This blog post is a quick review of how Robotic Process Automation (RPA) software is deployed as a solution, some five years after it became fashionable!
RPA promised a lot and let’s say that the jury is still out on whether it lived up to the hype. Part of the rationale for writing this was to consider the use cases of some successful RPA deployments.
A review of the various sales proposition statements for RPA reveals a consistent theme: “RPA automates repetitive tasks, allowing users to focus on more added value activities”. For many organisations, the application landscape is a hotch-potch of disparate tools and software; one or two cloud solutions plus a legacy system with generous use of spreadsheets thrown in for good measure.
A knowledge worker often needs a good deal of tacit knowledge to navigate these systems to complete a business process. The data integration between some of these tools and applications consists of knowledge worker’s hands on keyboard – transcribing key information because that was considered the best place to do it, the rationale is that the knowledge worker is “in” the application anyway and so they can “drive” the application by entering some key-strokes. As we all know, this task morphs over time into a significant amount of information.
I believe that RPA is incorrectly labelled, it should be called RTA, Robotic Task Automation. I can’t recall any situations where RPA is used to automate the entire end to end business process. The typical use-case is to automate individual tasks and sub-processes that roll up to provide an aggregated saving, i.e. thousands of instances saving minutes each.
So here’s a round-up of some example RPA projects:
A finance department transforming excel spreadsheet format data from a trading partner’s system into their own ERP. In this case, RPA is fulfilling the role of a sophisticated ETL (extract, translate & load)function. In addition to significant cost savings, errors were reduced and time to complete reduced by 80%.
A back-office of a specialist civil engineering company checking a web-based plant and equipment availability system before allocating resources to a project using a legacy, on-premise application. RPA fulfils a “middleware” function by connecting and updating disparate applications in real time. Result was improved customer experience by faster plant/equipment allocations, admin staff were assigned to higher-value customer-facing services and more frequent reporting and analysis.
A financial services company needed to classify in-bound emails before converting them to searchable PDF before submitting them to a workflow system. RPA is fulfils the tasks of orchestration of a linear process; four separate robot tasks: check email inbox, stage document classification and extraction, convert email to PDF, export it to a Content Management system and then update a workflow system with a link to the PDF and the information extracted during document processing. In addition to fewer errors (e.g. mis-directed documents, incorrect index data), the outcome was improved user productivity and reduced staff turn-over.
A logistics company using RPA to perform shipment status queries on external provider’s track-shipment portals and return results to their own customer portal. RPA provides a fast to implement, code-free, middleware function with the flexibility to use native web services or the web UI. Outcome was faster migration to customer enquiry self-service and at lower overall cost than developing a solution using alternative tools and techniques.
A corporate IT department used RPA to automate user-peak-load, stress-testing of a web application prior to go live. In this case, RPA played the role of test-script system simulating worse-case usage scenario; it’s ability to automate user input into web applications and record results was invaluable in providing the data needed to verify the real-life integrity and performance of system. Customer was able to make go-live decision with greater confidence and eliminated the scenario of hundreds of users on a new system having to endure teething problems.
In the projects above, user productivity and staff morale were cited as key areas for improvement. Reducing staff turn-over levels removes the disruption and productivity dip caused by on-boarding new team members.
RPA can be used to address a diverse set of tasks that are normally performed by humans and in most cases, can also be achieved using APIs and development tools. However, both options are expensive and in short supply. In the correct scenario, RPA can deliver prototype and production solutions quickly and cost effectively.
Please contact me if you’d like more details on how RPA can automate your tasks to ultimately improve processes and ensure a happy and productive workforce!
Karen Peters, Head of Financial Services Business Development, Sefas UK