Achieved a 90% reduction in error-prone manual work with custom build automation-tool.

I’m writing this blog post to share how I was able to help one of my clients gain man-hours savings and improve quality in a one-time error-prone manual activity.

We wanted to take data out of a legacy system where we didn’t have access to any backend services. We weren’t able to login to the database or interface to any of its API to extract information.

There were no other means to extract this information other than using the user-interface. Being a legacy system, you can imagine how bulky and slow this system was to interact with. We were left with two options; either we would manually extract this information or we would build an automation tool.

To give you an idea of the scope of work – we had 8400 sites to visit and to reach these sites we had to go through 5 clicks each. Once we reached the site, we had to extract the information and place it in on an excel sheet, produce the output report and save it on a specific folder.

We estimated this work to take approximately 2 min per site. This sums up to a total of 280 hours of error-prone manual work. Not very appealing for any user to go through this process.

So how do you solve this issue? Should you use automation for this?

Investing in an automation tool is one option to use, but the initial cost of seemed too extensive and time-consuming. To get this to work, you need to install this in the customer’s IT environment, buy the licenses and learn how to use the tool and then start configuring the tool.

This wasn’t going to give us a lot of time-saving. So was it really worth installing an automation tool for a one-time operation?

No, it wasn’t and I wouldn’t have been able to convince my client to make the initial investment either. So instead I decided to simpler and readily available tools to solve this problem: Microsoft Excel, VBA, and Javascript.

I developed an automation tool that was able to extract all 8400 target URLs, visit those sites, extract the required information, format the output in the desired format and store it in that specific folder – one folder for each document.

How fast did I do this?

It took 4 hours to develop, test and deploy the script. It took another 2 hours to run it and get the job done. 6 hours in total versus 280 hours of manual work. That amounts to more than 90% in man-hours saving.

Reach out if you want to hear more about where you can deploy practical automation tools to reduce man-hours.