RPA died. Get over it. Now focus on designing processes that deliver superlative experiences.


Seriously folks… there’s the hype, then the excitement, then even more hype… and then the realization that it wasn’t quite what you thought… and then, finally, coming to terms with the fact you’re no longer going to hit that elusive jackpot.  To hear some people still showering us with cryptic unaudited revenue numbers, and from neolithic analysts still parroting their marketing, just spanks of desperation to keep faking a market that simply isn’t there.  Can we just push the off-button on this charade, please?

What, exactly, is “dead” and where are the signs of life… when it comes to process software and enterprise automation?

The RPA that “died” is the poorly-defined “RPA” that got hyped up to create hockey-stick growth excitement for investors. It wasn’t defined correctly, was a mash-up of desktop automation with pure-RPA (unattended back office) and all the deals that got signed were “attended” so weren’t even “robotic”. 

The pieces of RPA that survive are the process orchestration tools (discover, design, automate and mine) that form part of what we see as the evolution towards “Intelligent Digital Workers” which augment human experiences and help with real customer-to-employee intimacy. Let’s also not forget these apps also need to be enterprise-grade, compliant with ITIL and security factors etc. Scale only occurs when the business designs and IT enables… The winners in the future are smart enterprises with leverage technologies to anticipate where their customers are going… often before their customers even know themselves.

It’s been a year since we declared RPA “Dead”… so what’s been happening since?

It’s been nearly a year since we penned our now-infamous blog “RPA is dead.  Long Live Integrated Automation Platforms”. Coming from the analyst firm that first introduced RPA to the world in 2012, this caused quite the stir.  In fact, one of the leading service providers even shutdown its RPA practice as a result and most of the others are left scratching their heads still trying to figure out where the money really is…Since the “dead” post, we’ve seen a swift realization from investors that the RPA “market” was being engineered by a small handful of marketeers attempting a reincarnation of the dot-com bust era where everyone goes nuts over robot butlers and a bunch of naïve enterprise clients who’d been oversold too many RPA licenses that they had any idea how to deploy.

We weren’t helped by a small handful of analysts who really should know better than to pontificate false marketing in exchange for an ego-stroking and glittering robo-stardom they’d never before experienced… and a great big Vegas party that precipitated the most embarrassing collapse we’ve seen in the history of process technology. Many good people had bet their careers on hype, false hope – and blatant lies – and are still on the job market trying to get their lives back on track. In fact, the whole fiasco very nearly destroyed the real market that these tools can help catalyze, if they are allowed time to develop and form part of a broader, integrated solution.  Our recent HFS Top 10 covers this form the view of 300+ current adopters, however, this market is changing very quickly and it won’t make sense in the future trying to put a lot of products in the same “market” that is changing into one that encompasses so much more than basic screen scraping, macros, and process loop recorders.

Instead, we need to focus on the development towards an intelligent digital workforce that help us deliver real customer and employee experiences

What isn’t dead is the fact that RPA created the path (and conversation) to a much bigger market that’s evolving, once you get real about business process issues and the true path operations leaders need to take to make them awesome.  But, if you can’t accept we’re in the early stages of a marathon, not midway through a 110 meters hurdles dash, we can define an exciting future for the world of automation.  But a “bot for every desktop”, or “hyper-automation”? Really, folks?  Can we just start talking again in plain English about what is actually realistic, what works and how we need to change ourselves to get there?  Can we start talking about an Intelligent Digital Workforce?  Can we start looking at how to move from dumb admin bots that keep old process loops and apps stitched together, and how enterprises can invest in intelligent workers that help us achieve much more intelligent interactions and experiences?  Can we focus on intelligent digital workers tuned to deliver (and learn) superlative experiences from processes we have designed to bring our customers and employees together?

The emergence of an Intelligent Digital Workforce is a key component of developing a OneOffice Experience.  RPA creates the foundation, but the next phase is to evolve to Intelligent Digital Workers

This shift toward intelligent digital experiences is a foundational element of HFS’ OneOffice Experience for Employee Experience (EX) and Customer Experience (CX). CX will increasingly be considered an umbrella term for the experience interacting with an entire organization, whether it’s the customer, partner, employee or any other entity. An EX culture is one where people work together shifting from transactional interactions to deeper relationships. Organizations need to ensure they get the balance right; which includes optimizing the use of emerging technology with a robust business case to improve CX to the long-term benefit of the business, getting the right information flows in place, eliciting strategic advantage and ensuring exceptional CX:

The HFS OneOffice Experience typifies how customer, partner and employee experience are coming together to drive a unified mindset, goals and business outcomes.  OneOffice conceptualizes how customer-centric experiences can be designed and supported by end-to-end processes across what we used to term front and back offices. Today’s RPA bots essentially are embedded in the “Digital Underbelly” where they form part of the foundational processing layer for enterprises, while the emergence of smarter tools that can truly augment humans are where the future of RPA lies. Iftoday’s current crop of software providers can develop their bots beyond the current static tools that really just keep old processes chuntering along. Digital Workers are emerging as the enabling technologies that are slowly becoming a critical component of developing CX design and delivering on the experiences smart process operators are designing processes to support.

HFS highlights five important principles of using Intelligent Digital Workers that all companies looking into implementing these solutions need to consider:

The Bottom-Line:  Intelligent Digital Workers are a powerful tool for connecting customer and employee experiences to drive a unified mindset, goals and business outcomes.  The RPA vendors need to get there if they want find their edge in the market

Experiencing a OneOffice enterprise with Intelligent Digital Workers looks different for every organization, but considering the HFS 5 principles will help your company define and execute on a strategy that benefits all of the stakeholders in your ecosystem rather than just having a “tick the box” approach to the technology.  Some companies will focus first on customer-facing, others will start with making internal processes easier and more intelligent.  Implemented well, Intelligent Digital Workers can better connect CX and EX, helping to provide the digital insights and intelligent support that a OneOffice experience requires. This is where the real market for process automation is heading… whether the current cast of RPA characters can make this shift is not inconceivable, but do they have the time and patience of their investors and clients to make the shift?  Time will tell… but not much time!

Posted in : Artificial Intelligence, Design Thinking, intelligent-automation, OneOffice, Robotic Process Automation



Leave a Reply

Your email address will not be published. Required fields are marked *

  1. You need to consider that most of the big companies have old systems , not much integration and processes that didn’t change since the beginning. RPA is a middle term solution, that provides a fast ROI while the full intelligent integration is implemented

  2. I like the HFS no nonsense approach to emerging tech

    Excessive hype -> overselling -> false promises -> failure

    RPA has delivered ROI for many organizations but it has serious limitations too

  3. I’ve been in RPA for 10 years with HPA, recently acquired by Cognizant.

    I definitely thinking there’s been too much hype with RPA with new terms being invented fasted than being implemented – but it’s far from dead. I think companies are finally cutting through the noise to see where the real value is. I covered this in a recent article late last year, and how to properly implement RPA to see real value: http://www.hpa.services/with-rpa-your-only-limitation-is-yourself/

    We’ve had numerous success stories with auditable numbers (in fact auditability is a critical aspect of our business, to ensure the accuracy and quality of the work performed.) In fact, a recent case study highlights just this: http://www.hpa.services/wp-content/uploads/HPA-Case-Study-Texas-Medicaid-Claims.pdf

    RPA isn’t dead, but the overblown and false promises finally are.

  4. The automation market might just take the MS Office route. What started as WordStar, WordPerfect etc. evolved to MS Office (Word) when GUI was perfected. In the RPA context we are at the stage where we are perfecting the GUI. Once this matures, Automation will become a way of life and a tool for other activities rather than an activity by itself…. similar to MS Office.

  5. Hahaha…such a funny post with unproven data .
    I have been in the automation industry since 4 years and even today we are getting hundreds of requests for process automation

    Yes , ofcourse , the scenarios has become more complex to handle with RPA that is where AI , ML and Datascience has came into the picture and RPA turned into Intelligent process automation & Hyper Automation

    If RPA is dying is microsoft , SAP , Oracle like big IT companies are fools so that they can put crores of money to develop their own RPA tools especially in the last half year ….?

  6. As ever an interesting and provocative read Phil Fersht. The title is rather decisive and I can’t say I completely agree with the sentiment as I believe that those who know what they are doing never really saw RPA as a panacea in the first place, it was always one ingredient in the cake, so to say. Too many people didn’t realize this until it was too late though.

    RPA does have a stand alone place under the right circumstances, it can be an extremely powerful tool but often (almost always) is only part of the answer and the other ingredients are being appreciated much more than they have been over past few years.

    What I do completely agree with is the integrated view of human, technology and process elements to design from success backwards keeping the customer and your staff at the core of your journey. This integration, seeing automation and emerging tech as an enabler for change rather than the change itself will help us all achieve much more from business and society.

    1. @Phil…RPA is indeed in trauma from a perspective of not delivering on the promise to the enterprise. A focus on designing, delivering, and deploying automation with the desired human experience as the foundation just may save the day for the enterprise.

  7. @William – – the RPA that "died" – is the poorly-defined "RPA" that got hyped up to create hockey-stick growth excitement for investors. It wasn’t defined correctly, was a mash-up of desktop automation with pure-RPA (unattended back office) and all the deals that got signed were "attended" so weren’t even "robotic".

    The whole goal was to sell as many licenses as possible based on marketing bullsh*t (such as "a bot for every employee"). That market died last year, but some people in positions of serious marketing spend are still pushing that message hard.

    The pieces of RPA that survive are the process orchestration tools (discovery, design, automate) that form part of what we see as the evolutions towards "Intelligent Digital Workers" which augment human experiences and help with real customer-to-employee intimacy. Let’s also not forget these apps also need to be enterprise-grade, compliant with ITIL factors, security factors etc. Scale only occurs when the business designs and IT enables… let’s not forget that.

    The winners in the future are smart enterprises with leverage technologies to anticipate where their customers are going before their customers even know themselves…

    Nice comments btw.


  8. Agreed, Phil. I will be speaking about that very thing next week at the AI for BFSI conference I am chairing. Operating models must change to be mire coherent. Companies still clinging to antiquated hierarchies and fiefdoms must give way to value chains and efficient process flow. RPA is just one of the tools in the tool box.


  9. @Craig – great to hear from you. Let’s face facts, RPA initially was exciting for business operators as it was a tool to stitch together processes and apps/systems without having to go through IT for everything. That only gets us so far and leaves RPA in a place that’s really just making old processes function better (which is OK, but has its value limit). The funniest statement of recent months was when an enterprise client said “our main impediment to process transformation is some obnoxious a**hole who insists on doing things his way”


  10. Always enjoy your perspective, Phil Fersht. Well stated, William Harris. Those that took the right approach to implementing and setting up an RPA program should have a good understanding that RPA is just one piece of true digital transformation. The chocolate chips to the chocolate chip cookie, so to say.

    To me, the biggest takeaway from this article is to think bigger. Whether your goal is to optimize your employee’s cognitive ability to add true value to your organization, reduce risk, or improve customer satisfaction; there are multiple facets to intelligent automation that should be considered. RPA may be a good starting point, but it should not be the end point.

    P.S. I would like a robot butler, can we make that happen? 🙂

  11. @Lindsey -the real value-path “RPA” is taking is as part of a process orchestration toolset that supports process design and execution to deliver great employee/customer experiences.

    The hyped version of the market is the one we have laid to rest. We are pushing the broader focus that helps business transformation, not creates a special robot butler on everyone’s desktop – we had that with Clippy in the 90’s =)


  12. Phil, based on your further comments re: attended/unattended, it’s worth noting that both cases can be successful still.

    I think the problem is mostly with incentives. Clients have a problem, and vendors have sold licenses, upsell with programs, etc.

    Sorry if this sounds pitchy but we worked out how to best align incentives long ago. By selling RPA as a service with a transaction based model (priced per item that runs *successfully* through the automation,) we are required to understand the root of the problems, and apply RPA in meaningful ways that – we fully know – may be temporary. You are now seeing other RPA vendors switch more to an as-a-service model as a result, which is great because clients aren’t the RPA experts, and we aren’t the experts on their business.

    When it comes to attended automation too, there’s plenty of ground to cover with RPA. Say what you will about Clippy – annoying as he was, the challenge Clippy faced was he had to interpret a myriad of possibilities and make sense of them, and he was injected mid-stream to do so. The user would do their work and in parallel Clippy would be interpreting what the user was doing. Attended approaches, virtual butlers, often take this approach – injecting the RPA overtop an existing process. Clippy was useful when you worked through him, rather than we worked through you. By that I mean, reshaping your process around the RPA – which means rethinking the entire process, soup to nuts, leveraging IT where needed so that the user only focuses on what needs their expertise, and the process itself remains consistent to cut out maintenance overhead. It’s hard work and it pays off – but you don’t get there by just buying a license.

    I’m big on analogies, so think about washing dishes by hand. Everyone has different systems… Some scrub and rinse one at a time, others scrub all, rinse all, etc. But along comes a dishwasher and it forces people to change the entire process. If you still scrubbed and rinsed, then you wouldn’t get the value our of your dishwasher. But if you change the process around the machine you become enabled to do other work. Additionally, and importantly, think of the dishwasher: it isn’t a robotic arm that works with a sync and faucet. There’s no AI or ML needed. Instead a custom environment is made for the robot where it can operate efficiently. That’s unattended automation.

    You can extend to attended automation by changing the appliance to a bread maker, where – once more – the process is changed around the machine, and asks the user to operate in a different manner, but frees them up from doing the repetitive work that is not a valuable use of their time. When it needs the user, it chimes, the user provides input, and the machine continues.

    In the same way that kitchen appliances changed how we work, RPA isn’t dead but we do need to adapt how we work to see the real value. And just like how we don’t have a robot chef making us full meals (not yet at least) we should limit the work we expect RPA to perform to the right kind of work.

  13. When are we going to recognize that RPA is in itself never going to be transformational. I liked your recent artical on the RPA manifesto and what real transformation is but 90 percent of the technologies HFS mention that will be needed by organizations to become “digitally” optimized busieness, are clearly not RPA at all.

    Thankfully, most people today are accepting that RPA has its place but only as a piece in a much larger and exciting Intelligent Automation stack, along with a greater and more important set of technologies.

  14. Thank you for this post.
    Unfortunately the company that would almost killed RPA from last yr — had this dismal live event in which the CEO claimed “2019 was challenging and they paid the price”.
    So, thank you for speaking facts. And thank you for remembering us who have been and continue to suffer.

  15. Nice one Phil! No doubt RPA has been both over-hyped and due to this has under-utilized and under-appreciated, IMHO (lack of understanding and change management being main drivers). I worked for two different RPA companies for four and five years respectively, and one of the biggest disservicees to RPA was/is the narrow view that it was a transactional, repetitive process replacement for current processes. From the use cases I’ve seen, it’s an accurate, but very limited view, particularly when we look at a world that’s emerging around intelligent digital workers as you’ve noted. I believe the best view of RPA, is to broaden it as a democratizing integration avenue through the user interface, in addition to ETL and APIs. If you look at machine learning and RPA for example, the cross-section is fascinating. I have a developing theory on an increased propagation of this form of “integration” coming from a couple different angles. I’d love to run it by your expert advice, ping me if you’d like to chat.

    Great read.

    Michael Kern

  16. Death of RPA was NOT natural and is a cold blooded murder. The biggest reason is be blamed by those managers of the organizations who are technically illiterates.
    Many of the use cases were identified wrongly. People still do not use right tool at the right place. Hence the investors are failing to ROI in proportion with investment. Sad to see these conclusions on RPA future. Hope people learn the subjects who intend to drive work. Clients must be guided properly to make this success.
    Thanks and good luck

  17. Hmmmmm…Phil sounds eerily familiar to a recent convo we had. As you know we began pivoting this way a few years back and were blasted by the other analyst because we didn’t have a record button or sit on a desktop and do scripting. Really?! Anywho, look where we find ourselves today…you are probably right it’s time for us to start speaking up!

  18. Some one needs to do a report on what percentage of 25k rpa deals have scaled to 500k deals… that will be a true test of revenue and sustainability … how many existing customers are scaling vs how many new 25k customers you are signing and at what terms and cost of acquisition … I am seeing in the market RPA being dumped on customers for zero cost just to plant a flag… in one case customer was told “keep it… you can pay us when you use it “. … 🙂 scary… seems like the one trick pony is starting to stop running

  19. I disagree RPA is NOT dead as a technological process requirement for ML, AI and eventually unsupervised intelligence (You have to start somewhere!). What is dead are the conceptions that many SALES folks in the industry managed to just nod their heads to questions from naive and hungry clients that dreamed of a silver bullet to fully automate back end system administrative tasks! These tasks developed by a truly dead product ERP which promised much but delivered heavy administrative human robots to manage the processes. RPA should have been positioned as a enabler to relive and automate repetitive back end systems processes (Straight Through Processing, STP) in stead people jumped on the band wagon and immediately drove the application to do things it was not designed for and that is removing humans completely from the process including the “complex” exceptions. RPA in true form should have been named Repetitive Process Automation, as the “Robotic” term lends itself to misconceptions of grandeur things!

    RPA is a step in the right direction towards enabling the Intelligent Workforce and a higher degree of integrations capabilities, allowing heavily invested legacy systems to be connected and managed by a centralized platform. Fully agree that the human workforce cannot be eliminated but can be utilized at a much higher skill set than pounding keyboards repetitively all day long! RPA is valuable, however it needs true transparency and champions that understand and can properly position the tool within the market space for what it is and not for what people think it is! My vote is for Human intelligence to be boosted by Artificial intelligence and not the other way around, otherwise “Skynet” here we come :).

  20. @vinay – Phil wasn’t saying everybody doesn’t want it and understandably you are getting hundreds of requests for process automation – many are reading the same hype. And whether RPA was around or not, who wants to pay someone to hammer away at a keyboard *IF* automating it were easy.

    I’ve been in the software automation industry for 41 years (this year) and with OpenSpan, now Pega RPA, for 15 of them. Large numbers of RPA followers have underestimated (or oversold) on RPA around the complexity of the applications as well as the processes – on the windows O/S (BTW – Mac has Automator built in). AI isn’t necessary to automate something rules based (most RPA is). Most RPA can only automate simple (or lots of services to get it to work). That will never change as long as the windows O/S is the desktop. RPA didn’t turn into Intelligent Automation – even Forrester only weighted it at 8% of the Intelligent Automation wave! Phil is correct.

  21. Is it reasonable to blame RPA technology providers? They’re in the business of selling licenses, not helping companies run more efficiently. As with any business tool, the onus is on the business to make it work for them.

    But, to say the ROI is pure hype is short-sighted. Companies that choose to build it themselves are learning as they go, thus cannibalizing most of the short-term ROI that would have been realized. Long-term, the barriers to success are well-documented. However, RPA service providers are producing tangible results for their clients because they understand how RPA fits in with the solution set, and have the expertise to bring initiatives to maturity.

    The hype lies in the timeframe to realization of that ROI. That is the single, biggest lie in the marketplace today. Who can blame them? You can’t sell as many licenses if you’re honest about how difficult it’s going to be.

  22. @MB Folger – I agree with your point that you can’t blame the tech suppliers for doing their best to sell licenses. However, selling on false hope, fake marketing, and – quite frankly – blatantly wrong business analogies has come back to bite the industry hard. Taking a realistic approach, where they look to sell the right solution to deliver realistic value over a realistic time period is what is required. Ultimately, the focus needs to shift from RPA vendors selling technology, to selling the business solutions the technology enables. This means more investents in consultative folks to educate clients, better partnerships with the quality services firms and consultants, and more investment in actual research that the industry identifies with… not merely paying analysts willing to copy/paste their marketing.


  23. As the world battles the corona pandemic, automation or RPA will be the key to successful business continuity. Now is the time to embrace it and assist employees in work remotely and swiftly. What do you think?
    Marketing manager at http://nuummite.consulting/

Continue Reading