If you are a typical company, you might be making decisions about your localization process based on perception rather than reality. To illustrate my point, a customer once complained “you never deliver localization projects early. You always wait until the last minute or deliver them late.” We pulled the data on all the customer projects and discovered that the majority of the time, we actually delivered ahead of schedule. But that wasn’t the customer’s perception. Why? Because certain projects that were important to them were delivered right on the deadline. And since they were anxious about them, they felt all projects were like that.
Because of this concern, they were making changes to their localization processes based on what they thought the problem was. Given that they didn’t actually have a problem, those changes may have made things less stable. Decisions on your localization process will consistently be better if you have a data-based, and therefore more accurate picture, of how your process works.
This is the start of a blog series where I will be discussing how to easily achieve a data-driven localization strategy, which allows you to analyze how your processes are working based on solid data. Then you can make informed decisions on how to modify your processes to make them more efficient. However, your first question may be, “What would I gain by using data-driven localization?” Well, with a data-driven localization strategy you can:
- Identify problems and causes and make good decisions on how to fix them.
- Identify potential problem areas and fix them before they become real problems.
- Monitor your localization process to ensure it continues to work well.
This will result in a more efficient use of your localization budget, better quality translations and faster turnaround. To start, all you need is data. You already have access to lots of data about your localization process from all the tools and technology you use in the localization process. Computer Assisted Translation (CAT) tools and your Translation Management Systems (TMS) both provide you with data. Additionally, localization process data can be obtained from your internal systems and from your LSP.
Having data is the start, but moving to data-driven localization means:
- Identifying and capturing data about key process steps
- Determining the numbers that indicate a process is successful
- Defining thresholds for what is good, acceptable and poor performance for each step
Depending on the type of content you work with, the process steps to measure localization effectiveness may differ. A good rule of thumb is to measure any step where content changes hands. And, as mentioned above, your current tools and technologies already collect data for these processes.
SDL can help you get started. All SDL technology products capture data from our Translation Productivity tools all the way up to our enterprise products. Additionally, your Project Manager can work with you to see what data from our internal systems might help you. Lastly, we have Language Process Consultants who can analyze your data and processes and use them to suggest and implement, best practices specific to your data, your processes, and your internal systems.
The next blog will show how to use your data to assess how your processes are working and identify problem, or potential problem, areas.