How Johnson & Johnson is implementing RPA
Implementing Robotic Process Automation (RPA) in Johnson & Johnson’s finance department has resulted in standardisation, improvements in work flow and greater accuracy. Sam Horton, Finance Strategy and Operations Director at Johnson & Johnson, describes the firm’s RPA journey, including the challenges that it faced and how these have been overcome.
Johnson & Johnson operates on multiple instances of the SAP Enterprise Resource Planning (ERP) System. This has led to inefficiencies, non-standard processes and challenges when it comes to big automations.
The company saw RPA as a promising technology to overcome these issues and automate processes. It partnered with KPMG on the project because there was not the necessary expertise in-house and RPA vendors tend to work through partners.
Within three months of starting the project the first pilot was launched. The long-term plan is for Johnson & Johnson to become self-sufficient in implementing RPA projects.
The RPA pilot: intercompany payments
Johnson & Johnson started with a pilot focused on automating aspects of intercompany requests, invoice creation and postings.
This area was chosen because it is a relatively straightforward process with significant volumes, but there are also issues around timeliness and consistency, which leads to reconciliation complications.
The RPA process runs as follows:
- When a request for an intercompany charge comes in, the robot opens the email and then the attached standard request spreadsheet template. This template has details of the sending and receiving legal entities, the general ledger coding for the sale and the expense in the respective entities, plus appropriate information to create an invoice which ensures tax and audit compliance.
- The bot checks and validates the data in the spreadsheet and rejects and returns to requester if the request is not completed correctly. When the request is correct, the bot loads it into a purchase order (PO) system.
- Once the request is approved within the PO system, the bot raises an invoice in the sending ERP (Enterprise Resource Planning) system, posting the sale and accounts receivable. It simultaneously posts the request to the expense and accounts payable in the receiving company system. It also ensures a clear audit trail, tax compliance and much more visibility on volumes.
Benefits of using RPA
There are three areas where using a robot to automate this process has brought benefits to Johnson & Johnson’s finance team.
First, it helps to improve the efficiency and accuracy of finance operations. The synchronised posting that is possible with a bot means that intercompany out-of-balances are no longer created.
Second, there are improvements to workflows and processes. RPA means that procedures are carried out with increased accuracy and in standardised timeframes. This has helped to ensure that work flows consistently through the departments using RPA. And the bot can then give statistics on the volume and timings of transactions in a very automated way.
Finally, there is the impact on staff resource. As the bot carries out daily manual tasks, it frees up those managing the bot to work on higher level operations. It also allows members of staff to focus on how to further improve the process. At the same time, people using RPA at Johnson & Johnson are learning new coding skills. Horton says that if a staff member is proficient with Excel macros and is hungry to learn new skills, then they can learn to code a bot.
The challenges of implementing RPA
The RPA journey at Johnson & Johnson has not always been straightforward. In particular, getting the buy-in from IT and working out responsibilities has taken time and delayed implementation.
One potential stumbling block in an RPA implementation journey is that the bot programmer must have keen attention to detail and really know the process. For example, the bot coding has to be specific enough to ensure the validation of data fields replicates what a human would review. This is where controls and human diligence are still of great importance as people still own the process and the bot will only perform as well as it is “coded”.
Another challenge that Johnson & Johnson faced was knowing which provider to use. They decided on Blue Prism, who met Johnson & Johnson’s criteria around functionality and security. However, Horton noted that the market is moving quickly and it is worth doing extensive research into current functionality provided by vendors and their development plans.
Johnson & Johnson were not focused on achieving a particular return on investment with the first pilot and will take stock after they have completed some additional pilots. However, despite the challenges highlighted, Horton is very optimistic about RPA, and the great benefits it provides coming from improved quality and compliance.
The FTE savings arrive when the, currently hidden, rework ceases to be required. It is important to keep in mind the offsetting costs of the new FTE needed for managing the bot and the ongoing maintenance and license fees, etc.
Horton would recommend RPA to other institutions interested in streamlining and standardising their financial processes.
RPA and the future of finance
Horton acknowledges that some finance roles will no longer exist or will change significantly. But new roles will be created, and there will always be a need for humans to manage bots and be accountable. Humans will still be in control.
Horton says that there will inevitably be some resistance to implementing RPA, as with any change. But she argues that this distrust is manageable when presented in the larger context of the overall benefits.
Horton concludes: “I’d rather be the driver of change. This is a really exciting opportunity; don’t let it happen to you, be the driver!”