Businesses now rely on analytics to help understand their processes, their products, and their customers. Having the right information available whenever business leaders need to make a decision is just as important as the information itself.
ASUG spent time with the authors of SAP S/4HANA Embedded Analytics to talk about how to benefit from the tools already available in SAP S/4HANA. Jürgen Butsmann, part of the global business development team for SAP S/4HANA and Thomas Fleckenstein, chief product owner of SAP S/4HANA analytics shared their insights with us.
Our conversation covered how you can use embedded analytics for operational reporting and to process analytics, how to configure SAP S/4HANA embedded analytics, and how to integrate with other SAP tools for data warehousing, business intelligence, and predictive analytics.
Sharon: Let’s start with the basics. What is SAP S/4HANA embedded analytics?
Thomas: Simply put, SAP S/4HANA embedded analytics is a tool set for both transactional and analytical processing that delivers on the promise of SAP HANA’s real-time database.
Sharon:How do embedded analytics help businesses?
Jürgen: Embedded analytics allow a business leader to make the right decisions, whether it’s about what to prioritize within a transaction or what to focus on based on facts and real-time information.
Sharon: What makes SAP S/4HANA’s embedded analytics better than past offerings or what’s out in the market?
Thomas: We’ve had embedded analytics for about 20 years. When data volumes and user numbers were relatively low and only a few experts worked in an ERP system, businesses could run a report, even if it consumed a massive amount of the central processing unit and memory, without disrupting the transactional processes.
That doesn’t happen anymore. User numbers have gone up, as have data volumes. Businesses expect things like a Materials Requirement Planning process to be done in four hours instead of six while avoiding a query running for purely analytical purposes on the same instance.
Not wanting to run analytical and transactional processes at the same time created the need for two systems: a system of record and a system of engagement. All ERP systems nowadays, such as SAP S/4HANA, are systems of record. A BI system or a data warehousing solution, on the other hand, is a system of engagement.
You use the same data to maintain a sales order as you use when you report on it, but the level of consistency between transactions and analytics is unattainable. What we see in the market is that companies are buying separate solutions with systems of engagement and putting them next to their system of records. That’s an old-school way of doing it.
With SAP HANA and embedded analytics, we do not have to extract data from a system of record and move it to a system of engagement anymore. A refined machine learning model is right inside the system, along with more agile intelligence.
Sharon: What’s the difference between embedded analytics and stand-alone analytics?
Jürgen: It’s really where the data resides and where the data is coming from. If we look at embedded analytics, we’re talking about data that is inside of the transactional system where the operational decision support happens.
When it comes to stand-alone analytics, we are talking about creating models that answer questions that are independent from the actual processes and data source. It means you are independent of your ERP system, but you also must harmonize the results.
Sharon: What are the architectural components required for SAP S/4HANA embedded analytics? How does SAP Fiori fit in?
Thomas: As the name implies, the embedded analytics run in SAP S/4HANA, which is on the SAP HANA database. As for the architecture stack, we have one or two components that are specific to analytics but are not separate. The first is an analytics engine that we took out of the SAP Business Warehouse (BW) code.
The other is the Virtual Data Model (VDM), which is a set of algorithms that takes the physical table in the SAP HANA database and transforms the data without replicating it. Customers can extend the models in the VDM or build their own.
Those are also the data sources for the SAP Fiori user interfaces (UIs). Whatever we do with respect to embedded analytics materializes for the end user in an SAP Fiori app, whether it’s for analytical or transactional processes. SAP Fiori is our one and only UI paradigm that we have within embedded analytics. And on the front end, that is the main architectural piece we need.
Sharon: Can you explain the prepackaged content and best practices available with SAP S/4HANA?
Thomas: There are two levels of content. First, we have content that comes out of the box once the system is switched on. You can currently use about 600 predefined analytical apps across different lines of business. A business can personalize those apps too. We have features and functions that allow customers to do their own thing. SAP will continue to ship new content with new releases, which will always be available right away to the user.
The second level of content allows users to create their own SAP Analytics Cloud stories, which are interactive dashboards. You will need to have a connection configured between SAP S/4HANA and the SAP Analytics Cloud tenant, and also have the authorization set up. These are considered best practices because it’s more a suggestion of how customers can use these dashboards.
Sharon: How does data warehousing work with SAP S/4HANA embedded analytics?
Jürgen: A data warehouse stores and combines data available from various data sources (SAP or non-SAP). That data is then transformed into structures that offer a harmonized view of everything, so it can be used for reports, graphs, dashboards, and other reporting formats.
SAP BW/4HANA is an enterprise data warehouse that aligns with SAP S/4HANA because it was built on the same principles. It has both data extraction and data federation processes. With the first, we’re extracting data from SAP S/4HANA or other data sources and into SAP BW/4HANA to create those harmonized reports. With data federation, we can pull the very latest data directly in real time into the structures of the historical data, allowing real-time evaluation of all comparable data.
Sharon: When is it useful to combine other business intelligence solutions?
Jürgen: The focus in embedded analytics and the SAP Fiori tools is on the enhancement of the transactional process and not necessarily on the best possible visualization. When we talk about business intelligence solutions, we are referring to the best way of visualizing the tools and not the interaction and integration inside of the business processes and the ERP.
Business intelligence solutions make it easier for users that require that visualization. For example, I wouldn’t expect the CFO to ever to log in to SAP S/4HANA to get one piece of data out of it. These types of users will typically go into their digital boardrooms or into their dashboard applications, such as SAP Analytics Cloud, to get the information they need.
Sharon: What can something like machine learning or artificial intelligence add to the data being analyzed? How does predictive analytics work?
Thomas: It’s interesting because on one hand, applying machine learning or artificial intelligence is simpler than most people think. But on the other hand, it’s a lot more complex. There are two steps to consider.
First, and the more complex of the two, is to come up with a precise question. You need to then go through a thorough analysis of your data, and finally you need to come up with a model that takes the available data and tries to answer the question with it. That leads to the second step, which is simpler because once you’ve trained that model through machine learning, it begins to show you predictive analytics.
That is something that customers get with SAP S/4HANA embedded analytics.
Sharon: What tips do you have for training teams on the new systems?
Jürgen: One of the first steps I’d recommend is to go through the book. Another tip would be to work within a demo system if you have one available to build your own cases. That’s certainly the best environment to improve your skills.
There are also a variety of training classes, workshops, and demos available through SAP and its partners.
Sharon: Can you summarize the book?
Thomas: It’s a comprehensive guide. We lay down the foundation and explain a lot of terms. There are sections that detail the various use cases and predefined analytics. We cover processes and best practices. We discuss roles and differences in business intelligence or enterprise data warehousing. We discuss predictive analytics and machine learning and much more. There is something for everyone.
ASUG members can log into their accounts to get their discount on SAP S/4HANA Embedded Analytics. To explore the latest advances in analytics, join us March 9–11 in Atlanta for the BI+Analytics Conference.