Robotic Process Automation – Big Bang or Go Slow?

With so much talk relating to Robotic Process Automation (RPA), it has become a continuously used word (after all, it is in fact an acronym) for anyone even remotely connected with technology and digital. It is now heard that often that it has simply stopped making people, stop in their tracks!

RPA has the potential to completely transform integral business strategies and processes. In using this technology, companies and governments can streamline processes, improve quality control, minimise errors and optimise resources. The benefits of RPA are so innovative that often business leaders consent to entire organisation wide RPA implementations. CIO’s frequently approve RPA projects in an urgency simply because the benefits of this technology are incredibly tempting.

The Big Folly!

The word ‘process’ is embedded in the very name of RPA. However, not all processes are the same. They differ from each other by virtue of the following parameters:

  1. Time taken
  2. Scale
  3. Repetitiveness
  4. Complexity
  5. Risk

Recognising these differences is the first, and perhaps the most important, step in ensuring a successful RPA implementation. Understanding the differences between the five above parameters, especially the last two, is the pivotal point in which rests the overall success of the RPA.

Complexity: Typically, about 50% of processes within an organisation are manual and repetitive. These processes can be automated using software robots that have been perfected over the last 5 years. Another 30% involve a higher degree of complexity and generally require some form of human judgment. Although these processes can be automated, they still require a strong collaboration between software robots and humans. The remaining 20% of processes are highly complex and are essentially completely dependent on human judgment.

Risk: Put simply, the greater the complexity of a process, the higher the risk of a potential failure and its overall impact within a business.

What Processes Should be Automated?

To understand this question better, we can look to the loan approval process used by a mortgage broker. The typical steps involved are as listed below:

  1. Log client requirements
  2. Gather initial data and documents
  3. Scan documents
  4. Calculate deposit, loan and purchase costs
  5. Determine borrowing capacity
  6. Review loan options from different banks
  7. Advise on 3 best options
  8. Select and customise best option
  9. Gather detailed data and documents
  10. Analyse, evaluate and lodge documents to bank
  11. Bank evaluates and makes a decision
  12. Loan is disbursed

The first ten processes take place at the broker’s end, whilst the remaining two steps reside with the bank. Now, let us analyse these processes in terms of complexity and risk:










If we were to implement RPA within the financial broking business, the first step would be to identify and prioritise the complexity and risk involved in the various business processes. Once this is established, the next logical step would be to start automating the least complex, low risk process.

No Big Bang!

The best approach is to start with the basic processes, succeed in those, and then as the organisation grows in confidence with RPA, bring about more complex processes into the RPA fold. Business processes that involve structured data, simple rules and are highly repetitive, should be the first to use software robots.

In the above example, logging client requirements and scanning basic documents are the first candidates for RPA. This is due to the fact that the rules for these processes are simple and repetitive and generally remain unchanged for the majority of clients. In addition, the potential error risks involved in these processes are far less detrimental in comparison to the possible risks associated with more complex processes, such as the lodgement of documents to the bank.

By implementing software robots to the least complex tasks first, organisations and their RPA partners can leverage on their learnings in order to fine tune RPA implementation as it spreads throughout the remainder of the business. It is better to fail fast whilst automating simpler processes than risk failure when the stakes are high. This scaled approach to RPA is in perfect sync with what the Agile methodology advocates.

Simply put, go slow with RPA!

To find out how Chrysalis can help you on your journey towards Robotic Process Automation, contact us today.

This entry was posted in Blog. Bookmark the permalink.

Leave A Comment