Validating RPA Processes Using the Process Assessment Questionnaire
In this period of Robotic Process Automation, AI, and IOT, there is little room for errors, delays, or poor technology investments. After determining which RPA tool best fits your unit area, the next question you need to ask is what processes are best suited for automation and which establish the “easy win scenario” and build the broader consensus that RPA must be fully implemented across various work streams to have a greater ROI.
As a technology leader, you need to ask: How does an organization select processes to automate? How do you measure “High Suitability” from “Low Suitability”? Do you need an initial “Guinea Pig” in order to set sail in the right direction or can you begin with a few best suited back-office processes? Which type of back-office processes are best suited for automation? What questions should be asked during process selection? What key measurements are needed?
It is highly recommended when searching for answers to these questions that you use a combination of the Process Assessment Questionnaire (PAQ), which can also be called the Suitability Analysis Questionnaire, along with the Process Selection Model (PSM) – encompassing evaluation through process interaction, maturity, line of business, process intelligence, and implementation.
Process Assessment Questionnaire (PAQ):
- Identification – Organization, business area, business process, systems and software utilized
- Execution Details – Total number of Full Time Employee’s utilized for this process, cycle times used in process (minutes, hours, days, weeks), average handling time (End-to-End), error rate calculated (how often/what percentage)
- Process Details – Standardization (logic defined, consistent, repeatable, or ad-hoc), throughput (Volume of Work, number of Tasks processed by FTE’s)
- Complexity Level – number of target systems, distinct screens, output destinations and input destinations, as well as execution scenarios (i.e. exceptions)
- Implementation Effort – Systems (number in execution), screens (number of distinct screens used), outputs (destination & formats), inputs (OCR/ digital data format), execution and happy path, security access (highly secure, or no access)
Process Selection Model (PSM) with 5 Evaluation Criteria Measurement Points for the Risk and Reward Model:
- Process Interaction – Volume, repetitive interactions, rule complexity, number of systems
- Maturity – Consistent, well defined, measurable, implied knowledge
- Business – Business significance, business priority, regulatory/compliance, business change, current costs
- Process Intelligence – Decision patterns, percentage of exceptions, sentiment analysis, visual analysis
- Implementation – Computer vision, operational costs, capital costs, cognitive technology needs, security constraints
To mitigate Risk with Suitability Analysis, you might want focus on these characteristics during key implementation stages:
- Process Optimization/Simplification – This is RPA in the works, not Business Process Management (like Appian). It’s necessary to evaluate the process optimization points and to be able to simplify and adjust ad-hoc after analysis from RPA implementations. Lessons learned will provide great dividends in how you can optimize processes while still allowing the bots to perform their functions
- Human FTE Impact Effect – Executive level vision on where potential mundane roles might or might not shift too, allowing time to evaluate best fit areas, role modifications, and re-deployment opportunities on-shore and off-shore
- New Exceptions and Workflow Configuration – reducing error handling and confirming smooth workflow configuration is vital:
- System vs. Business Exception reporting – Your RPA Analyst or System Architect will need to run daily exception failure reports to analyze stats like the percent of transactions
- Operational Manual – utilizing this function to coordinate RPA implementations from the RPA team to IT; compliance, audit, and governance is imperative for pre- and post-deployment
- Bottleneck via Manual Inputs – Avoiding the bottle-necking via bot to manual and back to bot; again, this can cause delays, make sure this situation is limited
Interested in learning more about how to validate RPA Processes? View our other articles on our featured insights page written by our executives and RPA experts or contact us directly at firstname.lastname@example.org.