I wanted to take a moment to address a common misconception that has been presented regarding Robotic Process Automation (RPA) and Intelligent Automation (IA). Many talk about how these technologies have the ability to solve problems in your business with efficiency, cost, and compliance. That is the wrong way to think about automation.
When processes are automated, they can and will ensure that things happen rapidly and in a consistent manner. That part is true. However, IF a process has issues, we will have only become efficient at being inefficient. Therefore, before you begin defining and coding your RPA bots and automation systems you must ensure the process is clearly understood, documented, validated. What does that mean?
While many business groups may work in silos, most processes do not. They have dependencies on systems, people, or even other processes. Therefore, when we think about our process, we need think more than just about our own space. We need to understand the up and down stream relationships between our process and where it hands off to others.
Example:
A company heard from their customers that the process of placing online orders was difficult. As a result, some opted to leave the company’s website and instead order from a competitor. To address this, the company decided to review the online order process. They had multiple meetings to review the steps and developed an amazing user interface that allowed customers to see the products and descriptions, and automated the ability to process the order and send confirmations. The customer now had a great up-front experience. Very quickly orders doubled. Unfortunately, the company then started receiving all new complaints. They failed to identify the relationship of the inventory system to determine if a product was in stock. They also neglected to recognize the hand off to the production team who was not equipped to handle the increased volume of requests. Orders went from a 10 day to 4-week delivery time. Automation in this instance may have made placing orders more efficient, but it also very quickly highlighted and even accentuated process deficiencies.
At the recent IA Exchange, Marissa Peters of USAA explained perfectly that you need to “know what all your work looks like, across the enterprise.” Not just in your space.
Now that we have a clear understanding of how the process works, we can start to look for possible candidates for automation. As we start looking, pretty much anyone who thinks their process is broken will have suggestions.
There are few quick questions that can help you identify if that truly is a good candidate.
1. How does the process work now?
- If it is not documented and they cannot explain it step-by-step, it is not a good candidate.
2. What do you want the automation to do?
- If they cannot describe the value the automation will provide, it is not a good candidate.
As I mentioned at the outset, automation is not about fixing problems. John Cottingim, Automation Director at Mars, noted the best candidate is most likely the process that no one is talking (or complaining) about. That is the process that is consistent, repetitive, and is done without issue. Processes that are highly repetitive and most importantly, stable, are the best candidates.
The takeaway is this:
- Your automation should be well thought out, strategic, and have a defined purpose.
- We should not automate for the sake of automating.
- Do not start automation before you understand the process and its impacts outside your group.
- Identify what you want to accomplish, or how you will measure, the automation before you begin.
For more information, give us a call. We would be happy to talk to you about your Automation goals.