There are many reasons why automation efforts fail. As is evident from the news headlines, large scale enterprises are struggling to make automation work and digital transformation stick. It seems like most enterprise automation, and transformation endeavors are failing, or flailing and firms are unable to quantify real benefits. This article examines the reasons behind automation pitfalls.
Top Ten Reasons Why Automation Efforts Fail
Not having a clear end goal
What are the SMART (Specific, Measurable, Attainable, Results-Oriented, Timely) goals are you striving for? Automation for the sake of automation does not achieve desired or enduring results. Without an overarching vision and a coherent strategy, automation will become a patchwork and will be a bandaid in place of radical interventions which may be apter.
Not understanding the holistic process
We at Skan are keenly aware of how a lack of 360-degree view of the processes and its innumerable variants results in an invisible enterprise. The lack of process transparency and insight leads to suboptimal automation and the bots will require months of hardening to account for the vagaries and variations in processes. The traditional method of process mapping – by an army of analysts drawing boxes and arrows – is not sufficient. Companies need to look at more modern such as process mining which leverages computer vision and machine intelligence to map detailed and nuances process paths to unveil the invisible enterprise. Not knowing the process is a significant factor in why automation efforts fail.
Not choosing the right business processes
From a superficial level, every process looks ripe for automation. But the reality is not every process needs to be automated. There are other ways to optimize the 360-degree process. It is vital for automation and transformation teams to analyze the processes to identify traits that indicate a high potential candidate process for automation. Some attributes to consider are routine and repetitive steps, steps that do not involve judgment or a decision, and high volume workflows, a value stream that includes several applications, and those process that have adequate systems support.
A tool first approach is one of the main factors why automation efforts fail.
The aggressive marketing and sales of top vendors in the automation space have pretty much made the recent automation endeavors a vendor-driven, tools first, the number of bots equals success approach. Of course, at the surface level, the vendors who have introduced a shiny new object and raised hundreds of millions of dollars need to promote their wares and earn their valuation.
However, for enterprises, tools first approach without a strategy, not identifying the right processes, and without clarity of how the process works are doomed to fail.
Using the RPA hammer
When you wield an RPA hammer every process looks like a candidate ripe for automation. The truth is there are several other methods on how to optimize processes:
- Re-engineering a process to eliminate wasted steps and make it function smoothly.
- Re-platforming with a modern system that is flexible and configurable to the needs of the enterprise.
- Standardization of one or more processes to address a particular business need.
- Elimination of a process is a perfectly acceptable solution when there is no clear and present justification.
- BPO (Business Process Outsourcing) is another viable option – mainly if the BPO vendor can automate and deploy the processes based on industry best practices and yet conform to the enterprise requirements.
Number of Bots is not a Success Measure
An executive of one of the RPA platforms has a motto: “A bot for every person.” Really? That is the bane of the industry. Enough said.
Task automation versus process automation
The general hierarchy of how to look at the process model of a company is as follows:
Value Stream > Major Process > Sub Process > Activities > Tasks.
So, when an automation effort is limited to task automation rather than end-to-end process automation, the benefits are fleeting and the ROI somewhat limited.
Of course, task automation is simple, whereas end-to-end process automation is a complicated endeavor.
Missing out on Better Approaches and Solutions
Again, this point goes to using the RPA hammer, and every process looks like a nail. Instead, today solutions like computer vision, natural language processing, deep learning, and other cognitive technologies may be a better approach to some intractable process problems. Or, more simplistically better-hiring practices, or reorganizing the work may yield superior results.
Automation is a part of the overall solution – the biggest pitfall for enterprises is when automation becomes the sole north star.
No training or change management
Work retraining, continuing education, and change management are all-powerful strategies to help employees to learn, grow, and do better work.
Similarly, concepts like Kaizen (continuous improvement) and self-directed workgroups are innovative methods to make the most of worker involvement and iterative refinement.
Lack of guidelines and guardrails on synchronizing work between humans and bots
How do you govern the proliferation of bots? When should one launch a bot or retire a bot? How do you orchestrate the sharing of work between two bots or a bot and a human? Companies are woefully unprepared to the brave new world of the future work where bots, robots, and humans co-exist and collaborate.
Satya Iluri is Vice President of Strategy, Marketing, and Customer Success at Skan, an AI-powered process mining and process discovery platform.