Introduction
Robotic process automation (RPA) is the hottest market the IT industry has seen since the rise of smart phones, search engines, e-commerce and social media. While many vendors are over a decade old, the revenue of some jumped 100-fold in just the last two years. But why? It's not just because RPA automates human tasks – or because it uses a little bit of machine learning to process documents and assist workers, or promises to change the nature of work and the value and purpose of enterprise workforces.
In fact, it's all of those things. But it's also because all of the innovations that were brought to the IT markets in the last decade, and designed to solve business problems and improve workforce productivity, actually caused new problems that are now diminishing workforce productivity. In this report, we explain this paradox, and predict where the RPA market is headed.
In fact, it's all of those things. But it's also because all of the innovations that were brought to the IT markets in the last decade, and designed to solve business problems and improve workforce productivity, actually caused new problems that are now diminishing workforce productivity. In this report, we explain this paradox, and predict where the RPA market is headed.
The 451 Take
Large enterprises with myriad task-oriented operations are aggressively pursuing ways to make use of RPA technology. They seek to lower operational costs, reduce manual data entry, and assist workforces to perform prescribed procedural tasks at scale. They also now view RPA as a means to augment workforce skills to make better, more informed, decisions. Over time, these 'digital skills' can help workers become multifaceted, and execute tasks, processes and decisions outside their core area of expertise. This has driven strong demand for RPA technology and services in the last two years, and spawned rapid year-over-year revenue growth for many RPA vendors. The belief that these platforms can also expand into other process-oriented application development markets is fueling speculation of considerable market expansion, and thus is driving higher valuations across the market.
Why Valuations are so High
UiPath was founded 14 years ago, Automation Anywhere 16 years ago, and Blue Prism 17 years ago. Until 2016, their main achievement had been staying alive. Today, the valuations of the top three RPA vendors are so high because, from a technology architecture perspective, RPA platforms are poised to expand into, and/or subsume, other tangential markets that cater to intelligent process-oriented applications development.
So the potential upside is considerable. However, and more importantly, RPA technology is viewed as a new platform designed to fully exploit the promised benefits of ML/AI, because it's applied to augmenting workforce performance – a new 21st-century approach to enable innovation and productivity improvement. Earlier productivity improvement efforts focused on process improvement. For the reasons we explain below, those efforts actually created new business problems that began to diminish productivity. The top three RPA leaders (and myriads of others) picked up on this and ran with it.
So the potential upside is considerable. However, and more importantly, RPA technology is viewed as a new platform designed to fully exploit the promised benefits of ML/AI, because it's applied to augmenting workforce performance – a new 21st-century approach to enable innovation and productivity improvement. Earlier productivity improvement efforts focused on process improvement. For the reasons we explain below, those efforts actually created new business problems that began to diminish productivity. The top three RPA leaders (and myriads of others) picked up on this and ran with it.
Context
The cloud era began in 2006, followed by the smart-phone era in 2007. The great recession began shortly thereafter, and enterprises sought new ways of exploiting innovative and disruptive technology to improve productivity and control costs in declining markets. Add the global shift from 'bricks and mortar' retail to e-commerce, and the digital transformation era was born. Line of business (LOB) leaders within enterprises, struggling to make ends meet, requested resources from their IT organizations to improve business processes for efficiency. ITOs couldn't keep up with demand, so LOBs looked elseware.
Early on, LOBs sought emerging SaaS offerings, but the market was still nascent and few departmental products outside customer relationship and sales management were available. So they turned to workflow engines and business process management suites that offered efficient on-premises tools to develop their own business process applications – many of which were built around electronic forms. As one process was improved, others followed. Within a year, these deployments were designing, developing and running dozens of departmental processes.
Soon, the SaaS market caught up, bringing other LOB and business function offerings to market that included marketing, finance and supply chain applications. LOBs turned to SaaS as a preferred choice – most of which added many new, configurable, user interfaces designed uniquely for different worker 'personas.' ITOs – sensing that SaaS, and indeed, other cloud services, could augment their strapped resources – jumped on board to assist LOBs with cloud-based strategies, and the move to cloud- and mobile-first computing.
Meanwhile, the big-ticket ERP vendors were busy acquiring technology and adding new functionality, products, and applications that yielded an overwhelming amount of data fields and screens to their software suites – also designed to improve workforce productivity. The big-data analytics frenzy soon followed, trying to make sense of all this new data and, of course, all the unstructured data contained in the large document troves found in any large organization, as well as that generated from the growing social media phenomena. To cope, enterprises turned to the hyperscale cloud service providers that now offered limitless compute power and storage – resources that made machine learning for analytics pragmatic and affordable.
But workers had too much data, forms, screens, tools, applications and systems. Data and process flows needed to be interoperable and automated, but weren't. Instead, the workforce had to figure it out themselves. Spreadsheets and cut-and-paste just wasn't cutting it. This spawned a host of new business problems that diminished workforce productivity, essentially negating the benefits sought from many process improvement and digital transformation initiatives. Cloud-based integration technologies emerged. Integration platform-as-a-service (iPaaS) offerings were sought to help solve the problem. Although valuable and useful, they only automated data process flows. Workflow, BPM, SaaS and ERP for the most part only automated business process flows.
What the market needed was a simple tool to automate structured and unstructured data and process flows together, at a granular field and form level to enable inter-application data and process integration between source and target systems. It needed to unburden the workforce, and make sense of things. And it needed to be smart – to interpret unstructured data and make it useful to the workforce and other systems. Once the approach became smart, firms decided to build on what was learned to expose new skills directly to the workforce, to augment their talents. They added some clever branding and marketing, and the RPA market was born.
Early on, LOBs sought emerging SaaS offerings, but the market was still nascent and few departmental products outside customer relationship and sales management were available. So they turned to workflow engines and business process management suites that offered efficient on-premises tools to develop their own business process applications – many of which were built around electronic forms. As one process was improved, others followed. Within a year, these deployments were designing, developing and running dozens of departmental processes.
Soon, the SaaS market caught up, bringing other LOB and business function offerings to market that included marketing, finance and supply chain applications. LOBs turned to SaaS as a preferred choice – most of which added many new, configurable, user interfaces designed uniquely for different worker 'personas.' ITOs – sensing that SaaS, and indeed, other cloud services, could augment their strapped resources – jumped on board to assist LOBs with cloud-based strategies, and the move to cloud- and mobile-first computing.
Meanwhile, the big-ticket ERP vendors were busy acquiring technology and adding new functionality, products, and applications that yielded an overwhelming amount of data fields and screens to their software suites – also designed to improve workforce productivity. The big-data analytics frenzy soon followed, trying to make sense of all this new data and, of course, all the unstructured data contained in the large document troves found in any large organization, as well as that generated from the growing social media phenomena. To cope, enterprises turned to the hyperscale cloud service providers that now offered limitless compute power and storage – resources that made machine learning for analytics pragmatic and affordable.
But workers had too much data, forms, screens, tools, applications and systems. Data and process flows needed to be interoperable and automated, but weren't. Instead, the workforce had to figure it out themselves. Spreadsheets and cut-and-paste just wasn't cutting it. This spawned a host of new business problems that diminished workforce productivity, essentially negating the benefits sought from many process improvement and digital transformation initiatives. Cloud-based integration technologies emerged. Integration platform-as-a-service (iPaaS) offerings were sought to help solve the problem. Although valuable and useful, they only automated data process flows. Workflow, BPM, SaaS and ERP for the most part only automated business process flows.
What the market needed was a simple tool to automate structured and unstructured data and process flows together, at a granular field and form level to enable inter-application data and process integration between source and target systems. It needed to unburden the workforce, and make sense of things. And it needed to be smart – to interpret unstructured data and make it useful to the workforce and other systems. Once the approach became smart, firms decided to build on what was learned to expose new skills directly to the workforce, to augment their talents. They added some clever branding and marketing, and the RPA market was born.
How RPA Works
Essentially, RPA platforms do three things.
- Automate document processing using scripts and some machine learning – mainly in the form of optical character recognition (OCR) and natural language processing (NLP) technology – to take structured and unstructured data from source fields, forms, screens, applications, systems and websites and place it in various target systems without human intervention, for example claims processing.
- Understand what's on worker computers and device screens. Workers access and use many applications, websites, databases and services to do their jobs. Individuals doing the same job may each do it differently, some more productively than others. Using machine learning to examination what they do, what works best, and what doesn't, reveals patterns that can be automated in software to improve accuracy, consistency and productivity.
- As examination perpetuates, learning can accelerate. Tasks soon can become skills that assist and augment the workforce with planning, problem solving and decision making. Bots may act unsupervised, acting autonomously when fully trusted.
The combination of these capabilities enables a range of automations that, until now, have been ignored, overlooked or impractical with earlier workflow engines, BPM suites, enterprise content management platforms, and other conventional application development environments.
Market Leaders
Automation Anywhere (San Jose, California) raised $300m in November 2018 from the SoftBank Vision Fund, an expansion to its series A round that raised a $250m, valuing the company at $1.8bn. It was led by New Enterprise Associates and Goldman Sachs Growth Equity, with General Atlantic and World Innovation Lab chipping in, bringing the total financing to over $500m.
Blue Prism (London) went public in March 2016. It is traded on the London Stock Exchange, and is valued at roughly $1.9bn.
UiPath (New York) secured a $568m series D round in April, led by Coatue, with participation from Dragoneer, Wellington, Sands Capital and T Rowe Price. The funding values the firm at $7.1bn. In September 2018, it closed series C funding, raising $225m at a valuation of $3bn, co-led by existing investor CapitalG and new investor Sequoia Capital. Accel, which led both the UiPath series A and B rounds, also participated in the round.
Vendor Landscape
- Accelirate (Edison, New Jersey) – Formerly AutomateWork, offers its Process Accelerator Framework to automate repetitive tasks as part of structured business processes. Capital sources undisclosed.
- Accenture (Dublin, Ireland) – Accenture acquired UK-based Genfour in April 2017 to tackle the immediate opportunity for ML in IT services. Accenture is publicly traded.
- AntWorks (Singapore) – The ANTstein RPA platform uses fractal science to enable intelligent automation. In July 2018, it closed a $15m series A funding round from SBI Investment, a subsidiary of SBI Holdings.
- AppBus (Philadelphia, Pennsylvania) – The AppBus Experience Platform (AXP) combines workflow automation, RPA, API integration and workspace management into an integral environment. It raised a total of $9m through seed and series A funding rounds from Forte Ventures and Osage Venture Partners.
- Catalytic (Chicago, Illinois) – The company positions its offering as automation as a service built on an AI-enabled SaaS platform. In July 2016, it raised a $11.1m series A funding round led by New Enterprise Associates. Other investors included BOLDstart Ventures, Pritzker Group, Hyde Park Angels, Hyde Park Venture Partners, Sam Yagan's Corazon fund, Lightbank and Chicago Ventures.
- EnableSoft (Orlando, Florida) – Offers Foxtrot RPA platform. Privately funded, it was acquired by Nintex in March 2019, financials undisclosed.
- Figure Eight (San Francisco) – Formerly CrowdFlower, Figure Eight's human-in-the-loop ML platform creates structured training data for ML models. It targets data-science teams that need to automate some of the complex tasks associated with data collection, data categorization and content modernization. In June 2017, it raised $20m in funding led by Industry Ventures, with new investor Salesforce Ventures participating alongside existing investors Canvas Ventures, Microsoft Ventures and Trinity Ventures. Total invested capital is estimated to be $58m. In March 2019, AI vendor Appen acquired Figure Eight for about $300m.
- EdgeVerve Systems (Bengaluru, India) – In July 2014, Infosys created a product subsidiary called EdgeVerve Systems focused on enterprise software for business operations, customer service, procurement and commerce network domains. In August 2015, its Finacle Global Banking Solutions division was transferred from Infosys and became part of EdgeVerve, which now houses Infosys' products and platforms. The EdgeVerve AssistEdge portfolio includes its holistic automation platform, which has an RPA module that can be used for unattended RPA or attended (desktop) automation. Infosys is publicly traded.
- Kofax (Irvine, California) – Kofax RPA automates the exchange of information from application and data sources – including websites and portals, desktop applications and enterprise systems – without coding. It focuses on driving employee productivity, adding insight to the decision-making process and delivering better customer experiences. Thoma Bravo acquired Lexmark Enterprise Software in July 2017 and simultaneously split the assets of that business into two. One portion became a new company under the Kofax brand – it comprises the old Kofax and ReadSoft businesses. The second portion, formerly the Perceptive Software business, was folded into Hyland Software, a separate portfolio company owned by Thoma Bravo.
- Kryon Systems (Tel Aviv, Israel) – The Kryon platform uses visual recognition technology to record and execute processes running on any application (web-based, legacy and desktop) and across multiple applications, without the need for integration connectors or APIs. In February 2019, it raised $40m in series C financing from OAK HC/FT, with participation from existing investors Aquiline Technology Growth and Vertex Ventures. In October 2017, it closed $12m in series B funding, led by Aquiline Technology Growth (ATG) and Vertex Ventures. Total capital raised is roughly $53m.
- NICE (Ra'anana, Israel) – NICE is a publicly traded global systems integrator currently valued at $8bn. One of its offerings is NEVA (Employee Virtual Attendant), designed to help employees responsible for front- and back-office business functions as they work with customer calls or internal business processes.
- Pegasystems (Cambridge, Massachusetts) – Pegasystems positions a version of its BPM technology as an RPA platform and as a robotic desktop automation offering. In April 2016, it acquired OpenSpan, an RPA vendor specializing in automating routine customer service representative tasks. Pegasystems is publicly traded.
- Redwood Software (Burnham, UK) – Redwood Robotics offers a catalog of pre-built robotic services enabling users to visualize, define and execute robotic business processes without the need for complex desktop-based scripting and reliance on development and maintenance teams. Its tooling enables 'the robotic enterprise' by 'robotizing' the manual tasks in the back-office systems that execute record-to-report, order-to-cash, procure-to-pay and other processes used in supply chain management and human resources. The firm is angel-funded.
- SAP (Walldorf, Germany) – the market-leader in ERP software entered the RPA market via its November 2018 acquisition of Paris-based Contextor, and is the first major ERP vendor to make such a move (but not, we suspect, the last). Contextor raised €600,000 ($778,000) in funding in one round from unknown investors in August 2015.
- Softomotive (London) – The company's ProcessRobot includes visual, no-code components, to enable the automation of auditing, error handling, security, log activity, change management and other business functions. In September 2018, it raised a series A financing round of $25m from Grafton Capital.
- Thoughtonomy (London) – The company's Virtual Workforce is an 'as-a-service' RPA platform that automates human activities by emulating the way they interact with technology – applications, systems, tools and devices – and the structured decisions they make. Capital sources undisclosed.
- WorkFusion (New York) – WorkFusion's technology was originally developed at MIT for fraud detection. In April 2018, it raised $50m in series E funding in a round led by Hawk Equity and Declaration Partners, bringing the company's total investment to $118m. Previous investors Georgian Partners, iNovia Capital and NGP Capital also participated. In May 2018, it announced the participation of strategic customers Guardian, New York-Presbyterian and The PNC Financial Services Group, and Alpha Intelligence Capital.
Future
The promise of RPA going forward is that of augmenting enterprise workforces with a variety of digital skills designed to enhance productivity, as well as the quality and consistency of business process execution, decision making and resolution management.
RPA technology will need to get better at using ML/AI to process structured and unstructured documents and data, and how to understand and learn from the behavior patterns of a variety of users and personas as they interact with software and devices. The goal is not only to make the workforce more intelligent but multifaceted, and to perform more tasks and make more decisions across a broad range of business functions by capturing and enhancing human skills.
In the next few years, many of the smaller RPA vendors are less likely to survive as stand-alone vendors. They will be subsumed into other platforms, including application development, digital automation platforms (DAPS) and potentially PaaS and hybrid integration platform (HIP) vendors. Indeed, Kofax acquired Kapow in July 2013, and Pegasystems acquired OpenSpan in April 2016. Accenture acquired Genfour in 2017, SAP acquired Contextor in late 2018, and Nintex acquired EnableSoft in March 2019.
Other vendors are coming to market with converged tooling. Signivaio's DAP includes RPA capabilities. AppBus recently came to market with an integrated digital automation and RPA platform. Even Dell Boomi added support to craft bots within its Flow DAP offering. Some DAP vendors may repurpose their offerings to craft bots of their own, using their existing technology.
Close examination of many DAP vendor architectures reveals that their low-code/no-code techniques are already equipped with RPA functionality. For example, IPSoft launched 1RPA in January, but says it will only use it in its 1Desk product to automate and execute processes rather than compete in the stand-alone RPA market. In fact, we predicted in a report back in August 2017 that DAP and RPA technology and markets will converge in the coming years. Larger DAP vendors will acquire or build their own RPA capabilities, while large, well-established RPA vendors will acquire or build DAP functionality.
Moreover, it's likely some RPA vendors will also be subsumed into machine-learning platforms of hyperscale CSPs like AWS, Google, Microsoft Azure (it's a great way to sell serverless functions), IT leviathans (IBM, Oracle, SAP) and the tool kits of global systems integrators. We expect the pace of similar acquisitions and new market entrants to accelerate in the next two years.
RPA technology will need to get better at using ML/AI to process structured and unstructured documents and data, and how to understand and learn from the behavior patterns of a variety of users and personas as they interact with software and devices. The goal is not only to make the workforce more intelligent but multifaceted, and to perform more tasks and make more decisions across a broad range of business functions by capturing and enhancing human skills.
In the next few years, many of the smaller RPA vendors are less likely to survive as stand-alone vendors. They will be subsumed into other platforms, including application development, digital automation platforms (DAPS) and potentially PaaS and hybrid integration platform (HIP) vendors. Indeed, Kofax acquired Kapow in July 2013, and Pegasystems acquired OpenSpan in April 2016. Accenture acquired Genfour in 2017, SAP acquired Contextor in late 2018, and Nintex acquired EnableSoft in March 2019.
Other vendors are coming to market with converged tooling. Signivaio's DAP includes RPA capabilities. AppBus recently came to market with an integrated digital automation and RPA platform. Even Dell Boomi added support to craft bots within its Flow DAP offering. Some DAP vendors may repurpose their offerings to craft bots of their own, using their existing technology.
Close examination of many DAP vendor architectures reveals that their low-code/no-code techniques are already equipped with RPA functionality. For example, IPSoft launched 1RPA in January, but says it will only use it in its 1Desk product to automate and execute processes rather than compete in the stand-alone RPA market. In fact, we predicted in a report back in August 2017 that DAP and RPA technology and markets will converge in the coming years. Larger DAP vendors will acquire or build their own RPA capabilities, while large, well-established RPA vendors will acquire or build DAP functionality.
Moreover, it's likely some RPA vendors will also be subsumed into machine-learning platforms of hyperscale CSPs like AWS, Google, Microsoft Azure (it's a great way to sell serverless functions), IT leviathans (IBM, Oracle, SAP) and the tool kits of global systems integrators. We expect the pace of similar acquisitions and new market entrants to accelerate in the next two years.
Carl Lehmann
Principal Analyst
Carl leads 451 Research's coverage of integration and process management technologies in hybrid cloud architecture, as well as how hybrid IT affects business strategy and operations. The markets covered in his research include enterprise architecture management (EAM) tools, hybrid cloud integration technology (including iPaaS and API management) and business process management (BPM) software.
Jeremy Korn
Senior Research Associate
Jeremy Korn is a Senior Research Associate at 451 Research. He graduated from Brown University with a BA in Biology and East Asian Studies and received
Keith Dawson
Principal Analyst
Keith Dawson is a principal analyst in 451 Research's Customer Experience & Commerce practice, primarily covering marketing technology. Keith has been covering the intersection of communications and enterprise software for 25 years, mainly looking at how to influence and optimize the customer experience.