Making the pivot to intelligent automation
Strategies for businesses looking for a technology boost
Diverse applied intelligence technologies are being explored by clients seeking to realize further value. But often it is not the solution in and of itself that produces the results; rather it is the synergies between technologies that proves to be the way forward.
How might this pan out in the real world?
Organization A may have looked to digitize shared service processes to realize greater cost efficiency, with enhanced accuracy, responsiveness and productivity across key process areas such as O2C (Order to Cash) and P2P (Procure to Pay), among others. As they seek to balance optimal operating costs with higher customer quality, they need to look beyond automation solutions.
Another organization, B, happy with the success of the initial attempts at implementing bots, is keen to expand automation to most process streams but also look to reconcile with legacy workflow systems.
Meanwhile, organization C might not have had an encouraging experience from its initial automation initiatives. It is exploring new technologies to boost the 'productivity' of its current workforce and yield new insights.
In summary, different organizations are now turning to a broader set of applied intelligence technologies in conjunction with automation (including RPA). Their objectives are typically to:
- Combine benefits of automation with optimization for a higher ROI
- Glean better insights into data and augment human capabilities
- Scale and institutionalize a portfolio of related Intelligent Automation (IA) technologies
The broad spectrum of IA technologies, however, seems to be posing a challenge for organizations, plausibly due to the recency of their advent and the abundance of choices. Challenges that have been highlighted indicate that organizations are still:
- Building a business case, at a discrete process/ function level, and inherently driven by cost.
- Examining IA technology solutions in discrete elements- or only looking at their integration at the level of functionality- rather than trying to assess the end-to-end impact- especially with regards to data.
- Continuing to deploy conventional governance mechanisms, structures and program coordination methodologies.
Borrowing from the old and the new, it’s possible to address these challenges, starting with the business case. Expanding focus to aspects other than cost - such as multiple KPIs – allows technologies to be considered for various use cases. Drawing up key customer journeys facilitates the way in which the architecture could evolve. A business case is also supported by a good hard look at processes top down to complement the more common granular analysis for deploying an automation application.
The need to allow for experimentation, and multiple pilots is key to a successful scale up during implementation. This means adopting an approach which allows for experimentation and failure, design thinking and advanced Agile techniques. This facilitates more scenarios to be tried in parallel in a short time period, than the conventional linear method of deployment to scale. One could go so far as to suggest using simulation techniques in cases where little or no prior application experience exists in the industry.
At the same time, successful decisions on the subsequent scale up, or the choices between competing technologies require a much more robust organizational entity than a program office. This can start with a CoE, but needs an organization-wide framework that addresses R&R on aspects of not just technology or process but also data, learning and change management.
The upside of automation is the amount of data, and its traceability. Reinforcing the overall data management capability including monitoring and mining can supplement the analysis an organization can do. It also provides for course corrections in the use of the technology.
Given the pace of change of applied intelligence technologies they would interface with (such as cloud, mobility solutions etc.), these suggestions, though not comprehensive, are key to a more successful intelligent automation transformation.