RPA is Unsustainable

One of, if not the only biggest problem with Robotic Process Automation is that most organisations fail to scale their initiatives and stuck at a very basic level of automation, so they could advance their processes only one step further compared to using VBA macros within an MS Office suite. It shouldn't have been like this. I think we can do better than that, so let's see what is holding us back.

Why do we need RPA in the first place?

Digital contenders appeared in most economy sectors, most are backed up by millions or billions of dollars of VC funds trying to disrupt entire industries with their new, fast and cheap way of doing business.

Traditional corporations mostly focused their efforts to improve their client-facing technologies leaving the rest of their firms under-prioritised, so they ended up operating with decade-old inefficient processes conducted on legacy technologies by a staff that has very limited knowledge on the technological possibilities of improving their own work.

These corporations had to realise that they might attract some customers by showing a shiny new interface they developed recently, but they are losing more when these customers start to realise the disparity between the price and the value they provide. Because there's no way to justify the higher price with slow processes.

They needed to find a way to quickly (and cheaply) improve the back-end, while they could still maintain focus on the front-end of their businesses. After trying a bunch of methodologies, they placed their faith in RPA to be the technology that's cheaper than system integration and that can hold together their business while they find out how they could actually react better to the increasing competition.

They brought in some consultants, dropped a couple of million dollars into it, and waited for the huge "digital transformation" they could experience on the way. After a few quick-wins, they started to scratch their heads as they didn't expect what was coming.

The Pipeline Dried Up

RPA needs structured digital data and well-defined processes. For companies, who have never seriously invested in a digital transformation of their operations, it causes a huge headache after automating a couple dozens of processes (a.k.a. harvesting the low-hanging fruits) with RPA to find new ones they could immediately automate to a relevant extent only by using this technology.

RPA tool providers might have realised this fact in their sales numbers and started to collect a bunch of other relevant technologies around themselves to bridge the "digital gap". To make adoption easier, they built their own connectors to these tools, and started to sell it as a brand new product called "Hyperautomation".

But Hyperautomation is just a buzzword

This is the point when we realise that automation is still not cheap if you need to first kick that dust from under the rug and start sweeping it up. Meaning that companies still need to catch up with all the delayed digital investments they saved the money on during the past decades.

So we must say it out loud: RPA in itself is still not enough!

Given this fact, the role of RPA developers and business analysts are becoming increasingly complex. They all very soon need to be able to design and implement workflows in BPMs, configure document templates in Intelligent Document Processing tools, have a good understanding on the way Machine Learning algorithms work and where it can be applied, understand the results of Process Mining tools, and keep up with other relevant and emerging technologies they could apply.

And that's when RPA is becoming unsustainable

Companies would love to believe that anyone can learn and use RPA with ease, within a short, two-weeks period of training. Reality kicks in: it takes at least 6-12 months of training and mentoring with hands-on development experience using best practices to properly educate a new RPA developer. The industry as a whole needed about 4 years from 2016 to learn how to deliver successful projects constantly and increase their chances for a successful delivery above 50%. Many are still applying subpar methodologies and development conventions.

In addition to learn the core technology of RPA, they'd need to learn the additional technologies that will extend the time needed to train a new developer from anywhere between 12 month to 3 years. This is a time-intensive investment, and these developers will then become more valuable than ever before.

And here comes the free-rider problem. Because you can bet, some firms would love to save the training time and fees by offering a bigger compensation package to these qualified developers. So the current practice of not really raising salaries after a thorough training will make even a below average level of RPA quality unsustainable due to increased employee turnover rates. Not to mention the catch-22 of this situation: if you don't train them, someone else definitely will.

We lost the focus from the real value of RPA

What justifies the cost of the elevated competency that will be required to automate business processes? Well, on the widely popular metric of automation that is the number of FTE or working hours saved, you will find that RPA doesn't worth the hassle. Then why continue?

If you consider that the lack of front-to-back automation will put you out of business against your high-tech competitors in no time, that will most probably give RPA a whole new shine. Still, shortening process transaction time is not really considered a key metric, only a minor side-effect of this technology. As long as it won't change, slow and expensive companies will remain... slow and expensive companies... only with more and more fancy technologies in an increasingly accelerating and less costly world...

To sum it up, keeping only RPA as a standalone technology is unsustainable as you get to automate more complex tasks and start focusing on front-to-back/end-to-end processes. So your team definitely need to evolve in terms of skills and need to get ready to transform entire business units, not just processes.

But due to the undescriptive metrics we commonly use to assess the success of our automation initiatives, some might find this upskilling unsustainable, considering the cost and possible turnover rate associated with that. So some might end up stuck with RPA as a standalone technology. And then due to technical limitations, the pipeline slowly dries up for them in the end.

#RPA #Budapest #Hungary #RoboticProcessAutomation