HFS endorses a toolbox approach for integrated automation, where the processes and applications, not the tool in hand, should determine the route forward. The most frequent use cases for robotic process automation (RPA) are industry-generic, back-office processes such as finance and accounting (F&A) with hooks into ERP (enterprise resource planning) systems. SAP claims, “77% of the world’s transaction revenue touches an SAP system.” So, it’s hardly surprising that SAP is directly facilitating process automation for its 440,000 customers across its enterprise systems suite. SAP shared a recent virtual update with HFS on its plans for automation.

SAP is assembling and curating the components of automation

SAP launched its intelligent robotic process automation (IRPA) platform at Sapphire Now 2019. It builds, orchestrates, and monitors process automation bots, spanning robotic desktop automation (RDA) acquired in 2018 (Contextor) and internally developed unattended RPA. SAP’s home-grown BPM features strongly in this narrative for process orchestration along with RPA. In addition, there is an optional interplay with SAP’s Leonardo suite of services encompassing machine learning (ML) for decision making and conversational AI bots. IRPA ships as standard with S4/HANA and can be used on earlier versions of SAP’s ERP. Prebuilt bot templates are available in the SAP Bot Store digital storefront.

The IRPA roadmap contains some features that are already commonplace in RPA vendors’ offerings, such as process discovery and unstructured document information extraction, but these can be achieved through partnerships anyway. SAP’s focus (rightly) is on SAP integrations with SAP software development kits (SDKs) and SAP Digital Boardroom.

Once RDA is required, you might wonder if SAP’s features for unattended RPA beats other tools from the enterprise perspective. If pricing is important, it might. SAP is running with a consumption-based pricing model for IRPA (per blocks of bot runs, not per bot or per seat). Consequently, developers do not need individual licenses, and you pay according to your usage, which removes the risk of being locked into unutilized or underutilized licenses. Free trials of SAP IRPA (three months) were launched in January 2020 and, we await updates on uptake and user experience to see its impact.

Embedded automation is still the long-term ambition

SAP’s long game is underpinning the intelligent enterprise and it has a new heightened appreciation for the experience economy. SAP is rich in operational data (O data) but relatively poor in experience data (X data)—the data that goes with processes and user experience. Presuming that it is more efficient to embed automation technology in core applications rather than place a layer on top, embedding automation is SAP’s stated goal. This was the original intent when SAP first approached the challenge of automating processes with machine learning ML (still an aim, but not yet in the driving seat) and when the Contextor acquisition was announced. Exhibit 1 corroborates the embedded automation approach, as enterprise users place the highest priority on enterprise software solutions to deliver on automation goals.

Exhibit 1: 39% of enterprise software solutions are counting on enterprise software to meet automation goals, while 12% are looking to RPA

Which technology has the highest priority as part of delivering your current automation goals?

Source: HFS Research, “State of Integrated Automation” 2019

Sample: Global 2000 Enterprise Leaders = 317

In the meantime, third-party RPA platforms and a host of SAP specific tools help SAP customers automate their processes

RPA platform vendors (Blue Prism, Automation Anywhere, UiPath, etc.) demonstrate progress with some of the common ERP process bottlenecks. But it is often piecemeal, more at the task level than process level, with origins in document processing, screen-scraping, and batch processing scripts. There are varying levels of orchestration and overarching workflow, and Winshuttle, Redwood Software, and BlackLine predate RPA as finance automation tools expediting many finance departments’ bulk uploads.

A piecemeal automation capability is adequate when the process is simple or comprises only a couple of simple tasks. But to tackle end-to-end processes with a series of steps, you need to span many systems and have a range of task types with varying complexity. Sometimes APIs are an option; sometimes RPA is needed to perform a task along the way—like an arm connected to the main body of workflow. SAP aims to span multiple applications with multiple tools.

A Copernicus moment has occurred; SAP’s systems of record are peripheral to new waves of intelligence and innovation

In recent communications, SAP has made a humble shift. Historically, SAP’s depictions of its interconnectedness with other systems placed SAP at the center of the universe and anything else in its orbit; this is no longer the case. Acquisitions have broadened the breadth of offerings, and, more importantly, SAP now places data, ML, AI (Leonardo), and innovation center stage and SAP applications as some of the multiple peripheral systems.

This mindset shift speaks volumes of SAP’s appreciation for the growing importance of data, data flows, and data-based decision making. It also underpins SAP’s multi-pronged approach of embedded automation in conjunction with BPM, RPA, and RDA. SAP developed an unattended RPA functionality; it is API-based and originated as BPM. SAP is also envisaging scenarios where use cases start as attended automation but develop over time into unattended automation.

SAP is realistic enough to know it is not the only system at play, and HFS predicts that processes that traverse internal and external systems are set to increase as more ecosystem-based business models come into being. A BPM/RPA interplay with process workflows handing off to RPA at various points, therefore, makes sense.

The Bottom Line: SAP is banking on embedded automation outperforming third-party automation tools, especially in terms of longevity and resilience—promising less bot breakage than with third-party tools.

Data underpins intelligence in an organization, and processes form the basis of operations. APIs are central to SAP’s BPM approach, with dedicated connectors and something in between surface automation and APIs “using the protocol of underlying applications” as an intermediary layer.

We believe in the power of AND, not OR, and integrated automation as articulated in HFS’ APIs and RPA. A critical factor here is SAP’s touchpoints on the process—or at times, the lack of touchpoints. For processes that span non-SAP systems, SAP needed Contextor’s attended desktop automation capabilities to be integrated into SAP’s solution to traverse third-party systems’ user interfaces.

SAP’s IRPA is comparable to Microsoft’s Power Automate, giving users an embedded option when thinking about process automation, albeit one with less-rich functionality than many RPA vendors’ offerings. But both giants are in catch-up mode to RPA, and in SAP’s case, finance automation too, where vendors have tackled operational pain-points head-on in terms (and with tools) that business users understand.