Cutting costs, not corners: Transforming processes with automation
Gary Thompson, Transformation Consultant and member of the Definia Associate Community, explores how technology is reshaping automation to enhance processes and drive cost optimisation.
For practical tips on starting your cost optimisation journey, check out the first and second instalments in this series.
Think you can fix it with technology? Jumping straight to automation can be tempting, but it’s a mistake to digitise broken processes. As discussed earlier in this series, automation doesn’t fix inefficiency, it accelerates it. Process improvement must come first.
Decoding the automation landscape
New technology is constantly emerging, and process automation is at the forefront of that change. Companies are increasingly using AI to streamline operations, and automation is getting faster, better and easier to use, thanks to new tools. Let’s unpack some of the technical jargon within the automation environment.
Bots vs. brains: Robotic Process Automation (RPA) and agentic AI
RPA, or ‘bots’, and agentic AI are different software-based approaches to process automation. RPA is great for transactional, repetitive tasks based on rules, such as data entry or processing invoices, but it’s not flexible enough to handle more complex jobs requiring judgement or additional context.
In contrast, agentic AI operates more similarly to humans, with the capacity for learning and planning. It adapts to changing conditions, making decisions and corrections without human intervention, reacting to real-time data. This makes it well-equipped to handle complicated, evolving processes that need smart behaviour.
Low and no-code automation solutions
Bots and AI focus on what the automation does, whereas low and no-code solutions refer to the way an automation is built. With no-code platforms, teams can build, test and run workflows without developers, using visual tools instead.
Low-code solutions work similarly but are better for more technical users with some coding skills, and are suited for complex, scalable solutions that need system integration or custom features.
Orchestrating end-to-end automation
Beyond basic bots and scripts, there are Business Process Management Systems (BPMS) tools to help map out task flows, automate repetitive parts of a process and ensure everything runs smoothly. In addition, you can use iPaaS to link different apps, document automation to direct information where it needs to go, and cloud orchestration to scale it all. Together, you’ve got everything you need to fully automate processes from beginning to end.
The power couple of automation: RPA and agentic AI
Technology is becoming smarter and more adaptable, but the core aim of automation remains the same: faster processes, fewer errors and lower costs. Machine learning now enables straight-through processing (STP) — fully automated financial transactions without manual intervention.
By combining RPA with agentic AI, the result is automation that’s both efficient and intelligent. While RPA automates repetitive tasks, agentic AI brings decision-making and context awareness, enabling it to handle complexity, correct errors and improve over time, supporting continuous process improvement. For example, RPA can extract customer enquiries from an inbox, while agentic AI analyses tone, chooses suitable responses and flags issues for human input.
This combination processes both structured and unstructured data, creating flexible end-to-end automation that reacts to live data. However, organisations need strong governance, clear plans and a workforce ready to shift from managing people to managing processes.
Effective automation needs people, not just technology
Effective process automation requires more than implementing tools; it’s about preparing your organisation and adapting to change. Success hinges on how people collaborate and add value in this new environment. Key focus areas include:
- Redefining roles. Teams should guide tasks rather than just executing them. Managers must shift from making decisions to leading workflows. Empower internal advocates to drive adoption, resolve challenges and share success stories.
- Design for flow, not hierarchy. Often, processes (and therefore automation) cut across different teams and functional silos. So, build processes that follow the work, not the org chart.
- Ensure clear governance and ownership. Who oversees automation? Is it Operations, IT or a combination? Is there one clear process owner, bridging silos from start to finish?
- Engage the workforce from the start. Explain the ‘why,’ let employees shape the ‘how,’ and you’ll get more buy-in and smarter solutions.
- Be modular and stay flexible. Automation should grow with your business, not restrict you to rigid workflows.
TLDR: Success factors that matter
It may seem straightforward, but a few key areas can make or break your automation success. Gartner says 40% of agentic AI projects will be cancelled by the end of 2027, so the correct setup is crucial. Here’s how to approach it:
1. Fix first, then automate. Cleaning up processes first sets the stage for better performance and cost savings.
2. Take time to identify the right processes for automation. Not everything can, or should, be automated. Focusing on the best-fit tasks helps deliver quick wins and sets the stage for scaling automation effectively. Typical goals should include:
- Reducing costs or headcount
- Freeing up capacity for growth
- Streamlining tech systems
- Enabling straight-through processing with minimal human input
3. Choose the right automation tools. Select tools that are easy to use, fit the task, scale across the business and connect with your existing systems, such as your ERP.
4. Setup clear governance and a long-term strategy. Automation needs ongoing oversight, accountability and flexible solutions that can adapt as your business evolves.
5. Address staff concerns and take them on the journey. Job loss fears often arise when automation or AI is mentioned, but real success depends on your people! Automation works best when your workforce is ready to adopt and grow with it. Build a team that blends BAU expertise with technical skills, communicate clearly and make them feel part of the solution, not sidelined by it.
6. Get IT and the wider business working together. Process automation isn’t just a tech project or a business initiative, it’s both. So, everyone needs to understand the tools, how they differ and how to effectively use them together.
7. Don’t rely too heavily on solution providers. Retain in-house ownership and bring in partners selectively, ensuring your business stays in control and avoids long-term dependency. Establish a Centre of Excellence (CoE) for Continuous Improvement to centralise automation efforts, refine operations, enhance controls, and build the skills needed to support and scale change across your organisation.
Make technology your competitive advantage
The best tech transformation projects begin with clarity: clear processes, clear goals and clear ownership. Whether you’re optimising operations, implementing new systems or scaling automation, we help you avoid costly missteps and accelerate value delivery. Ready to unlock your potential? Get in touch with the team today.
