other hand, does not use forecasts to determine longer-term future requirements, using the reorder point to trigger purchase orders and production orders. In practice, the decision about whether to use MRP or CBP depends on whether generating a forecast is worth the effort involved.

Forecasting Demand Is Costly and Time-Consuming

It makes sense for consumer goods manufacturers who sell large quantities of laundry detergent, for example, to use an MRP forecast for planning because washing machines are in continuous use. So, as , an SCM Business Processes senior consultant at SAP, explains, “There are likely to be only relatively minor unexpected fluctuations in demand for laundry detergent.”

But not every industry can generate forecasts simply and reliably. At companies in the mechanical engineering industry, for example, planners have to plan several hundreds of thousands of materials. “For them, the effort of creating a detailed forecast for every single material is just too great,” says Gulyássy. Therefore, many companies use statistical algorithms to generate forecasts for active items. However, this automated approach cannot adequately account for the specific aspects of certain materials.

Push Control in Procurement: Forecasts 80 Percent Accurate at Best

The downside of MRP, as a push type of inventory control, is that “Quantities are pushed into the process even though the company does not have an accurate picture of customer demand,” says Gulyássy. At best, he says, the likelihood of a forecast – even a good one – being accurate is usually between 70 and 80%.

The downside of CBP, as a pull type of inventory control, is that the long-term perspective is lost. Suppliers rely on forecasts to produce and deliver parts reliably, and they even give their customers discounts where forecasts are available. A preview of future quantities is required for capacity planning and availability checks, too – but is not possible as a standard in consumption-based planning. A blend of push and pull control would therefore be ideal.

MRP and CBP: Advantages to Both

The compromise companies adopt is to vary the use of pull and push control according to material or material group. “This means that they have to classify all their materials,” says Gulyássy. For example, the purchasing department might use consumption-based planning and others might use MRP. That would mean suppliers wouldn’t get a forecast, but the company would be able to perform accurate capacity planning and availability checks.

In practice, this compromise means that resource management is not ideal, because companies often compensate for uncertainty by building in buffers. So, sometimes stock levels are too low and sometimes too high. “That could mean that a company works in three shifts when two would be sufficient,” explains Gulyássy, “or that it spends money on expensive and higher-performing machines to provide flexibility – when it doesn’t actually need to.”

Demand-Driven Planning and the Theory of Constraints as a Basis

The two new materials planning methods leverage the benefits of both the push and pull strategies – though each focuses on just one of those two strategies. The new methods are derived from the planning philosophies of demand-driven planning (demand-driven replenishment planning) and the (bottleneck-oriented planning) that are possible with the 51·çÁ÷SCM consulting solutions from release 2018. The idea is to create the optimal pull-push hybrid and therefore benefit from the advantages of both planning methods for each individual material at the same time – without losing information about future quantities in the multi-level context.

Leveraging the Benefits of MRP and CBP in SCM

With the customized MRP type add-on in 51·çÁ÷Supply Chain Management (51·çÁ÷SCM), 51·çÁ÷gives customers the option of using two new algorithms. The “DD” (demand-driven) approach is based on consumption-based planning, which is augmented by aspects of MRP such as, notably, the preview function, which, if required, also takes account of dependent requirements. The “TOC” (Theory of Constraints) approach is based on demand management but also follows the reorder point strategy for triggering procurement and production proposals and has consumption-based planning embedded in it.

At operational level, planning is consumption-based, while demand management creates a forecast of the longer-term perspective. In both cases, this happens directly in production, not in a simulated environment. “This means that planners have immediate and full transparency across the entire planning period in the current stock/requirements list,” says Gulyássy.

DD and TOC for Planning: Four Benefits at a Glance

Now, no matter whether they use 51·çÁ÷ERP orĚý, companies can supplement the familiar materials planning methods with one of the new algorithms – provided that they have the customized MRP type add-on. The benefits include:

• Both new methods enable a preview of future quantities, which is calculated from known sales orders or is based on a forecast, that is reflected in the planning of the entire bill of material and bill of distribution structure at all levels.
• The functions for capacity planning, previews of future requirements for suppliers, and availability checks (similiar to MRP) can be used at all times.
• Consumption triggers replenishment, which means buffers that were often required in the past for subordinate bill of materials and distribution levels are no longer necessary.
• The preview of future requirements can be created directly from the operational environment. A separate simulation, such as long-term planning for consumption-driven materials, is not necessary.

DD: The More “Reactive” Sibling of TOC

The two add-ons for 51·çÁ÷ERP and 51·çÁ÷S/4HANA were developed by 51·çÁ÷consulting in conjunction with customers.

One of them, a tool manufacturer from Switzerland, benefits specifically from TOC. “It can perform consumption-based planning without the added effort of a forecast – and still has an overview of future requirements,” says Gulyássy.

Another, a mechanical engineering company, is using DD, the more “reactive” sibling of TOC, which is particularly suitable where forecasts are difficult to generate.

This story originally appeared on the 51·çÁ÷Germany News Center.

Most companies that switch to 51·çÁ÷S/4HANA choose one of the two standard methods. As market researchers at discovered last year, more than 80 percent maintain their existing process and IT architecture and corresponding modifications (brownfield approach) or choose to leave their historical data behind and make fundamental changes to their processes (greenfield approach).

That said, a third option has grown increasingly intriguing – the selective data transition.

“Companies are recognizing the advantages of pinpointing the historical data, organizational units, and processes they want to keep while modifying and innovating other specific workflows and structures,” reports Alexander Rombach, director of Consulting Practices at 51·çÁ÷Services.Ěý

Selective Data Transition: Mapping New Business Models While Retaining Proven Practices

There’s an entire spectrum of individual approaches in between the two standard models. While a system conversion takes place on the basis of established data and processes, a new implementation uses fresh or migrated data and the processes provided by 51·çÁ÷S/4HANA – along with innovations like fast close, finance ledger solutions, and new features supported by artificial intelligence (AI).

“The greenfield or brownfield approach isn’t always a good fit for a given customer’s requirements,” Gregor Staubach, an 51·çÁ÷software architect who oversees selective data transitions on the customer side, points out. He cites a number of examples of why access to historical data might be necessary or required by law, including in connection with product returns, complaints, or warranty claims, or with the documentation obligations common in the aviation and defense industries.

Choosing Between Shell Conversion and Building from Scratch

Companies can therefore tailor the third option to their particular strategic goals. Those that want to gain flexibility and implement new business models will likely tend toward a fresh implementation. Meanwhile, a system conversion is probably the better fit for organizations that prefer to stick with their existing business models and processes. In order to take individual preferences into account, the selective data transition approach includes several different levels. Here, 51·çÁ÷Services uses the terms “shell conversion” and “building from scratch.”

While a shell conversion largely preserves a customer’s particular processes and structures and only comes up with new developments in specific cases, the build-from-scratch approach focuses on redesigning processes and retaining those from the past only when they fit right into the new template at hand. In the real world of business, it is common for companies to carry over around two-thirds of their processes and data while seeking to start fresh with the rest.

“That’s when it’s faster and more cost-effective to go with a selective data transition instead of a full greenfield implementation,” Rombach says.

Typical Scenarios for a Selective Data Transfer

The software architects at 51·çÁ÷Services face an array of common situations in this regard. These include maintaining certain company codes or requirements, converting charts of accounts, and needing to integrate a company in short order. For 51·çÁ÷Services experts in particular, the following areas are important:

• Redesigning financial accounting
  Companies want transparent business figures, and some of those making the transition to 51·çÁ÷S/4HANA decide to limit their efforts to implementing a modern financial system based on the standard suite. The goal here is to apply the corresponding finance template, which might include the cloud version as well, while preserving all other processes and histories in areas like supply chain and logistics. A redesign project in accounting and controlling could lead to a uniform, ledger-based chart of accounts, for example, or simplified controlling structures and hierarchies.
• Cleaning up after acquisitions and split-offs
  Many companies have a system landscape that has grown and evolved over the years. Other companies have been bought and sold, integrated new enterprise resource planning (ERP) systems, expanded or adjusted processes, and implemented other systems in parallel. In these cases, it makes sense to identify company codes and data that are no longer needed and exclude them from an upcoming migration to 51·çÁ÷S/4HANA in order to start with a new system that’s streamlined and up to date.
• Multitasking during preparatory projects
  Those who first need to harmonize systems and start preliminary projects for 51·çÁ÷S/4HANA may end up committing many employees to these activities and losing time they need to move to their new suite. Here, carrying out selective data transition projects at the same time in targeted areas – to modernize a financial system or pare down the data to be migrated, for example – can be a good idea. Preliminary transformation projects involving things like financial harmonization or modernization can, in other words, be completed in the same step while transitioning to 51·çÁ÷S/4HANA.
• Bringing IT landscape architecture up to speed
  There’s no shortage of companies with landscapes that have grown over time and no longer provide the best possible support for their business models and strategies. In such cases, establishing split architectures during a selective data transition is one way to move on. Doing so involves integrating financial and logistics systems into a central instance of 51·çÁ÷S/4HANA. Consolidating several systems – for specific regions, for instance – into a single 51·çÁ÷S/4HANA installation can also be a very valuable step. “It’s an efficient and streamlined method that combines with the innovations of 51·çÁ÷S/4HANA to bring dynamic growth to a company’s business,” Rombach explains.
• Utilizing data from acquired companies
  After purchasing another organization, a company has to import its IT into a new infrastructure. At the same time, however, access to business-critical data – figures and trends from years past, which the company’s controlling department will continue to need, for example – should still be ensured after the systems in question have been merged.

Selective Data Transitions: A Catalyst for 51·çÁ÷S/4HANA

In the experience of 51·çÁ÷Services, the ability to use the innovations 51·çÁ÷S/4HANA offers is not enough of an incentive for many companies to make the move.

“When it presents the opportunity to update their architecture, put it in order, and perhaps even harmonize it to provide optimal support to their new business model, however, a selective data transition can act as a catalyst for 51·çÁ÷S/4HANA,” Rombach says.

He compares the transition to renovating a house: There’s no need to rebuild everything from the ground up if you can just replace the heating, but what if that’s not the only issue?

The third option might include new windows, some roof insulation, and maybe even a new addition. “It’s the entire renovation and modernization effort that needs to be consistent and worthwhile,” Rombach continues.

Keep What Fits and Focus On Implementing Specific New Components

For a good example of what the third option might look like in IT, look no further than a paper producer that was already using 51·çÁ÷S/4HANA, but then purchased a company that hadn’t yet made the transition.

“We integrated 51·çÁ÷ERP into the system they were running,” recalls Thorsten Spihlmann, another solution architect at 51·çÁ÷Services. “In another project, we modernized the landscape and proceeded with an incremental rollout, which is also typical of selective data transitions.”

As a positive side effect, the customer is now using only around 10 percent of the old vendor data from its previous heterogeneous system. In this case, the third option proved to be the right one.

Selecting Historical Data for the Transition

A selective data transition is meant to strike an ideal balance and enable companies to choose what historical data and processes they want to keep going forward.

“It’s often the most cost-effective way for companies to switch to 51·çÁ÷S/4HANA, particularly when they plan to retain around two-thirds of their established processes,” Rombach reveals. In other words, it’s about keeping what fits and focusing on implementing specific new components.

The notion that this approach will be a popular one in the future is also born out of LĂĽnendonk’s research. According to its analysts, most companies see “integrating local IT systems into 51·çÁ÷S/4HANA” as their biggest challenge – and it’s one that will probably continue to be very specific to their situation.

Learn more about the possibilities the selective data transition approach offers in moving to 51·çÁ÷S/4HANA during the webinar “.” Register for this webinar and explore the concept behind selective data transitions and the prerequisites for a lean selective data transition.

A study by management consultancy firm last year found that among successful companies, 70 percent were using algorithms and machine learning and 48 percent were using digital assistants and chatbots. Among their less successful competitors, these shares were 31 and 25 percent respectively.

Managers in many industries have realized that embracing new technologies puts their business ahead of competitors. An Oxford Economics study revealed that 93 percent of managers believe that such technologies are vital to competitive advantage. In a joint study, the Economist Intelligence Unit (EIU) and 51·çÁ÷found that half of those companies that adopted machine learning early saw in it the greatest potential for making their businesses more profitable.

Studies aside, managers still want to know whether investments here really will deliver such significant benefits in the medium- and long-term, whether these benefits can be measured, and how companies can fit innovation projects in around all their other imperatives.

New, Faster Way of Prototyping

In SAP’s innovation factory, consultants from the 51·çÁ÷Services organization have come up with a way of rapidly testing new ideas, building prototypes, and creating and discussing business cases. It enables midsize companies and corporate groups to discover new technologies and learn how they can innovate in a practical and methodical way.

“Many companies take a highly academic approach to innovation,” Juan Galeano Ventura, program manager at 51·çÁ÷Services, says. “But we want to change this by being problem-oriented, and following a method that is flexible and allows companies to change their thinking.”

Even companies already migrating to 51·çÁ÷S/4HANA can benefit from the innovation factory.

“It is also about making 51·çÁ÷S/4HANA even more compelling by enabling extra innovations along the way,” Lars Friedrich of 51·çÁ÷Services explains.

Two Teams, One Approach

The innovation factory method helps companies build early prototypes and generate new business ideas. There are four stages:

1. Set Up Teams: The company and 51·çÁ÷each put together one team. The 51·çÁ÷team comprises an innovation architect, a process architect, and experts in innovative technologies, such as machine learning, IoT, and chatbots, as well as a data scientist and a program lead. The customer provides its project lead, a sponsor from management, business and technical experts, as well as the people who will be using the software that results.
2. Generate and Test Ideas: This stage is not only about discovering that playing around with ideas can be fun, it is also about getting the teams to identify which ones will provide new impetus for the business. The teams also create a backlog – a list of promising ideas – and choose the best ones to pursue.
3. Build Prototypes: The teams then spend 30 to 50 days working on a prototype solution before deciding whether to scale it and then use it operationally.
4. From Proof of Concept to Business Case: At the proof-of-concept stage, the teams test whether the technology is feasible and whether it is easy to use. They then produce the business case.

Fresh Ideas Along the Way

The business case quantifies the financial benefits of the new solution to the company and therefore determines the prototype’s future. The business case matters because the project’s initial funding ends at this point. From then on, the team must fund itself, Friedrich explains. Team members from the company spend one or two days a week on the project and can return to their regular roles for the rest of the time.

“The project manager still needs to invest about half of their working time,” says Friedrich. Because the innovation factory comprises a clearly defined approach and helps ensure the teams have the right mix of skills, the resulting prototype is less likely to be consigned to the desk drawer.

Jump-Starting Innovation

“It takes courage, and investment up front. The rest is a bit like a startup,” Ventura says. “Some things work, some things don’t. If the first prototype goes live, that can really jump-start innovation at the whole company.”

Ventura also notes that what tends to happen in projects like these is that as time goes on, the ideas are less about the technology itself: “It’s about getting creative and recognizing the practical benefits to the company.”

There are already early success stories: One major German manufacturer that tried the innovation factory approach generated 60 ideas. It whittled them down to the 10 most promising, and turned them into prototypes. Three are now about to become part of its 51·çÁ÷software and company processes.

This story originally appeared on the 51·çÁ÷Germany News Center.

Innovation managers from the 51·çÁ÷Digital Business Services organization in the Middle and Eastern Europe (MEE) region are skilled in spotting topics that generate value for companies and for making these ideas happen.

Andreas Spahn knows the stumbling blocks that prevent companies from pursuing innovation. Most are set up to optimally manage the business model that is making them successful right now.

“All the company’s structures, processes, tools, and cultures are geared toward doing what it has always done,” says Spahn, an innovation and technology architect for the Middle and Eastern Europe (MEE) region in 51·çÁ÷Digital Business Services. “These days, most business leaders and managers know that they have to innovate. The difficult part is putting ideas into operation to help ensure innovations are a success long term.”

For example, if an employee has a good suggestion on how to automate a process but has nowhere to take their idea, then they may not follow it up. Other companies have the opposite problem: They pursue too many ideas at the same time without prioritizing them.

“Many companies find themselves overwhelmed by the pace at which technology is advancing and by how fast business models are changing,” says Spahn. “Or they are unsure how to fit innovating in around their operational business.”

A recent study by Deloitte found that the new technologies and trends (32 percent) themselves are putting companies off innovating. Security concerns (30 percent), lack of technical expertise (25 percent), organizational culture reasons (23 percent), and lack of leadership and digital management skills (23 percent) are other hurdles to innovation.

Innovating Out of the Status Quo

There is no one right way to go about digital transformation. Each company is unique in terms of its size and culture, and how much experience it has with digital technologies. Companies that are particularly innovative tend to have appointed a chief digital officer, set up digital labs or incubators for startups, or actively go out and scout for technologies to drive digital transformation. Others stick to a zero-defect mentality.

When Maria Fay, Spahn, and Lars Friedrich visit companies, the 51·çÁ÷innovation managers start by asking questions: How is the company progressing with its digital agenda? Which topics is it implementing and how? What other new technologies is it considering? Does it already have a strategy for involving a network of partners? How will Ěýthe success of the innovations be measured?

Relevant digital topics differ from industry to industry, and every department has different priorities. “One company might be on a cost-cutting drive, making process efficiency its priority,” explains Spahn. “Others want to grow and are therefore investing.”

To address different needs, innovation managers have adopted a standard approach. The starting point is an assessment to bring together the business department and IT perspectives and identify joint innovation priorities. Then, selected projects are launched and implemented.

“Together, we identify which topics will really benefit the company and look at whether building a scalable solution really is the answer,” says Spahn, noting that many companies todayare looking to build and sell digital platform models and new digital products and services.

Formats for Practical Innovation Management

How can a company turn a visionary strategy that looks decades into the future into an innovation management structure that sets the parameters of its digital strategy and implements them gradually across business divisions? Friedrich’s answer is a range of formats to suit different requirements. Some customers want to optimize processes, some want to use machine learning for rapid prototyping, and others look to create new digital platform business models.

Friedrich has each customer start with an innovation assessment, a three-hour workshop at which the company and 51·çÁ÷present what innovation means to them. The idea is to understand each company’s narrative and gear it toward the future. “Successful organizations have to think exponentially if they are to tap into openness and transparency in entirely new ways,” says Friedrich.

This is where the future-maker format comes in. 51·çÁ÷experts ask the company to think about which technologies will matter in 2025, 2030, and 2040, and how these future developments influence the business models and business platform models of today. Similar to Gartner’s hype cycle, the technology radar from enables companies to look closely at the key technology trends right now and in the future. It rates them for the situation today and for the industry in which the company operates.

Offered by 51·çÁ÷Advisory Services and devised by change management experts at 51·çÁ÷and Germany’s University of Mannheim, the digital capabilities assessment looks at corporate culture. Digital transformation involves change. Companies must be willing to test new ideas early, accept errors, and team up with competitors. They must also make it all about self-organized teams, not the technology. “This type of culture is not typical of traditional manufacturers or companies that have a conventional management style,” says Friedrich.

Innovation managers from 51·çÁ÷can also put companies in touch with other innovation entities in the 51·çÁ÷ecosystem, such as startups in the and the program.

“When this happens, there is tremendous potential for meeting and addressing specific customer requirements and creating seamless end-to-end processes by integrating 51·çÁ÷solutions,” says Fay. According to her, innovation managers can co-innovate with customers to develop ideas that come about in the exploratory stage and turn them into new 51·çÁ÷solutions.

Hack2Sol and Mode-2-Garage: Tangible Results Within a Week

For a company’s innovation culture to flourish, it must become clear quickly whether ideas will work and should be pursued. One customer in the chemical industry has a dedicated “innovation factory” Ěýand is currently working on machine-learning prototypes with 51·çÁ÷Digital Business Services in MEE.

“In just a few weeks, we can tell whether [we] have found the right approach,” says Friedrich. Events such as , innovative workspaces such as , and HackBots make it simple for companies to team with 51·çÁ÷to build prototypes rapidly. “We often have tangible results within a week that end-customers can validate after each innovation sprint,” notes Friedrich. “What’s more, a company doesn’t have to invest huge sums of money either.”

Going about innovation this way is often vital for getting more people behind an idea, helping prove that is feasible, and creating a solution that is not built on assumptions and genuinely helps end users and customers.

Friedrich explains: “Besides technology, companies also learn what it means to work in an agile environment and quickly focus on the ideas that will benefit users most.