14 Feb

Scaling Enterprise RPA. We’ve Been There, Done That. What Lessons Have We Learned Along The Way?

While most companies have finally realised that when we talk about robotic process automation (RPA), we’re talking about software robots, not tin cans imbued with artificial intelligence; and while some businesses might have even begun exploring their own RPA trials, few organisations have the practical, hands-on experience required to engineer large-scale, organisation-wide RPA capabilities. Because of this lack of internal expertise and skills, many businesses make expensive mistakes unnecessarily in their pursuit of the promised productivity outcomes and transformational benefits of RPA. Forewarned is forearmed, so let’s take a look at the biggest lessons we’ve learned so that you know exactly what to keep in mind as you approach an RPA initiative.

Read the first part of this article: Six Things To Consider When Prepping Your Enterprise For RPA

Lesson 1: Strategy and stakeholders need to be completely aligned.

This means it’s essential to get IT, Security, Business Operations and GRC together around the same table, as soon as possible. Because RPA tools are highly flexible, easy to use, and applicable in several contexts, across functions and departments, it’s easy to let everyone get on with their own thing for fear of trampling innovation, but without centralised control and governance, it’s even easier to get lost in a tangled mess of duplicated efforts and messy, isolated projects, instead of a streamlined, focused automation program.

Dangers of letting everyone get carried away on their own RPA tangent:

  • Robots can be targeted at the wrong tasks, solutions can overlap, which results in a mismatched assortment of tools and techniques, all of which can hinder future scaling.
  • Key risk practices might be applied inconsistently – or overlooked completely, which could have catastrophic consequences in terms of business continuity, formal maintenance schedules, system documentation, IT security protocols, robot inventories and the measures necessary to preserve human process knowledge.

Achieving RPA at scale is best done within a common environment across the organisation, in terms of shared security, risk and quality standards applied under centralised control and governance procedures in order to minimise risk and maximise learning.

Lesson 2: Change champions need to drive transformation throughout the organisation.

People respond to other people before technology. With this in mind – who better to explain and evangelise the positive implications of RPA than trusted individuals across the organisation? Hand picking the right agents of change is an important task. 

These individuals will need to be highly-motivated creative thinkers that can see the opportunities in RPA. They need to be able to communicate and convince their peers that the kinds of tasks that automation targets are those  mundane, repetitive, high-volume grudge tasks that form part of high-turnover roles that usually result in dissatisfied, disengaged employees who would rather be doing something more interesting and more valuable to the organisation.

RPA makes that possible. Human brains of course, are more better applied to non-robotic tasks, like empathetic problem solving, innovation, personalisation, discretionary decision making, in-depth analysis and human-to-human engagement, which is why RPA is important in allowing teams the breathing space to be retrained to focus more of their time on these higher value, higher satisfaction tasks. Robot capability can also be evolved through machine learning to allow for increasingly complex tasks to be automated, which means that it’s necessary to map these tasks against revised roles, processes and systems, to identify and capture further benefits. This doesn’t happen by itself – and your RPA plan needs the technology strategy and the people strategy to operate from the same page. 
Failure to achieve this alignment between people, technology and business objectives will at best, cause delays in training, redeployments and team development, and at worst, lead to uneasiness as employees feel uncertain about their future in the face of automation.

Lesson 3: Start small, remain agile. Everything is about constant communication.

It’s pretty simple to get started with RPA – that’s the beauty of it. One of the best things about the technology is the ease with which businesses can test, learn and grow in the safety of sandbox environments. While it might seem a big thing to undertake a centralised, rigorous program of pilots, tests and reviews, this is the only way the demands of board members, investors, regulators and stakeholders are going to be met. On the plus side, it ensures the RPA program is off to the right start, by expanding on the internal knowledge and expertise required for the RPA program. 

On the other hand, the straightforward ease with which it’s possible to prove concepts with RPA can lead many businesses to think they can fly before they’ve even learned to walk or crawl.

Getting one robot running is pretty easy. Getting hundreds of robots up and running across multiple diverse processes and integrating this automation across the organisation? Much harder. Constant communication with stakeholders and employees is necessary – everyone needs to know what’s going on and how it affects them. 

  • An initial piloting phase should typically take between 1-3 months. A process of wider business consultation needs to happen here, and the different business areas need to contribute their ideas for the RPA program as to what could be automated to make their lives easier and help them work better with others. 
  • Scaling up from piloting then requires the establishment of a formal structure and operating model, centralised control, robust governance, proven business cases and a long-term road-map, which is going to involve  systematically rolling out smaller projects across the wider program to deliver benefits in tandem.  
  • The “go live” phase is usually a 9-16 month program delivering implementations in waves, as per the agreed roadmap. 

Taking a structured, robust, holistic approach is the only way to build a sustainable automation capability that unlocks the significant benefits of enterprise scale RPA.

Lesson 4: It’s going to take a complete IT environment assessment.

Robots need to be managed operationally and maintained technically – just like any other technology system. By thinking of robots as virtual workers, it makes sense that they would require procedures and rules – just like human workers – and when these updated, a change strategy needs to be in place to ensure that robots run as they should, processes are up to date, systems are harmonious and human workers are informed. RPA programs require continuous management – new procedures must be rigorously tested and program leaders need to be open to complexities that may present at any stage. In the same vein, all changes to underlying technology need to be monitored and changes put through their paces by the virtual team.

Because robotics tools are non-invasive and require no integration with existing/legacy applications, it’s tempting to bypass IT and assume that RPA won’t require the input from the enterprise IT technology team. This would be a HUGE mistake, as it is critical to program success to ensure that RPA systems become a significant part of IT’s worldview and strategy for security, reliability, scalability, continuity and fault tolerance across the enterprise. By working in partnership with the IT business function, the Operations team will be able to drive productivity outcomes and transformation benefits faster and harder.

Lesson 5: it’s just hype to think that robots are the whole solution

With RPA, we’re not looking to substitute one component for another – we’re not simply replacing human resources with coded robots. Instead, we’re shifting to a whole new way of doing things. The biggest benefits in terms of transformation comes from the right combination of RPA tools, process engineering and human determination working toward an enhanced experience for everyone in the organisation through greater automation. As such, software robots should take their place as an incremental investment in automation, analytics and artificial intelligence that will form a solid footing for change – driving transformation, modernisation and innovation into the future.  

A short-sighted view that concentrates solely on cost reductions will not yield the full benefits of automation, as few processes can be neatly and entirely automated using an RPA tool alone. It’s necessary to view a process from end-to-end, bearing in mind that it might be you will often need to pull together multiple tools and techniques, including mini bots, natural language processing, data analytics, process re-engineering and a bunch of other tricks and tips. 

At the same time, cost containment can’t be ignored completely and to save costs, some investment will need to happen (for example, digitisation) to support the right combination of tools required. This makes solution design one of the most important, yet complex areas of the automation program –  it is necessary to identify the high value areas for RPA in order to determine which combination of capabilities must be applied in order to create optimal efficiency – this requires a broad focus from the start, and must be backed up by a long-term plan and business use case that will support ongoing tweaks, improvements and facilitate innovation.

Turn it around with these critical success factors

Sidestepping these costly mistakes is all about getting pre-implementation expectations and strategy right. It depends on strong governance of RPA across the organisation that will drive ongoing monitoring, maintenance and improvement. 

These directives hinge on six critical success factors: 

  1. The Automation Centre of Excellence: Your organisation needs a team who is accountable for automation governance, idea generation, skills development, process assessment and organisation-wide support and hand-holding where necessary. This business function should verify that best practice is implemented, efforts are not duplicated and reusable, sustainable RPA tools and resources are developed. 
  2. Choosing the right tech tools: There are many automation tools out there, with their own capabilities, strengths and weaknesses.  Your organisation needs a new kind of capability assessment and solution design that can be tailored to automation projects that carries out continuous analysis of various technologies to optimise the selection of the right tools for the right processes. 
  3.  Structured for success: A strong infrastructure support network is critical to running a virtual environment, with the server hosting and management, product installation and service capabilities necessary to seamlessly support large-scale rollout. 
  4. Control your robot risks: Strict monitoring and security governance is the only way to  ensure all tools and related infrastructure developed in RPA are compliant with IT security measures, regulatory rules and risk policies.
  5. Operational governance: A comprehensive governance framework is necessary to guide the step-by-step RPA implementation through change management, process updates, service demand management, all the while communicating constantly with stakeholders.  
  6. Define and focus on value realisation:  Implementing RPA is about achieving operational efficiency, productivity, improving quality,  and boosting customer satisfaction, and so much more. Successful RPA programs all have one thing in common:  a relentless focus making business value visible throughout the organisation.


We’d love the chance to to pop at your offices to show you how software robots can help your humans work smarter, not harder, while significantly reducing your operating costs. Let’s talk about robotic process automation! 


Share this

Leave a reply