RPA is dead. Long live Integrated Automation Platforms

The biggest problem with enterprise operations today is the simple fact that most firms still run most of their processes exactly the same way as they did 20/30/40 years ago, with the only “innovation” being models like offshore outsourcing and shared service centers, cloud and digital technologies enabling those same processes to be conducted steadily faster and cheaper.  However, fundamental changes have not been made to intrinsic business processes – most companies still operate with their major functions such as customer service, marketing, finance, HR and supply chain operating in individual silos, with IT operating as a non-strategic vehicle to maintain the status quo and keep the lights on.

Enter the concept of Robotic Process Automation (RPA), introduced to market in 2012 via a case study written by HFS and supported by Blue Prism, which promised to remove manual workarounds and headcount overload from inefficient business processes and BPO services.  However, despite offering clear technical capability and the real advantage of breathing life into legacy systems and processes, RPA hasn’t inspired enterprises to rewire their business processes – it’s really just helped them move data around the company faster and require less manual intervention.  In addition, most “RPA” engagements that have been signed are not for unattended processes, instead, most are attended robotic desktop automation (RDA) deployments. Attended RDA requires a loop of human and bot interplay to complete tasks.  These engagements are not the pure form of RPA that we invented – they are a motley crew of scripts and macros applying add band-aids to messy desktop applications and processes to maintain the same old way of doing things. Sure, there is usually a reduction in labor needs – but in fractional increments – which is rarely enough to justify entire headcount elimination. Crucially, the current plethora of “RPA” engagements have not resulted in any actual “transformation”.

The major issue with RPA today is that it is automating piecemeal tasks.  It needs to be part of an integrated strategy

Real research data of close to 600 major global enterprises show just how not-ready we are to declare any sort of robo-victory. In our recent survey of 590 G2000 leaders, only 13% of RPA adopters are currently scaled up and industrialized. Forget about leveraging RPA to curate end-to-end processes, most RPA adopters are still tinkering with small-scale projects and piecemeal tasks that comprise elements of broken processes.  Most firms are not even close to finding any sort enterprise-scale automation adoption.

RPA provides a terrific band-aid to fix current solutions; it helps to extend the life of legacy. But does not provide long-term answers. The handful of enterprises that have successfully scaled RPA across their organizations have three things in common:

  1. A unifying purpose for adopting automation,
  2. A broad and ongoing change management program to enable the shift to a hybrid workforce, and
  3. A Triple-A Trifecta toolkit that leverages RPA, various permutations of AI, and smart analytics in an integrated fashion.

So HFS is calling it as we see it. RPA is dead! Long live Integrated Automation. And by integrated we mean integrated technology, but also, and all importantly, we mean integration across people, process and technology supported by focused objectives and change management. Integrated Automation is how you transform your business and achieve an end-to-end Digital OneOffice.

Integrated Automation is not about RPA or AI or Analytics. It is RPA and AI and Analytics.

Business problems are not entirely solved by one stand-alone technology but by a combination of technologies. While only 11% of the enterprises are currently integrating solutions across the Triple-A Trifecta, there is emerging alignment. The supplier landscape is also starting to realize that clients will buy integrated solutions (see Exhibit 1) and examples below:

  • RPA products are seeking to underpin AI and data management capabilities. WorkFusion was arguably the first to combine RPA and AI with its “smart process automation” capability. Other subsequent examples include Automation Anywhere with its ML-infused IQBot, Blue Prism announced its AI Lab to develop proprietary RPA-ready AI elements, and AntWorks embeds computer vision and fractal science in its stack to enable the use of unstructured data. What these products having in common is their use of robotics to transform tasks, desktop apps and pieces of processes.  Hence, we need to refer to these “RPA” products as Robotic Transformation Software products which is a far more appropriate description.
  • AI and analytics focused products are starting to embrace Robotic Transformation Software, instead of undermining it. IPsoft launched 1RPA with a cognitive user interface. Xceptor’s data-led business rules and AI-based approach to automation leverage RPA to help extend its functionality. Arago is starting to go to the market where it can help orchestrate RPA capabilities within its platform.
  • Enterprise software products are integrating the triple-A trifecta capabilities in their products. SAP Leonardo aspires to harness the emerging technologies across ML, analytics, Big Data, IoT, and blockchain in combination. It also acquired RPA software company Contextor (late 2018) similar to Pega when it acquired OpenSpan in 2016 adding RPA functionality to its customer engagement capabilities.
  • System Integrators are orchestrating the Triple-A Trifecta across multiple curated products. This typically combines some of their IP and service capabilities. Accenture launched SynOps in early 2019, offering a “human-machine operating engine.” Genpact’s Cora, a modular platform of digital technologies, similar to HFS’ Triple-A Trifecta, is designed to help enterprises scale digital transformation. IBM’s Automation Platform includes composable automation capabilities that orchestrate responses and alerts between Watson and Robotic Transformation Software solutions. KPMG’s IGNITE brings RPA, AI and analytics tools together with KPMG IP and services.

Click to Enlarge

Integrated Automation is not just about Technology. It is Technology + People + Process.

The real point of Integrated Automation is actually to move beyond the tools. Yes, the Triple-A Trifecta offers more functionality, but it still does not work unless you change your business, your people, your processes.  Integrated automation is the effective melding of technology, talent, organizational change, and leadership to get to the promise land. It requires the integration of the Triple-A Trifecta change agents in your toolbox and their application across the original trifecta of people, process, and technology. If you keep throwing technology at a business problem, you will have more technology rather than a solution.

Technology has overpowered the discussion today without adequate focus on people and process:

Source: HFS Research 2019

Integrated Automation is not a Product or a Service. It is a Product and a Service.

Just like we realized that throwing bodies at a problem does not solve the problem, we need to recognize that merely hurling software at business process will not drive transformation. The real genius lies in understanding what to use when and how. The software also needs to come with support and services. Otherwise, we’re just selling more snake oil and magic. Strategic and collaborative relationships of the future will be formed by providers that can consult as a trustworthy advisor and execute as an “extension” of clients’ operations. Enterprises need partners to drive innovation, contribute investment, apply automation and new ideas, and focus on delivering business outcomes – and that requires a combination of services and software. An ecosystem approach with symbiotic relationships between service and product companies is a must-have ingredient for automation to succeed and truly be transformative. It is imminently clear that no one can be everything to everyone.

Adoption is not the measure of success for Integrated Automation. It is about Change Management.

Fifty-one percent of the highest performing enterprises see their cultures as holding them back in the digital transformation journey, while only 36% of the lowest performing enterprises identify culture as a problem to progress. Providers need to offer change management approaches that are agile, measurable, and iterative to be impactful. Scaling up digital initiatives and enabling the right governance models are also critical points. The ability to codify “business outcomes” in contractual agreements, pricing structures, and performance measures is also a vital element to drive change. While there is no nirvana around pricing, it needs to be implemented based on every client’s unique requirements and context. The flexibility to put skin in the game with innovative and non-linear commercial models is essential to drive real change.

Integrated Automation will not be effective with a functional approach. It requires an end-to-end “OneOffice” strategy.

Less than 12% of the enterprises we surveyed have an enterprise-wide approach to automation. This strong focus on task-level and process-level automation remind us that automation often takes place in functional silos, with parallel but unconnected initiatives. The ability to balance task-specific and process-specific pilots and production instances with broader enterprise mission and vision is certainly daunting, but it is precisely what needs to occur to enable scaled and successful automation programs.

The collaboration between business and IT is another crucial issue. While automation initiatives require IT involvement, the programs are generally impacting and enhancing business processes—which requires participation from business constituents who understand the functions in question. The ideal leadership mix, then, is a combination of IT and business. However, our data shows that just one-fifth of respondents have created integrated IT and business leadership teams to grapple with automation strategy and deployment.

Bottom Line: Integrated Automation utilizes the power of AND, not OR! 

We are lucky to live at a time where we have a multitude of established and emerging change agents at our disposal: global sourcing, design thinking, Robotic Transformation Software, AI, Analytics, IoT, and blockchain, among others. But, unfortunately, most of the discussions in the market end up becoming a comparative discussion versus integrative discussion – man versus machine, offshore versus automation, RPA versus AI, consulting versus execution, and so on. These change agents must work together rather than operate in silos to solve real business problems. The power of AND is much greater than OR and Integrated Automation is all about the power of AND. Thus, RPA is dead. Long live integrated automation!

Posted in : Artificial Intelligence, Business Data Services, Cognitive Computing, intelligent-automation, Robotic Process Automation



Leave a Reply

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

  1. Wonderfully written article. Exactly what we are trying to get right at our company. I guess the first step is realizing you have a problem. We are keeping our eyes open and through all the excitement of RPA and AI, trying to have good discussion and planning to ensure that we are not creating more problems than what we are trying to solve. Exciting times though.

  2. Integrated automation is nothing but adding new features, connectors to extend the scope of automation. RPA products are now fighting a battle to be be the first to develop integrators for all sorts of applications that are part of a process. Be it Tableu for Analytics, Abby for OCR or Supervisor to control RPA.
    RPA was ground breaking and just like any other product, it requires new offerings to survive. What Workfusion, AA etc. have done is the standard now- support the ecosystem. The silver lining is such extension of offering will address the question of why not APIs?
    Piecemeal automation is rear and thoughtless. The ROI emphasise will never allow automation in pieces. Why would one spend £5K to automate a minute part? RDA is still valid for the cost and mandatory manual checks.
    RPA is not dead, it is evolving.

  3. Hi All,
    I am working as dot net developer, but I am planning to change RPA UIPATH, how is the future about RPA UIPATH.

  4. Excellent article , as rightly pointed out we at times get obsessed with technology and forget the people and process part .

  5. Relgo Networks, an innocuous company in India is a great example of integrated automation, with role based defined processes automated in an integrated environment and delivered to suit personal needs and specific requirements quickly and deployed in cloud.

  6. Very well written article. I dont think RPA is dead but RPA or automation will continue to drive integrated solutions along with AI/ML, Deep learning, IOT, and Analytics. Deployment of integrated solutions will definitely bring certain disruptions,especially for human resource. Hence retraing and reskilling will be equally important. As Dr. Michael Hammer says change in mindset is required if we want to reengineer business processes. Entire model has to change, roles will change and so on…

  7. Excellent article. Points mentioned are absolutely correct and rightly mentioned, the power is in AND.

  8. The article makes sense. However, there are fundamental issues in the way technologies are adopted in enterprises. With the barrage of technology inflows, there is always the inclination to throw technology at problems and gain the first mover advantage. It is evident when we look at the technology adoption lifecycle of any recent ones, be it cloud/AI/Machine learning. Process and People lags technology adoption in enterprises. There has to be a paradigm shift in the way we consume technology. Plan more, Think more, Deliberate more and then execute rapidly. At some point, CTOs/CIOs have to resist the urge.

  9. Yes, any automation cannot work independently, need continuous train to improve results, means automation is goes well with .AND. rather than .OR.

  10. @Sunil – read the piece, not scan the headline. – We re-badge RPA as Robotic Transformation Software because that’s what it is. only a small portion of it is actually “process automation”… most of it is desktop apps and screen scrapes. PF

  11. This is a good overview of the components of Intelligent Automation.
    The point that it is about systemic transformation is especially well-taken. Radical re-thinking of existing processes will drive the greatest benefits, while functional and financial constructs such as earned-value, phased initiatives will provide the foundation for the deep changes – and rewards – IA will provide.

  12. Reality check and much needed level setting for all the Automation & AI aficionados and experts! Business problems are not entirely solved by one stand-alone technology but by the right combination of technologies.

  13. Brilliant perspective…people + process + technology drive true transformation. Quoting HFS: “throwing bodies at a problem does not solve the problem, we need to recognize that merely hurling software at business process will not drive transformation. Strategic and collaborative relationships of the future will be formed by providers that can consult as a trustworthy advisor and execute as an “extension” of clients’ operations.“


  14. Great article on the usage of RPA by combining technology, process and people which are the three pillars and need to be handled cautiously while bringing any change management. Quality of change and it’s acceptance will play a vital role in any change Management exercise.

  15. Absolutely an interesting article. Standalone technologies are no longer going to solve the problem. They should complement each other and it also open ways for new innovations.

  16. RPA is dead: Phil Fersht and his team at HFS Research are one of the analyst firms that seems to get it right, all the time! This great article declares that RPA – IMHO the over hyped simple answer not to change your way of working, but instead automating what was working 20/30 years ago – is dead. In summary, a more hollistic integrated automation platform is needed. Recommended article worth reading!

  17. Very well written article highlighting the key elements of automation. While technology has always been an enabler but it will never be an end in itself. It all boils down to combinations of change management and various forms of technology (Analytics + AI + IoT + Blockchain + Quantum +…whatever is next)

  18. “Robotic” was never ever a real thing in process automation. To all those who could not and still cannot define exactly what a process is, it perhaps meant something. Because by definition, a process is a series of linked activities that has a measurable customer outcome. Robotic would mean unintelligent repetition of tasks; not intelligently connected activities. Any process expert worth his/her salt would tell you that continuous reengineering at the level of Sub-Processes and their Activities is what would produce Process benefits and not just the mundane automation of Workflows at a Task / Transaction level. Because, intelligence is embedded in the Controls of the Process that is dictated by Governance, Risk & Compliance (GRC). In realtime, roles across the ecosystem should be able to provide Intelligent Engagement and Experience that is Intuitive; not predefined or hardcoded. Thank God, RPA is dead. But what is replacing it? Can we get to the bottom line of how Intelligence is defined, derived and distributed in realtime across devise, channel and mediums in a secure manner to herald the oncoming of a new life?

  19. Well written and completely spot on Phil. This is exactly what we have been working on with our customers at DXC. Trying to change the mindset of the organization and educate them that just throwing RPA at a problem does not create any sort of solution. RPA AND AI AND ANALYTICS AND MORE is really the key to unlock the true value of transformation. It’s time to move into the future and it takes the right technology, the right strategy and the right partner to achieve a better future state.

  20. As you know Phil, I am in full agreement with your main concept and points here. And I think you will concur, I saw this on my radar when we first chatted many years back., Knowing that RPA / RDA was only ever going to reach scale if it was PART of something bigger. I saw this years ago and is why we sold OpenSpan to Pega, and focused on technology to make our advanced Deep Robotics (and now bundled) part of an industry recognized Digital Process Automation (Integrated Automated, OneOffice, Digital Ops or whatever it’s called) platform.

    I will go further. RPA + AI alone does very little because AI isn’t a thing. This is another myth (with some exceptions) thrown out by the RPA vendors to keep their big investors happy and now used as the excuse it’s why you could not scale. What exactly is AI to RPA? You quote Workfusion (AI) but that’s paper automation – just one part of the process that in itself is rapidly becoming digitized. There are so many different AI and Machine Learning technologies built for different purposes and many have had $billions invested to make them unique (using AWS AI for instance to find a cat in the middle of a pack of dogs isn’t going to help predict when a customer might churn or if an email can be automated without a human ever needing to be in the loop). However, the RPA vendors are throwing AI (bolt-ons) around as if it is something bigger and as if it’s going to suddenly transform or digitize a process with RPA. As you say though, there are many other technologies, people and processes (and IT and Biz alignment) that must all come together to solve the problem of 20/30/40 year old legacy. The whole concept of RPA (speed, ROI) is thrown out the window if, in order to use it, it costs as much as automating the process properly. In 2012 and for the last 7 years, RPA has told the world you don’t need any of this, RPA was supposed to be so easy, it’s plug and play. Yeah, right.

    We shall see how this pans out but I absolutely think our go-to-market with DPA (Digital Process Automation), which includes RPA, RDA, AI, ML, no-code, Case Management, Rules engines + about 15 other core technologies, fully “integrated”, is the future. Rather than band-aids, DPA is changing the way the world builds software so you never need band-aids in the future. I think it’s real.

  21. Platforms are poor substitutes for well thought out products – ‘enterprise IT history’ is replete with platforms falling by the wayside. Fact is that despite all manner of marketing hype notwithstanding, there remains an important automation problem to be solved and the current crop of products and services vendors are still unable to get meaningful and complete products to customers.

  22. I love this piece Phil. We believe #RPA can do a lot of good things but needs to be part of a bigger picture if aimed for scale.

  23. Well, this article is quite obvious behind the tl;dr. In my RPA practice we do not even call RPA without ML a true RPA. Any Senior-level engineer is obliged to have ML experience. So it is mostly about how you call it, not what it is.

  24. The problem is to identify real time scenario identification in RPA. The deployment isn’t a easy thing for RPA. It’s not completing helping the process to automate. At the end manual intervention is needed and still needs the headcount

  25. Wonderful analysis of the dead bots. Truly we make a mistake by throwing automation or people on a bad process.
    Intelligent automation with human empathy is what I call it getafix magic potion

  26. Phil,

    In some decade of the 19X0s, we said that we needed to straighten the winding cowpath, not just pave it over so we can drive faster.

    We failed! RPA is dead. Long live IAP.

  27. In that case RPA died a while back. The challenges I have are around dated functionality in RPA tools, platform architecture if not cloud based and interoperability & availability of APIs between systems and apps

  28. Wonderful and succinct article and agree wholeheartedly. Intelligent Automation is the future. RPA is just another tool we use in our holistic Intelligent Automation approach to providing outcomes at kobo.ai

  29. RPA is evolving fast in conjunction with Machine Learning and AI. RPA has to be looked at as part of an overall plan. It will be an enabler in near future.

  30. This is really good analysis of RPA vs Integrated Automation. There is no doubt that without people and process transformation technology alone cannot reap in desired outcome and ROI. Of course, RPA, AI and Analytics together result in real benefit but it also require organizations to build their MODEL correctly which will tell as to how their ACTIONS impact the GOALS.

  31. Well written article. Most important aspects for success of RPA+AI+Analytics is not only the technology but

    1. A good change agent who can prepare proper business case
    2. An integration of Process owners, Policy Makers and Technology team in any organization.
    3. Involvement of CXO level to drive this integration of People + Technology
    4. Most importantly, patience and readiness to give away legacy systems. Things will not change overnight for true transformation.
    5. RPA is not dead, but it will be coupled with AI + Analytics. And it can not be a plug and play solution, it has to be designed well as end to end solution. It will be different solution for different company needs. Patience and readiness of change from top down in company hierarchy will make the company successful
    6. Finally, RPA+AI+Analytics will not need companies for survival. Companies need RPA+AI+Analytics for their Survival.

  32. First of all RPA is by not dead by any measures as suggested by the clickbait title, I can assure that consultancies will make good money on this basic concept in the foreseeable future. There’s seem to be a strong need from a corporate community to gather around vague ideas that provides it with hope and vision for the future, without really articulating what’s the underlying substance. It’s no brainer that business is evolving towards integrated solutions, it’s been discussed years now, the problem is, like with many other concepts, it’s easier said than done.
    Vision for having RPA/ML/AI is nice but it’s not something you can deliver in near future in a standardized manner if systems providers are not involved or even responsible for solutions. As of today enterprise AI is a buzzword, ML space is hugely fragmented and from business standpoint these solutions simply don’t deliver any substantial benefits in a recurring manner.

  33. RPA as known is still evolving and I completely with the Author that we need to change as change only thing that is constant in life but changing for something good is what Integrated Automation all about…

  34. It is Sire.. The Baby is an adult coming into the uninsulated world of Automation where every product can transform the business and I hear you Phil Fersht and I hope everyone else acts upon the fact that you’re research depicts (the bigger circle of tech in comparison to people and process). The new adult RPA called “RTS” along with other college mates would still need to understand and learn the values of life that you’ve called out as ‘OneOffice’. Love it.

  35. @Thomas Kielman – clickbait is only there for people like you who clearly only scan headlines and spout off. We re-badge RPA as Robotic Transformation Software because that’s what it is (or what aspires to be). Only a small portion of it is actually “process automation”… most of it is desktop apps and screen scrapes.



  36. Something that the leaders I have worked with have always envisioned and what I have preached to customers. Certainly glad to hear that from HfS!

  37. Top post. Thanks for the “AI is not a thing”. Even though I work for a vendor using (and touting) AI, I get tired of the misuse and misappropriation of the term.

  38. The focus on technology vs. transformation is disheartening with too many companies declaring victory after “go-live” and losing the plot around realizing value and the heavy lifting of change management. I’ve also seen company after company going forth and buying 50 to 100 licenses of a software tool without any idea how or where to deploy it. Changing the mental model to focus on an integrated platform may help move the focus away from the tech and towards all the elements needed to achieve transformation.

  39. The focus on technology vs. transformation is disheartening with too many companies declaring victory after “go-live” and losing the plot around realizing value and the heavy lifting of change management. I’ve also seen company after company going forth and buying 50 to 100 licenses of a software tool without any idea how or where to deploy it. Changing the mental model to focus on an integrated platform may help move the focus away from the tech and towards all the elements needed to achieve transformation.

  40. This is the beginning of the decline of hype….my only two cents is that soon we will see the acquisition of RPA companies by large software companies so as to embed RPA in their core software.

  41. RPA is basically just a glorified AutoHotKey/AutoIT/whatever wrapped into a marketing bs and friendly looking IDE on top.
    Unattended automation is almost always better off without RPA, for RDA scripting is good enough.

  42. Loved reading the article. All this while we have been hearing of the Triple A Trifecta being the future of automation. With ML & Analytics coming to the fore and being deployed as pilots across, along with Automation, it really paves the path for better outcomes from Intelligent automation than just automation. The graphic on integrated automation platforms is just so true, we can see the convergence and adoption across.

  43. RPA is not completely dead but it was mostly a hype with less than 50% success stories. Integration of AI/ML with RPA is a continuation of that to keep the ball rolling. AI/ML itself has a lot of future but nothing much to do with the user stories with integrated process automation.

  44. Good article .. capturing all the points. Clearly articulating the change through which automation industry is undergoing.

    RPA has gone well beyond its hype curve and now evolving to sustain and provide better value to business. Integrated/ smart/ intelligent solutions are the way forward. As long as companies are in operations they have to adopt to such technologies. Making organisation lean and agile is still the top priority of C-Suit and this evolution is an enabler.

  45. Like the tag Integrated Automation Platform …long overdue move for next generation Enterprise Software. It’s what we pioneered with a no code approach and well proven now over 18 years handling complex case management with constant change supported. BUT very disruptive for existing supply chain putting people first in hands of business. Good forum here plenty debate ….

  46. RPA is not dead but has been mis-sold and abused by consultants, providers and analysts alike! Specifically, robotic automation should not be criticised for what desktop automation does (like reviewing a BMW 5 series and using criticisms of a Renault Cio). They are different, but both have their place in transformation programmes and continuous improvement. Secondly, people should be careful not to repeat same mistakes with platforms, when in many cases platforms are short hand for integrated offerings that combines applications, robotics and ERP as opposed to genuine role-based integrated platform solutions (which incorporates robotic automation).

  47. RPA is not dead, but as the big players are entering that battlefield it will evolve rapidly.

    I was CMO for Contextor, I am now in charge of marketing & communications for SAP Intelligent RPA, and I can tell you that as SAP Leonardo we are working very hard on
    – seamlessly integrating RPA with NLP / chatbots, ML / data intelligence etc
    – providing off-the-shelf pre-built intelligent bots for major SAP products, such as S/4HANA, C/HANA, Concur etc

    RPA is not dead, but long live to Intelligent Enterprise Automation!

  48. Absolutely, the combination of IoT + AI + RPA is set to power the digital enterprises of the future. One without the other puts you in a difficult spot.

  49. Great article. Continuous insight needs continuous adaptation. Otherwise, you’re left with faster but still static automation.

  50. Agree .. there is this trend in Tech teams to indulge in new technology and apply it to some existing process or application without giving a proper thought on its applicability and without clarifying the business case .. it’s very childish behaviour accentuated by this compulsive need to be seen as adopting innovative technologies !!

Continue Reading