Automating tasks is not a new phenomenon. We have been slowly and steadily getting accustomed to automations over the last decade or so. Take accessing your music library on your smart phone for instance. So far, we would need to pick up the phone, unlock it, go into the music library, and select the album or song we wish to listen to. This in itself was a phenomenal breakthrough, having an entire music library on one device. But we needed faster and swifter access than the steps I just described. Hence, automation came to the rescue in the form of Alexa, Siri, Google Home etc. Now, all you need to do is say, “Hey Siri, please play Wish you were here from Pink Floyd”, and it does!
Automation is here to stay, and is evolving rapidly. Most of the robotic process automation tools today configure rules into software bots that run non-intrusively on heterogeneous systems. It is way easier to train these bots just by imitating the task user does across underlying applications. The expectations from such RPA implementations are to amplify the business value through significant productivity increase, reduced turnaround time, and improved quality of execution.
But, the underlying data from actual implementations show a different facet. As per a recent report from Ernst & Young, 30% to 50% of initial RPA projects fail1. A Deloitte study suggests that 63% of the enterprises did not meet the delivery deadlines for the RPA projects2. Only 3% of the enterprises responded that their automation program has met more than 70% of its objectives. So what has gone wrong in their automation programs? What challenges the businesses typically face while implementing RPA which is failing to give desired benefits?
To diagnose this better, we first need to understand how do businesses decide and prioritize the processes and subsequent tasks to automate. Currently, most of the enterprises are hiring consultants who interact with subject matter experts (SMEs) and operation agents for process knowledge to do automation evaluations manually. Following are the typical challenges faced by consultants during manual evaluations:
These factors make the automation ineffective and increase the risk to automation program. The success of any automation program strongly depends on the deep understanding of how processes are actually getting handled on-ground.
So, what’s the way forward to unleash the value and potential of automation?
The answer is automated process discovery. This approach gathers the actual on-ground execution data, analyzes it, and backed up with substantial empirical data it creates the automation blueprint. It also not only eliminates the human biases from automation evaluations, but also enables process SMEs to keep the SOPs updated with on-ground innovations or to do audits of the ways tasks are actually getting executed. Needing minimal interactions with stakeholders for process understanding, this accelerates the automation assessment with rapid discovery at large scale.
The automated process discovery would also elicit opportunities of re-designing the subsequent tasks to amplify the value derived from automation by providing visibility into end-to-end process execution based on actual on-ground data.
The automated process discovery is the need of the hour to enable Automation CoE to baseline the as-is process by collecting and analyzing empirical data, and assess continuous improvement and RoI. It accelerates and amplifies the true value of automation.
So, what are we waiting for?