The diversity of customers now trialing and adopting Intelligent Automation exposes the complexity and challenges of doing it right—from small pilots and discrete sub-process deployments to, increasingly, major transformation programs claiming RPA at the core (see Sumitomo announces largest ever RPA deal; but is it really all RPA?). One of the biggest and most common issues customers face is how to scale across the enterprise, successfully moving beyond their early adopter phase.
Customer experience, as it applies to end users of RPA and broader Intelligent Automation, is increasingly critical as customers begin to look beyond the technical differences of tools and consider the context of what it’s like to work with specific tools and services. Questions are asked, for instance, about how people and processes will interact, how much configuration is required, how the tools will respond to changes, and how tools can deliver against expected outcomes (see The HfS RPA Customer Experience Benchmarking Report: The Big Picture View).
| Customer experience is dependent on many things beyond technical features and functions, not least its applicability to the customer’s specific business context and environment |
We interviewed global insurer Allianz, with over 75 other organizations, for the HfS RPA Customer Experience Benchmarking Report. Allianz has been deploying RPA and RDA (robotic desktop automation) for the past few years.
Like many large organizations implementing automation, the biggest challenge for Allianz has been stepping up from the proof of concept stage and scaling across different divisions and geographies.
It had started with a prescribed solution, but quickly learned it needed to adapt. It is now exploring alternative tools and services as it drives the automation agenda through its organization.
Allianz runs SAP as its core ERP system for most operating environments group-wide today. With this single platform, automation became the natural next step, promising to drive out additional costs, improve efficiency, and gain greater flexibility and scale from its existing IT and process investments. Digitalization overall is a CEO-led initiative to achieve true customer centricity with an ambition to become a digital leader by 2018 (source: Allianz).
Allianz Germany’s RPA project manager Michael Rode explained to us that the RPA approach was chosen after realizing that it is impractical, costly, and time-consuming to invest in traditional API-led integration to achieve the same result.
Taking a Best-of-Breed Approach to Intelligent Automation
It’s a common misconception that the differences between the available RPA tools is small. In terms of features, many appear similar on the surface, but it’s clear from our CX research that even seemingly subtle differences impact how tools can be applied and aligned to the client’s own operating environment. These factors become increasingly important as customers determine what they want to achieve from their automation investments (see The RPA Premier League – process transformation comes to the fore).
There can be large differences, for instance, between desktop automation and back-office automation, in part due to considerations about whether the approach should be a lights-out automation without human involvement (RPA), or augmented by humans (RDA). Technical considerations become critical, too, because different tools are inherently suited to different tasks, which can be dependent on the technology underpinning them, whether a process is server, desktop, or ERP-based. Integral to these conversations is making sure the customer has the right tools and related expertise for the job at hand.
In our view, this means customers should consider a multi-vendor approach as dictated by the tasks themselves. During 2016, Allianz chose desktop RPA tools Blue Prism and UiPath for its front-office customer applications, but its back-office operations were looking for a solution more tuned to administration and ERP-related processes, for which they chose Redwood Software.
One of the important reasons for adopting Intelligent Automation is its rapid time to deployment—the ability to fail fast and learn fast, creating a feedback loop to change adoption and practices. This also means needing to be pragmatic if the organization is to scale RPA. Allianz is therefore taking an approach to its RPA selection, which allows challenger providers to compete for and win business as deemed appropriate.
Allianz Is Aiming for Flexibility and Scale from Its Existing Infrastructure
Allianz has literally thousands of people in back- and mid-office functions—from credit risk, analysis, and underwriting—with all of these processes dependent on access to accurate and timely data. Responsiveness to customers and speed to decisions are key for the business in a highly competitive market. Overcoming manual workarounds that slow things down is therefore critical.
| Allianz recognized that different tools are suited to different use cases |
Allianz began working with Redwood Software in August 2016 to explore RPA adoption in the finance function around its core SAP ERP. Redwood’s SAP expertise was a seen as a key differentiator for Allianz as it ensured a rapid resolution for problems.
A proof of concept confirmed they could achieve benefits they could otherwise not achieve using the existing desktop robotic automation tools. Scale adoption was slowed while the distinction between RPA tools was explained to procurement. Allianz finally agreed on a policy for tactically deploying a desktop RPA to support some of its insurance-led processes that worked with legacy applications. But they also realized that more complex business processes using ERP and other associated systems were better suited to Redwood. Allianz therefore took a pragmatic approach in adopting different solutions for different jobs.
| Allianz advice-don't forget the process quality and re-engineering |
The first Redwood robot went into production in November 2016 in a master data reconciliation process for investment accounting, and as part of the broader RPA scaling. Allianz runs Redwood and the other desktop solution side-by-side because they each have strengths in different areas.
It’s important to point out that Allianz got its house in order before using RPA—partly with the SAP implementation but also understanding that they may have to do some process re-engineering. The software may not be enough.
To further stress this point, Rode maintained that process quality was a prerequisite for a successful RPA implementation. However, this often only became apparent when looking at the process during testing, where the RPA system threw up some challenges and required frequent changes. Allianz realized that process re-engineering was needed, which drove up the initial costs.
Encouragingly, Redwood’s robot re-usability and the end-to-end process capabilities brought costs in line with expectations. Since then, Allianz has expanded its use of Redwood into other areas, with success processing structured data, where it is being used to extract, catalogue, and cross reference master data. In highly data intensive sectors like financial services and insurance, there is lots of manual checking and cross-checking of information. Rode sees plenty of additional use cases for this technology in other data heavy tasks. Time saved is another big benefit, since robots are now able to pull data from systems in advance for claims handlers to use directly.
Allianz has set up its own small internal Centre of Excellence with the aim of gaining broader stakeholder buy-in internally, for instance within the $2.4 billion Allianz Technology division, where we understand Redwood has replaced a competitors’ solution. Allianz is set to grow as one of Redwood’s large international clients over the next few years.
Intelligent Automation Isn’t a ‘One-Size-Fits-All’ Approach
Allianz’s challenge reflects what we are increasingly hearing in conversations with customers—applying automation to business process is complex, and does not lend itself easily to a ‘one-size-fits-all’ approach. The process and the desired outcome should dictate the type of automation, not the other way around.
Each technology and provider has its strengths and weaknesses. For instance, RPA is well suited to rules-based automation of structured back-office business processes; API-led automation does well integrating discrete micro-services and applications; desktop automation can neatly knit together common desktop activities. Customers want the best of all worlds, but will often have to compromise in some way.
Compromise isn’t great for the end user, which could help explain why early pure RPA implementations haven’t delivered the expected returns. In fact, our research shows only 58% of customers are satisfied with their RPA implementations today. A source of dissatisfaction is that end users are often attempting to knit together the long tail of discrete business processes and can miss the big picture that’s needed to transform operations end-to-end.
Intelligent Automation to Deliver the OneOffice Vision
Any definition that sets out to objectively represent a subjective term like “digital business” tends to fall short. It will include organizations that don’t quite make the grade or that exclude companies because of an arbitrary failing. However, one of the characteristics that could make an established organization more digital is increasing the speed at which it makes decisions. For an organization to operate effectively online, it needs to be able to deliver decisions to its customers, employees, and partners in seconds rather than hours or even minutes.
Fast decision making for traditionally slow processes like insurance claims or underwriting is a massive differentiator. It enables companies to package solutions in innovative ways, sell products and services differently, and deliver services via electronic channels. It potentially opens new markets and new opportunities with customers. This disruptive speed, at least when it is measured in seconds, demands the removal of manual processes and an end-to-end view of the process operation. It requires better data management, better predictive analytics, and, frankly, a better platform for business, which requires automation.
The traditional view of automation may be focused on cost saving and other productivity gains, but early engagements are demonstrating that it can drive end-to-end process delivery, with automation of the process across the front, middle, and back office. In a digital era, this is the only way to bring about the speed increases and data quality improvements required for a cutting-edge business.
Exhibit 1: The Digital OneOffice Framework
End-to-end process management and fast decision making become more critical as organizations pursue digital operating models like the HfS Digital OneOffice . Automation is at the heart of the OneOffice and should be considered part of a holistic vision for transformation across the organization, connecting the back, middle, and front offices, so that the customer experience is not eroded by organizational silos and poorly aligned processes.
Why Allianz considered Redwood as an alternative
Plug and play robots for ERP-based processes
A clear sweet spot for Redwood is automating ERP-based processes, which are often complex, heavily data centric, and governance-led.
Redwood’s robots target processes that sit directly around the ERP, such as data extraction for the finance team or other departments. These can be legacy processes or manual workarounds inherited from the core system. It means there is a great deal of commonality between organizations, and it therefore makes sense to deliver pre-built robots that can be re-used by different customers.
Redwood has developed 35,000 pre-built robots, which it also maintains and supports, so customers aren’t required to start from scratch and build and maintain their own. Flexibility is built in, so that customers and their partners can also build robots for themselves if needed, or indeed Redwood can build new robots and will support and maintain those that have applicability and can be re-used for other customers.
RPA Customer Experience key to driving new thinking in adoption patterns
In our RPA CX research, we have identified six key service dimensions contributing to overall CX in RPA deployments:
- Out-of-the-box functionality
- Integration and support
- Security, governance, and controls
- Flexibility and scalability
- Embedding intelligence
- Delivering outcome objectives
Redwood’s robotics attempts to address the first dimension—out-of-the-box functionality—via its plug-and-play robotics, which lets users drag and drop robots to build their processes and drive rapid deployment in to the enterprise.
There is also the option to create, test, and monitor new robots within its process studio and then run them at scale. A key feature is the orchestration engine, which manages all the robotic and human interactions, complex process flows, SLAs, and process escalations in a top-down and visual manner and can form the basis of autonomous robotic processing, if customers so decide. This also has a big hand to play in scaling adoption across the enterprise.
Overcoming scaling challenges
Flexibility and scalability are the core of another key CX dimension. Redwood takes a technical approach to automation, which can have its advantages. For instance, it typically sells to IT, which means it is already well-aligned with corporate IT culture and demands for governance and compliance. The challenge and opportunity is to then gain buy-in from the operations side of the business.
Providing tools and services that support the overall CX becomes critical at this point.
Redwood’s orchestration engine provides strong governance and controls for operations managers, who can clone and re-use entire robotic process chains with just a few mouse clicks. An example would be the ability roll out standardized processes from one region to the next, giving significant scaling potential across a federated service center model.
The system also orchestrates the hand-off between robots and humans and vice versa, which can be particularly time saving in complex environments spanning different processes; an example is employee on-boarding, which spans many different departments (HR, IT, finance, etc.).
Supporting this orchestration layer are visualization tools, which provide the ability to monitor and assess processes at the organizational and individual level. This is a fault-tolerant set up, with alerting, dash-boarding, and SLAs built-in. The platform monitors the SLAs and can inform and escalate in the event they are breached.
Actionable data: foundation for the ‘touchless back office’
Embedding intelligence across the enterprise automations is another vital piece of the CX picture. This can really only be achieved through access to good actionable data and an end-to-end view of the processes themselves.
Virtualizing the robots across existing infrastructure helps ensure costs are kept under control. Then ensuring the data has a pathway from the back through to the front office is key, so that users can rely on automations when using desktop applications such as spreadsheet and mail applications, PDF viewers, Salesforce, databases, and other UI-based systems.
| Removing the barriers to accurate data and analytical insight is vital for maximum business impact |
Redwood’s preferred option is to connect directly with the systems and applications, into which the core data is held, although they can also access from the UI if needed. This eliminates complexity and builds a more robust approach to automation that removes a potential point of failure.
The big picture is to eventually achieve a completely autonomous robotic system, with minimal to zero human intervention—in other words, a ‘touchless’ back office. But, this is often dependent more on a customer’s readiness to take that bold step.
This means Redwood’s systems need to be able to self-regulate and respond to changes rather than rely solely on humans for exception management and monitoring. It means building a wider remit than operating as a pure execution layer, requiring investment in machine learning and artificial intelligence, which in turn could open avenues of opportunity around data analysis and discovery at the early stages of the RPA journey.
Advice for Automation Leaders as We Evolve to a Customer Experience-Led OneOffice View
- Align solutions to use cases. What becomes clear when looking through the lens of customer experience is that different solutions are inherently suited to different customer requirements. This should be of no surprise, but understanding the nuances in specific customer contexts can be extremely tricky. This is what we’re aiming to achieve with our RPA CX research.
- Best practice often involves exploiting best-of-breed ecosystems and partnerships. In the digital world, providers can’t be expected to be all things to all people. This aligns itself perfectly to providers specializing to offer expertise in specific areas and use cases. In the case of Intelligent Automation and RPA we are seeing more specialization across desktop, mid-, and back-office use cases. As buyers become more aware of this and more knowledgeable, it could result in clients choosing specific providers to tackle specific activities.
- An excellent customer experience is reliant on people and process change. Allianz serves as an example that customers need to think beyond the tactical choice of a solution and consider how it applies in context of their own operations. Our conversations with customers underline time and again how important people’s domain knowledge and oversight are in creating the best environment for automation and minimizing disruption.
- Business outcomes should be front-and-center of any automation initiative. Taking a narrow view of this simply around cost (FTE) reduction is not enough. Customers often tell us how long it takes to realize the cost savings. But, when looking at the bigger picture, productivity improvements, accuracy, and customer responsiveness tend to stand out far more quickly.
- Data, data everywhere, but how to benefit from it? Often overlooked in automation initiatives is exploiting the newly acquired digital data in a meaningful way—going above and beyond simple data visualization to predict and self-regulate processes in order to become more autonomous. Customers should look to their providers to offer solutions to these questions, which in turn could supercharge their outcomes and customer experience.
- Scale automation investments to get to the next level. Customers tell us that one of the biggest challenges they face is scaling their automation investments. Getting stakeholder buy-in across IT and operations from the outset has to be front-and-center to ensure that once a solution has proved capable at the PoC stage, it is going to be endorsed for further adoption.
Conclusion
Operations leaders in highly regulated sectors, such as financial services, need to be able to rely on automations that are compliant with the existing line of business systems. The cost of ripping and replacing legacy systems is proven to be unnecessary in light of Intelligent Automation. Viewing automation solutions in the context of where they are to be deployed and how they align with existing business requirements is key.
Customers shouldn’t be afraid of working with multiple solutions as long as the solutions have strong governance and oversight from inside the business. Only by testing and learning in this way can progressive companies evolve to their ideal of what good Intelligent Automation looks like. Pushing the boundaries in this way and collaborating with different providers and internal stakeholders is a sensible way to transform their slow, cost-focused businesses into newer and agile digital businesses.
HfS would like to extend a special thank you to Redwood Software for its support of this study. See Redwood positioned at the intersection of ERP, cognitive and ERP and Redwood gets under the hood of the OneOffice.