Organisational investments in automation and AI are rising, driven by promises of improved efficiencies and greater value from new technology. Most discussions focus on assessing whether this investment makes sense: Which tools should we adopt? Which processes should we automate? How much efficiency can we gain?
While these questions are important, our recent research suggests organisations may ignore the more important question: what happens after the technology arrives?
To answer this question, we followed three organisations over several years as they implemented Robotic Process Automation (RPA). While RPA used to be associated with simple software "bots" that automate repetitive tasks, our study revealed a much richer story. In none of our organisations was automation a one-off technological intervention. Instead, automation involved a progressive process that transformed organisational work overtime. This progressive process encompassed not only the evolution of the technology itself, which moved over time from simple rule-based automation towards intelligent solutions that incorporated AI capabilities, but also a gradual evolution of organisational work changes which shifted from incremental improvements in existing processes to dramatic reconfiguration of how work is performed and structured within the organisation.
To capture this evolution, we developed a practical framework that can help managers think not only about how work automation happens with RPA, but also about how to realise value from broader AI investments.
Stage 1: Improve existing work (convergent change)
Most automation journeys start with a straightforward objective: improving efficiency.
In the organisations we studied, RPA was initially deployed to automate repetitive and routine tasks. The benefits were familiar and immediate: faster execution, fewer errors, improved compliance, and greater operational efficiency. Employees were also able to spend less time on mundane activities and more time on tasks requiring judgement and expertise. But pockets of resistance also emerged, mainly driven by a lack of understanding of what the technology entails and fear of unwanted consequences.
At this stage, automation is primarily viewed as a tool that supports existing ways of working. Work structures are adapted to support the technology, but the change is minimal. The focus is on solving specific business problems and delivering measurable operational benefits within the existing organisational structures.
Success at this stage depends on identifying use cases where the is clear alignment with the organisational goals so that automation can create visible value quickly.
Stage 2: Expand automation across the organisation (diffusing change)
Once early benefits become visible, interest grows in expanding them across other areas.
Automation starts spreading beyond its initial applications. Other teams recognise its value, new use cases emerge, and organisations become increasingly willing to experiment with different forms of automation. Work structures begin to change to accommodate the technology diffusion. At the same time, people's understanding of the technology evolves. What was initially seen as a tool for improving efficiency becomes viewed as a capability that can reshape how work is organised.
Our research found that this expansion is not driven by technology alone. It depends heavily on organisational factors such as leadership support, stakeholder buy-in, user learning, and the efforts of automation teams to demonstrate and legitimise the value of automation.
Success at this stage depends on building legitimacy for the technology by demonstrating its value and minimising its risks to gain organisational buy-in.
Stage 3: Transform work (divergent change)
The greatest value from automation often emerges much later, and relates to uses of technology beyond what was originally envisaged.
As automation becomes more deeply embedded, organisations begin to rethink processes rather than simply automate them. Workflows are redesigned, job roles evolve, and automation is integrated into broader digital transformation initiatives. Instead of focusing only on individual tasks, organisations start reconsidering how entire processes should operate.
Significantly, this transformation is also driven by changes in automation technology itself. Several organisations in our study moved beyond basic rule-based automation and started combining RPA with more advanced technologies, including AI-enabled capabilities. What began as simple task automation developed into more sophisticated forms of intelligent automation capable not only of supporting increasingly complex processes, but also offering opportunities to perform work in a completely different way.
Success at this stage comes from leveraging the accumulation of prior effects of automation across the organisation to create opportunities to reimagine work around evolving technological capabilities, rather than simply deploying them to achieve the original objectives of better value and more efficiency.
This finding is particularly relevant today. Many organisations are investing in generative AI and other advanced AI technologies, often expecting transformational outcomes. Our research suggests that transformation rarely comes from the technology itself. Instead, it emerges progressively as organisations learn how to integrate new technologies into work, develop new capabilities, and rethink how value is created.
Three lessons for leaders
Our findings offer three practical lessons for organisations investing in automation and AI.
1. Focus on business problems, not technologies.
The most successful initiatives started with clear organisational challenges which translated into strategic objectives that clearly resonated with people on the ground. Automation was valuable because it addressed business needs which were visible to users, not because it was a novel technology with strategic potential.
2. Treat automation as a change programme.
Technology implementation is only one part of the journey, and it was by far the easiest one. Building trust, developing people capabilities, and creating organisational structures to support the change are the most important mechanisms that enable automation to scale successfully.
3. Plan for evolution.
Automation technologies do not stand still. Neither do organisations. Leaders should view automation as a platform for ongoing innovation rather than a one-off efficiency project. The organisations that achieved the greatest benefits were those that allowed both the technology and their ways of working to evolve together.
Looking Beyond the Technology
As investment in AI accelerates, it is tempting to focus attention on the capabilities of the latest tools. However, our research points to a different conclusion.
The long-term value of automation and AI depends less on the technology itself and more on how organisations progressively integrate these technologies into work. Transformation is not a single implementation event. It is a continuing process of learning, adaptation and organisational change.
The organisations in our study did not begin with a vision of AI-enabled work. They began by automating routine tasks. Yet over time, their technologies evolved, their ambitions expanded, and their work processes changed. What started as simple automation became part of a broader transformation of work.
That observation offers an important lesson for leaders today. The future of intelligent automation is unlikely to arrive as a single technological breakthrough. Rather, it will emerge progressively, as organisations move from improving work, to expanding automation, to reimagining how work is organised altogether.
The most important question leaders today ask may not be whether we should invest in this technology to automate work.
Instead, it may be: What does better work look like, and how can we help our people get there?
Source
Bunduchi, R., Chiș, D.M., Mihăilă, A.A., & Crișan, E.L. (2025). The progressive transformation of work with robotic process automation technology. European Journal of Information Systems.