The Automated, No-Code Data Stack
Learn how Astera Data Stack can simplify and streamline your enterprise’s data management.
Automation Decision: The Key to Competitive Advantage
In today’s fast-paced and ever-changing business environment, companies need to stay agile and efficient to remain competitive. One way to achieve this is by automating business processes. However, deciding which processes to automate can be a daunting task, especially when there are limited resources and time.
In this blog post, we will explore how a Multi-Criteria Decision Analysis (MCDA) model can be used to make informed decisions about which processes to automate. We will use the Simple Additive Weighting (SAW) method to build our MCDA model and provide an example with five potential automation decisions.
Choosing Factors and Weights
To build our MCDA model, we need to first identify the relevant criteria and assign weights based on their relative importance. The following factors are commonly considered when deciding which processes to automate:
Complexity – How complex is the process? More complex processes will be assigned higher scores. This is because companies want to automate complex processes because they present a greater chance of human error. (Weight: 0.3)
Frequency – How frequently is the process performed? More frequent processes will be assigned higher scores. This is because companies want to automate processes that occur regularly. There is no point in automating something that only happens once every five years, for example. (Weight: 0.2)
Time Consumption – How much time does the process take? Time-consuming processes will be assigned higher scores. Automating tasks that take up a lot of time is a way through which companies can save time and money. (Weight: 0.2)
Current Cost – How much does it currently cost to perform the process? Costly processes will be assigned higher scores. High current costs signal that the process has scope to be made more efficient. (Weight: 0.15)
Implementation Cost – How much does it cost to automate the process? The higher the implementation cost, the lower the assigned score. This is our first relationship since a high implementation cost stands in the way of the automation project. (Weight 0.15)
Human Expertise – How much human expertise is required to perform the process? A high degree of human expertise equates to a lower score. Not everything can be automated, and this factor has been included to reflect this reality. The human expertise score also reflects the handling of sensitive information such as payroll. (Weight: 0.1)
Each factor is assigned a weight based on its relative importance in the decision-making process. In this example, we have assigned the highest weight to the complexity of the process, followed by frequency, time consumption, current cost, implementation cost, and human expertise.
Let’s Put Our Model to Use
Let’s consider a retail company that is considering automating five different processes. For each process, we will evaluate it based on the above criteria and assign a score on a scale of 1 to 5 (with 5 being the best score).
| Criteria | Weight | Sales Order Processing | Inventory Management | Payroll Processing | Customer Service | Social Media Management |
| Complexity | 0.2 | 3 | 4 | 4 | 4 | 3 |
| Frequency | 0.2 | 5 | 4 | 2 | 5 | 3 |
| Time Consumption | 0.2 | 3 | 4 | 3 | 2 | 4 |
| Current Cost | 0.15 | 4 | 3 | 2 | 4 | 3 |
| Implementation Cost | 0.15 | 3 | 5 | 4 | 3 | 2 |
| Human Expertise | 0.1 | 4 | 3 | 2 | 2 | 3 |
| Total Cost | 3.65 | 3.9 | 2.9 | 3.45 | 3.05 | |
Download this MCDA Model Template!
In this case, inventory management is the best process for the company to automate since it has the highest weighted score (3.9). The second-best process for the company to automate is the sales order process (3.65). On the other end of the scale, the payroll process (2.9) is the process that is the least suitable for automation.
Note that this model can be customized for each decision that an individual company makes. For example, you may increase the weight of the implementation cost to 0.2 or 0.3 if resources are incredibly tight and you want to get the best bang for your buck. On the other hand, you may increase the human expertise weightage to reflect the fact that your company works with highly sensitive data.
A Final Word
Automations are becoming more and more common by the day. To make the best decision for your company, it is vital that you follow a specific framework rather than your gut instinct. The MCDA method is a great way to simplify complex decisions.
Automating a given process over another is not a binary decision. You might, for example, want to consider semi-automating a process rather than fully automating it. The MCDA method should also prove handy to you there.
Finally, once you decide which process to automate, it’s necessary to set up the automated data pipelines to achieve end-to-end automation. While structured data has a pre-defined format and is easy to process, unstructured data lacks a specific schema and requires an AI-based document data extraction solution like Astera ReportMiner to ensure your data pipelines have accurate and high-quality data.
Authors:
Farid Khan
