Understanding why government automation projects fail

Access to government services has improved with new technology, but despite the millions of dollars invested into digital transformation, agencies remain infamous for being slow and inefficient. 

Even when government IT budgets increase, citizens are still complaining about long wait times to apply for or access benefits and secure essential documents like a driver’s license. However, with the recent White House Blueprint for AI Bill of Rights guiding how automated systems should provide meaningful impact to the public, government agencies at all levels can be bullish about new artificial intelligence investments that improve efficiency and accessibility.

Yet, government CIOs still need to be wary. Research from McKinsey suggests that 70% of digital transformation projects fail, meaning the public purse is funding useless technology.

According to a 2022 ABBYY, a major reason behind the failure of automation upgrades is that the “goals of the automation project were too vague,” suggesting agency leaders aren’t sure what they are trying to achieve through automation. 

Understanding where to start with digital transformation

With unclear goals at the heart of failure, it’s imperative that government agencies take a new approach to instigating change.

A recent Deloitte report on government tech trends calls on agencies to upgrade legacy IT systems but avoid replicating bad business practices at scale. Assessing what is bad and where improvements are needed is an obvious starting point, but how can agencies determine priorities? 

IT consultants will often sit with managers to determine where problems lie and which areas are ripe for intelligent automation. However, this approach is always based on limited information about how processes actually work. In fact, a recent study found that 60% of employees admitted to diverting from their organization’s standard workflow for better efficiency, mainly to improve customer service. 

No matter what process agencies are trying to improve, only a data-driven deep analysis will provide a true, accurate and unbiased picture of operations. By digging into workflows, process intelligence identifies exactly how each process is accomplished, how systems and people interact, where delays occur and even where employees deviate from the norm. Dashboards provide visibility and create key performance indicators to guide prioritization, giving managers a factual understanding of which areas will benefit from intelligent automation.

Process intelligence to prevent automation failure

Process intelligence features a two-pronged approach to workflow analysis: task mining and process mining. Process mining scrutinizes back-end systems, examining event logs to detect root causes of variations. Task mining focuses on the user, analyzing employees’ inputs like keystrokes and mouse clicks and the sequence of steps they take to complete a task. By having data at an individual level, agencies can determine who does what and why and how best to standardize processes or offer additional training. 

Once agencies have a detailed picture of how their processes actually work, the next step is to make changes toward streamlining the workflow for optimum efficiency. This is where process simulation comes in. Advanced process intelligence services will include a simulation capability that evaluates the impact of hypothetical changes in processes. This analysis allows agencies to gauge how adding staff or technology to address a bottleneck might impact citizen and user experiences before they make investments. It also allows them to continuously monitor processes for variances, adherence to standards and opportunities for better solutions.

One key area ripe for process improvement is digital onboarding. Acquiring a new ID, registering a vehicle or enrolling in school or a health care plan all require lengthy processes and complex identity proofing procedures, leading to one-in-five government applications being abandoned, according to an ABBYY survey. 

Agencies need more seamless onboarding experiences, but they must also incorporate speedy mobile document capture and identity proofing with real-time biometrics. The survey showed the top technologies being implemented to make these improvements were process intelligence (21%) and intelligent document processing (21%). Respondents reported that reducing abandonment by 50% would increase revenue by 20%, allowing for more money to be allocated to other government services and staff.

Public expectations of good experiences are higher than ever. Agencies have an opportunity to change their reputation with the right mix of technology and human interaction, but they must start with the right approach and ensure automation projects succeed.

Brian Hettinger is senior director, mobility at intelligent automation company ABBYY.