Robotic Process Automation (RPA) in 2017

Thoughts on Robotic Process Automation (RPA) in 2017


Below article includes 3 parts:

1. Looking back at 2016: a very dynamic and exciting landscape.

2. 2017: RPA Express, the first free RPA software by Workfusion: potentially a game changer, and a stimulus for Mergers and Acquisitions (M&A).

3. 2017: towards more intelligence.


Part 1 – Looking back at 2016: a very dynamic and exciting RPA landscape

In 2016, the global RPA licensing market reached about USD 700 million according to Transparency Market Research, almost twice the revenue generated in 2016. And the pace of market growth is increasing. According to SSON (Shared Services & Outsourcing Network), the amount of new adopters is now doubling every 6 months.

The growth in RPA market is also evidenced through the growing number of new vendors, the size of the pools of robots, and by M&A:

  • We started the year with Pega, which acquired RPA specialist OpenSpan to enhance its own Customer Relationship Management (CRM) application and Business Performance Management (BPM) platform
  • A few months ago, Information Service Group (ISG) acquired Alsbridge, and I suspect RPA was a key consideration as ISG seemed to be lagging behind in this field. Some weeks ago, CA Technologies announced its intention to acquire Automic. This acquisition trend is expected to accelerate in the upcoming months (refer to Part 2 of this article)
  • More and more new RPA vendors are entering the market (e.g. Antworks). These vendors tend to be more specialized (by industry or function)
  • The largest amount of robots hosted by a company was about 500 in 2015. This number has almost reached 2000 robots in 2016. 2017 will see even larger pools of robots

RPA is certainly becoming a more established field, and confidence has emerged.

  • The success of Blue Prism’s IPO has demonstrated investor confidence in the RPA market. The trend of Blue Prism’s share price confirms that point (it has more than quadrupled in 10 months!)
  • Initiatives launched by most vendors in 2016 were essentially focused on making RPA more accessible and actionable by their clients: free trial versions (e.g. UiPath, Automation Anywhere), improved training platforms including video tutorials and library of materials, structured implementation methodologies and RPA projects quality standards (e.g. Blue Prism), easier configuration interfaces (e.g. Workfusion)
  • I see less and less clients starting their RPA journey going through Proofs of Concept (POC) or Proofs of Value, but directly implementing pilots. This shows the growing confidence companies have in the technology

Even the way the RPA implementation projects are led has evolved. RPA is more often now included, from the inception, as an integrated part of larger initiatives like ERP, shared service centers or other large transformation projects



Part 2 – 2017: RPA Express, the first free RPA software by Workfusion: potentially a game changer, and a stimulus for M&A

In December 2016, Workfusion announced the launch in March 2017 of RPA Express, the first free RPA software.

  • This was a surprise to me. Firstly, as the RPA market is not yet as mature as to see such aggressive offers, which typically come once software development costs have been paid back. Secondly, as I don’t see the cost of licenses being an obstacle to adoption of RPA for my clients. Education, change management and project rigor are the largest constraints
  • According to Workfusion’s webinar on 22nd December 2016, RPA Express was launched to enable companies a faster and easier journey in the digital transformation (RPA being the necessary first step before cognitive). Given Workfusion’s background in the cognitive market, RPA Express is a way to counter the largest traditional RPA actors in the market (Blue Prism, Automation Anywhere and UiPath) which are more focused on RPA versus cognitive. Blue Prism, the pioneer and pure play RPA specialist, might suffer from such competition, and look for alliances or acquisitions in the cognitive space to keep up. I also understand this launch to be a successful strategy to advertise Workfusion, and gain a seat within the top 4 market players
  • In order to successfully achieve this new product launch, and be able to answer to the demand for free RPA (which is expected to explode), Worfusion will have to strengthen its relationships with partners, and launch important training programs. Note that the consequences in terms of image of any potential product defect will be amplified by the high volume of users, and might affect negatively the whole RPA market (whatever the vendor, if the product does not work, clients might generalize it to the technology itself – “RPA does not work”)
  • RPA Express is a game changer in the RPA market. It will force other vendors to focus on value added functionalities, differentiators and new business models, like integrating consulting, or RPA as a service (managed services). If this results in a reduction of the traditional vendors’ revenues, it will stimulate M&A in the RPA market
  • Overall, this will add some spice in the 2017 RPA market, and this is pushing the market in a direction where it was already moving: towards more intelligence




Part 3 – 2017: towards more intelligence

In this context of vendors looking for differentiators and ways to add more value, 2017 should see an increase in the range of the robots’ functionalities:

  • Integrating more intelligence, while always making RPA more accessible and actionable by the users (refer to Part 1 of the article)
  • Climbing the value chain to more cognitive: adding to RPA (which only “does”), the functions to “think & learn” and “interact” with the environment
  • Moving, in the longer term, to Artificial Intelligence (AI), including the ability to autonomously drive other robots

In a continuation of what we have seen in the second half of 2016, expected functionalities to be added to RPA in 2017 include:

  • Connection of RPA with data analytics systems (big data) to analyze actual data and predict future trends (e.g. share price trends): The robot is able to understand, think, decide and act (e.g. sell or buy shares) on the basis of the outcome of the analysis
  • Combination of Natural Language Processing (NLP) and cognitive: Virtual assistants (e.g. chatbots) able to interact with internal or external clients or operators, to facilitate their work. Chatbots will enable a better / faster adoption of RPA, making it more friendly / interactive
  • Combination of Natural Language Processing (NLP) and Machine Learning: Enabling the understanding of unstructured data received via text or pictures (E.g. unstructured data from incoming invoices using different formats). The robot is able to learn by itself through repetition, build patterns and understand new formats based on what it has learnt by experience
  • Robot configuration assistants (a next step in machine learning): The robot will be able to learn by itself; solely through observing a human executing tasks on a computer, identify the tasks repeated regularly (e.g. daily, weekly), propose to robotize them by itself, and then configure them. When considering the investment in adopting RPA, the largest amount is always the configuration cost, and this functionality will help reduce it drastically
  • Integration with Artificial Intelligence (AI) robots (e.g. Watson, AlphaGo): Given the current state of technology, talking about AI has to be focused upon a specific narrow topic. Let us take the example of the finance function: finding mistakes in finance compliance audits, and summarizing information out of reports spanning a thousand pages. More and more RPA vendors are working on such integration (e.g. partnership between Blue Prism and IBM Watson), but this still remains in its infancy

If you combine the above features into a single robot, this could give us the following example of a use case:

(1) Data analytics application identifies a new or unusual trend (e.g. cost or event correlation, or cost higher than usual);

(2) This triggers the chatbot to interact with the human (e.g. “the cost of X are unusually high, would you like to analyze / adjust?”);

(3) The human takes the decision (e.g. adjust level of prices to X level) and answers the chatbot, which triggers RPA;

(4) RPA performs the modification in the system.



Leave a Reply