The challenge of IA implementation

Leaders tend to overlook the fact that implementing intelligent automation means a full commitment to a digital-first operating model, in which technology is supported by people, and not the other way around.

Robotic process automation  (Alexander Supertramp/

By now, government leaders see the promise that intelligent automation (IA) offers in cutting costs, improving efficiency and moving talented people off of rote tasks so they can perform higher skill work.

And there seems little doubt the use of IA will only expand since the release of a recent Office of Management and Budget memorandum that specifically encourages agencies to use such technologies as robotic process automation (RPA) to "reduce administrative tasks." So the future for automation is bright.

But with the growing momentum for IA solutions comes a renewed focus on implementation. The findings of a recent survey my firm conducted of mostly commercial entities mirrors what I'm seeing among federal agency leaders.

Fundamentally, these leaders tend to overlook the fact that implementing intelligent automation means a full commitment to a digital-first operating model, in which technology is supported by people, and not the other way around. What's needed is a comprehensive IA strategy that goes beyond just automating a legacy system, for example. Piecemeal efforts designed to merely cut costs of legacy processes -- with no overall plan -- won't really move the needle.

The survey and my own observations reveals that while many leaders have high expectations for the potential of intelligent automation, they face a range of obstacles to implementation. One obstacle is the simple fact there are often not enough resources to implement IA solutions. Another is the difficulty of clearly defining goals for IA deployment and accountability for results -- some agencies just haven't developed a comprehensive, agencywide strategy for IA that matches their overall mission. Still another obstacle is concern about the impact on employees.

On the latter point, I believe that as IA adoption matures, humans and virtual robots will work side by side. Obviously, robots will be able to process data far faster than humans, but human judgment will be required to define the problems that need to be solved and prioritize the solutions.

Agency leaders should also think about creating centers of excellence to develop technically adept employees and recruit specialized talent as needed, especially as agencies advance from RPA solutions to true cognitive automation where artificial intelligence is employed. This model can also address other concerns leaders face when implementing IA, such as governance, security, demand management, program management, talent management and training.

Here are a few other considerations for agency leaders as they plan for IA implementation:

  1. The use of IA is potentially transformative and the agency will need entirely new operating and business models, requiring long-term planning. Obviously, complete buy-in from leadership is required.
  2. Formulate a comprehensive approach to automating the service delivery model including centers of excellence, shared service centers and business and self-service providers. Approach IA spending holistically across all technology platforms. Before investing, develop a solid business case.
  3. Before automating, think deeply about the agency's mission and pinpoint which processes should be automated first based on what's likely to produce a healthy ROI. The specific technology solution, while an important enabler, should be a secondary consideration.
  4. Consider the "operating model" in all of its forms. Operational and technology infrastructure, organizational structure and governance, and people and culture are all critical to deployment, especially their impact on core business processes. Measurement and incentive systems should change to align with operating model disruption.
  5. Think about ways to transform from within the agency while maintaining uninterrupted operations.

Implementing IA to have a meaningful impact on an agency requires much more than solving a technology puzzle or automating a single process. Like any successful transformation of operations, it takes planning and strategizing, complete support from leadership and wise investment in technology. If agencies are up to the task, the payoff can be profound.