Robotic Process Automation (RPA) reached the peak of Gartner’s hype cycle for Application Services in 2017 and is a circa $500m market today. Depending on which analyst house you ask, RPA and its associated technology market is estimated to grow at a CAGR of 40%-50% reaching circa $9bn by 2023.
It’s surprising what magic some smart marketing and a bit of branding can do. That’s right. The technology and the associated constructs themselves have been around for more than a decade, mostly unnoticed, until a certain genius called Pat Geary rechristened it to ‘Robotic Process Automation’. It’s not surprising that RPA figures in the top echelons of CIOs’ digital agendas and will probably continue as such over the next 3-5 years. However, market surveys claim that up to 50% of RPA initiatives have failed over the last 5 years.
So why have such a large proportion of organisations failed to scale RPA to deliver enterprise grade benefits? Where they have seen benefits, why have the results not lived up to the hype or sustained over time? I think the reasons are not just the usual failings of big-bang tactical change for instant results, but also how organisations have perceived the hype of RPA itself. Here are a few learnings from personal experience and anecdotal evidence from the RPA battlefields.
Automation strategy before shopping spree
Even before the first penny is spent on licensing a tool or platform, IT leaders should really reflect on what automation means for their business and sector. What value can be unlocked via automation in an unconstrained view of the future? Shun the cost-play lens and put on the value creation hat. Cost-base improvement is a fringe benefit if an organisation looks at automation as a holistic enabler of value and performance. Yes, this may be an aspirational view of the future, but it will help calibrate current efforts and mitigate, at least partially, the risk of creating supremely automated bad practices!
Ensure benefit realisation accountability
Yet another reason for automation efforts failing to live up to the hype. This isn’t much different from the all too familiar failing of large-scale change. However, the contrast in results achieved between when this does get done well versus when it doesn’t, is significantly magnified when applied to RPA projects.
RPA isn’t the destination
Automating labour-intensive, repetitive tasks to bring about a reduction in headcount and improve a function’s cost-base and accuracy metrics should be the start. It will create the critical mass of success and showcase the potential for what automation can do for the business. However, stopping at this stage is like letting go of 90% of potential benefits whilst basking in the glory of the 10% achieved. Rules-based automation is relatively easy and is great as a starting point, but the aim really should be to combine a suite of capabilities (RPA, AI, Data Governance) and tools to achieve transformational automation. It is a journey and a marathon at that. It doesn’t pay back in 3 months, but if the strategy is not just cost-play, is long-term and the organisational stamina high, the payback in 24-36 months will live up to the hype, even exceed it.
Plug value leakages
In my experience significant value leakage occurs when process candidates for automation are selected based on flimsy logic. RPA works best for labour intensive, rules-based and repetitive processes requiring minimal human intervention. But not all repetitive processes are alike. Sure, it is easier to automate a back-office HR or Finance process, but does that make it a great process candidate on that basis? How about transforming a front-office process as a pilot, or maybe a supply-chain process such that it directly adds value to your main line of business? It will take additional time, resources and maybe even a multi-tool solution. This brings us back to my previous point. A few percentage points of improvement in a process that is barely visible beyond the department floor versus unlocking value that directly, positively and significantly impacts the P&L statement in year 2.
Isn’t such a tough choice, is it?Author: Shakti Mohapatra