Like cloud-native enterprises, many large or long-standing enterprises aim to automate as many operations as possible. As a result, many of them are overly ambitious about process automation goals and are seeking to roll out company-wide digital transformation initiatives. Ambition is good, but many of these initiatives take years to complete and often require ripping and replacement of legacy systems.
Few organizations believe that end-to-end process automation will change the way we think across people, processes, and technologies. Let’s take a look at the three most common process automation mistakes and how organizations work together to fix them.
- Strategic automation initiatives deploy too fast
There’s nothing wrong with being strategic, but thinking it’s too big is a common pitfall of overly ambitious automation projects. Taking on strategic work too soon carries the high risk that an organization will not see business value for a long time. As a result, developers can get completely stuck in forming complex platforms without understanding their use cases.
Instead, start with the most urgent or important projects and try to break down the larger strategic initiatives into components. There is one way to approach this:
- Start with a pilot project: The goal of this project is to define and validate both the architecture and the stack. This pilot project is often set up as a proof of concept (PoC). However, working with that pilot is important to actually learn all aspects of a workflow solution throughout the software development life cycle (SDLC).
- Accelerate to Lighthouse Project: You need to work on the lighthouse project immediately after a successful pilot. This project should have a broader but more realistic range that can be leveraged to showcase the architecture, tools, and value of workflow automation to others and teams within the organization.
- Progress to Massive Transformation: Leveraging lessons learned from lighthouse projects to enable project team members to perform center of excellence (CoE), eliminate inter-team silos, and drive change across the organization To
I call this approach “The art of gradual change.. ” Ideally, plan your entire process ecosystem, including people, systems, and devices working in the background, before working on a large automation project. Start by modernizing the processes that have the most impact on your customers. Then design a transformational approach that fits your business or customer needs, not the requirements of your technology stack.
- Processing automation projects in silos
A step-by-step conversion approach is recommended, but that does not mean “siling” or lack of structure. If each team chooses its own tools, it can be difficult to achieve organizational-wide changes and end-to-end process automation. Technology decisions are a commitment for years, sometimes decades. These decisions and the resulting maintenance do not only affect the current team in the trench.
As mentioned above, the CoE approach helps clear silos in your organization and share best practices on what worked and what didn’t in previous automation projects. Ideally, this group maintains a list of approved tools and frameworks that can be reused company-wide, rather than dictating arbitrary criteria.
In addition to tools, CoE can also maintain start-up guides, project templates, and reusable open source components / libraries that your team can leverage. In addition, you can act as an advocate of automation by running a community and raising awareness of new automation initiatives within your company. Within this framework, more teams can be inspired by the potential for automation within the department.
- Cannot adopt microservices architecture
One of the key factors to deal with is how to build software in-house. Adopting a microservices architecture for a legacy company is not as easy as it sounds. Legacy systems are often introduced that make it difficult to remove the seat.According to one estimate, there are still more than 200 billion lines Code written in COBOL, A programming language decades ago. Large-scale replacement of these systems can cost $ 4-8 trillion (or more).
That’s where the step-by-step conversion approach comes in handy. For example, many companies place RPA implementations on top of legacy systems (such as systems written in COBOL) for surface-level automation. A good approach to these scenarios is to modernize in three main stages.
- Coordinate all these RPA bot-driven local task automation along with your end-to-end business process
- Sunset these bots one by one in priority order
- Invest in rewriting the underlying business logic as a microservice. Microservices can be retuned along with end-to-end business processes.
The advantage of microservices-based automation workflows is that it enables a decentralized architecture where each team “owns” their own separate processes. If a problem occurs in one process, you can easily control and fix it. From there, process engines can “promote” these microservices-based processes throughout the organization and integrate them in meaningful places.
In summary, end-to-end process orchestration does not occur in a vacuum. Projects need to be started on a small scale, involving stakeholders throughout the organization. Without a clear pilot project, large-scale strategic automation efforts can fail to prove business value. Organizations can work together to define priorities, develop best practices, and deploy the necessary technical changes to ensure successful end-to-end process automation.
Sign up virtually to learn more about process automation CamundaCon 2021 The event from September 22nd to 23rd.