5 Pitfalls of Automation

Bodil Ekström is one of the founders of Releye and has been working with the digitization and automation of customer processes for over 20 years. Here, she shares the pitfalls you should avoid in an automation project.

1. Too ambitious

Once you’ve decided to start automating business processes, there’s a risk that you’ll want to begin with the most complex process, or that you’ll want to automate several processes at once. This can make it harder to establish a structure and get an overview of the work at the outset. When the process becomes too complex, it’s also easy to get bogged down in deciding which data sources to use or other specific details.

Instead, start where you are. Automation works very well for routine, recurring, and resource-intensive tasks, whether internally or for the customer. Consider these aspects and set priorities, and calculate, for example, the lead time you save internally and/or for the customer. Then select a workflow, which will make it easier to get started and measure the results.

2. Hard-coded processes

An automated process isn’t something that will always look the same. It’s easy for an automation to be implemented in a hard-coded manner and for people to then “forget” that it’s running in the background. The need for changes may stem from new technology, legal requirements, changes in customer demands, or business rules. Therefore, changes must be easy for the business itself to make. It is important to understand that a process must be adjustable and adaptable based on new information. Working on continuous improvements within an automation team is highly recommended.

3. All systems must be integrated at once

This also has to do with the level of ambition. The risk is that the automation initiative will become nothing more than an integration project, where the goal is to integrate all data at once. This leads to long lead times and makes it difficult to demonstrate results and roll out the solution iteratively. Select the process(es) where you can determine early on that you have access to the right data and where integration is possible. Also, perform test runs with data and verify that the underlying data is up to date.

4. Automation solutions are viewed as “a ready-made product”

There is a common misconception that automation software is like any other IT product. You buy it and then click a play button to get started. It’s better to view it instead as a platform—a powerful tool—where you can automate workflows and use various actions and rules to determine what needs to be done. The platform is an engine for the business—not a finished car. It offers the functionality to automate, but the business needs to decide when, what, and how.

5. Create the conditions for success

Business process automation is not the same thing as AI. An automation platform is not a “smart” product from the start. It does not learn or adapt on its own. It does exactly what you have configured it to do, using the information you have selected, when you want it to do it, and when you want it to stop. If you set up a process to contact the customer at a specific time and it turns out to be incorrect, the process will continue until you change it. That is why it is important to work together as a team that has some form of approval process to verify that what you intend the platform to perform is also correct and up to date. However, AI can be a component of the platform when processes are defined and automated and there is a wealth of data available.