Collect & Clean: A Business Case for L&D Data Aggregation and Cleansing

So in this post, we’ll set out the business case for using Watershed for data aggregation and cleansing. It also outlines Watershed's value—even if you don't use the reporting and dashboard functionality.

What's L&D data aggregation and cleansing?

When you have several learning platforms that each contain some of your L&D data, it's unlikely that you'll be able to easily work with and report on that data as it is.

As a first step, you need to combine that data or aggregate it into one place before you can use it. This process involves:

• moving the data or setting up automated data feeds, and
• mapping and matching the data so you can understand it as a complete dataset.

For example, if the same learner is joe.bloggs@example.com on your LXP and jbloggs2 on your LMS, you need to be able to match all of that learner’s data from both platforms.

With any large, unprocessed dataset, you will likely have some data that is not useful or even inaccurate and misleading. So, in addition to aggregating the data, you'll need to check and cleanse it.

You'll also need to:

• review the data for errors (either in the original systems or introduced by an error in the aggregation process),
• look for gaps and missing data, and then
• work to address the issues to ensure good data.

How does Watershed combine and clean your L&D data?

A global construction equipment manufacturer uses Watershed for its learning analytics reporting. They have a variety of dashboards for their learning team’s internal use and other dashboards created for managers reporting on their people.

The company also pushes data that has been aggregated, cleansed, and processed in Watershed onto Power BI to provide reports to their executives—as executive leadership is used to and prefers this format.

So, processing the data in Watershed and then pushing it to Power BI empowers the learning team to provide reporting and analytics to various audiences using the most appropriate tool for each group.

In another example, Visa uses Watershed as analytics middleware to provide aggregated learning data to the various operations groups across the organization. The ability to export to a BI tool means the L&D team can leverage the in-house expertise of data analysts who are used to working with particular BI tools (i.e. where the L&D team can't do all the required analytics themselves).

Another global client uses Watershed as their primary learning analytics platform, but some learning data teams were accustomed to using BI tools instead. One of these teams was initially hesitant to lean on Watershed instead of going directly to the underlying data sources.

But ultimately, they found a lot of value in using Watershed to pre-process data before importing it into their chosen BI tool, which reduced the processing time in generating the required reports.

As a result, the BI tool has much less work because Watershed has already processed and flattened the data while also removing unnecessary fields and enriching it with demographics data from the HRIS system. In other words, Watershed does all the heavy lifting before passing the data on.

How does Watershed support L&D data cleansing and aggregation?

Any Watershed implementation involves aggregating and cleansing your data. This essential step helps ensure pristine data, which means avoiding any confusion from bad data.

And many of our clients have found this pristine data is also helpful for teams in their organization who are a bit more data-technical and are used to using BI Tools for analysis. Several of these teams are using Watershed simply for data aggregation and cleansing as a step before bringing the data into their BI Tool.

Watershed helps in six ways:

1. Aggregation. We set up automated data feeds with all your learning systems. Having all the data in one place means you can create a single data feed for learning data into your BI systems.
2. Cleansing. Setting up those data feeds involves a certain amount of checking and reviewing the data to fix issues. Additionally, Watershed has features that resolve issues with the data, such as the Activity Editor.
3. Data review. Use Watershed to review and sense-check the data. Sometimes issues only become apparent when data is represented in a visualization. For example, you can use the Activity Report to check quiz data for issues.
4. HR data enrichment. Watershed combines learning and HRIS data to match learner data across multiple systems and unique identifiers. It also enriches the data with additional information, such as job role or department. You can include all this data in a feed to BI systems.
5. Data transformation. Learning data comes into Watershed from course systems using a variety of data formats. Watershed converts all that data into xAPI, an industry standard for structuring training data. You can transform this data into a  tabular data structure to match that expected for the BI import.
6. Data extraction. Watershed offers a wide range of API, manual download, and BI tool connector options for getting the data out. These offerings include options to extract data precisely as it is stored in Watershed or get processed data organized to your specifications.

Why should you invest in Watershed if you already have a BI Tool?

When selling Watershed internally, you may come across the suggestion that you don’t need a learning analytics platform because you already have a BI tool. But BI tools and learning analytics platforms aren’t competing products.

Unlike BI tools, LAPs are explicitly built for learning data, making life easier for L&D professionals and line managers who want a simple dashboard to view data about their learning provision and their people.

Watershed can also make life easier for a data team accustomed to regularly using BI tools. Data aggregated, cleansed, and enriched in Watershed is much easier to manage in a BI tool than the source data—helping even the data team generate better reports faster.

Watershed can also make life easier for the data team who are used to using BI Tools on a regular basis. Data aggregated, cleansed and enriched in Watershed is then much easier to work with in a BI Tool than the original source data, helping even the data team to generate better reports faster.

So, the business case for Watershed in this scenario is twofold. First, Watershed puts data and easy-to-use analytics in the hands of people on the ground—the L&D team and line managers. Second, Watershed makes that same data available to the data team for deeper analytics and reporting to senior leaders.

How can you convince stakeholders of the value?

We've seen teams impressed by the time, frustration, and money saved after they've tried extracting data from Watershed instead of combining the source data themselves. So the best approach is to show your stakeholders the benefits.

For example, run a pilot that involves aggregating and cleaning up some source data from multiple systems. Show them how easy it is to extract that data from Watershed for use in their favorite BI tool.

Understand your stakeholders and how they will benefit from learning data aggregation and cleansing.

Next Course: Why L&D Needs Data Automation

Having all your learning data in one place and an analytics engine to crunch the numbers isn’t just an essential first step in reporting and analytics. This data can also inform automation processes—such as awarding a credential when learners meet the criteria, alerting a line manager that an employee’s compliance certification has expired, or automatically delivering a report when metric values fall outside predefined thresholds.

In the next post, we’ll explore the business case for automation and set out the value of using Watershed to trigger events and tasks based on your data and the rules you set up.