5 Steps of Learning Analytics [GUIDE]

STEP 1: Plan & Gather Data

The first step is to start gathering all of your data in one place and a common format. Keep in mind there's often a good amount of work and planning that goes into data collection, formatting, and aggregation—but there's also a lot of value from this critical step. And the sooner you start, the sooner you'll have access to key insights.

You can’t have too much data—as long as it’s good data.

Useless data can quickly become distracting and cumbersome. That’s why you must be intentional about gathering the data you need to answer your specific questions.

Once you’ve collected your data, it’s essential to ensure you have low-friction access to it by putting it in an effective format and system. After all, understanding your data will be difficult if you have to contact your IT department whenever you want to ask a new question.

STEP 2: Review & Clean

Once your data is in one place, the next step is to familiarize yourself with what’s there. Think of this step as quality assurance for Step 1 and a foundation for later steps.

• Check your data. Take some time to explore your data for accuracy. Ask yourself if the data seems reliable.
• Identify gaps. Ask yourself if there are holes in your data that will limit you down the road. Can you identify additional data you should be collecting?
• Eliminate junk data. Have you found excessive noise or junk data? Too much junk data can get in your way and even affect the performance of your LRS and reports if left unchecked.
• Test your data. Evaluate learners and learning experiences, configure a few reports, or even ask others to double check for things you may have overlooked.

STEP 3: Operationalize

Once you have a solid data foundation, it’s time to put it to use. Look for some quick wins to gain internal support for your learning analytics project without having to dig too deeply into the data. Stakeholders may be cautious of your new reports at first, especially if they think any existing functionality has been removed.

So, start by reviewing your existing operational reports and then add a sampling of a few new findings or areas for improvement in your first iteration.

Give stakeholders those basics, and you’ll have them eating of out the palm of your hand.

STEP 4: Explore & Analyze

As you operationalized your data, you probably started to learn a lot about what is happening in your organization. Now, start asking why it’s happening.

Begin by looking at unexpected findings from your reporting, including positive and negative deviations from the norm, and ask yourself why these particular outliers occurred.

You designed your learning program with a business goal in mind. So, at this point, you can look at your learning program holistically and measure its impact on the larger business goals. From there, you can identify the actions and behaviors needed to achieve specific goals, determine the learning required to support those actions, and then design the program needed to inspire that learning.

STEP 5: Build & Refine

Learning programs and initiatives are never done—they are modified or replaced. So the next time you begin a new learning program, design it with analytics in mind. You’ll know you’re an analytics pro when you start to weave data into all your design decisions.

Remember, every journey begins with a single step.

There are different ways you can build on what you've learned to get even more insight into your next program. We recommend you follow the Seven Steps of Learning Evaluation to design a mature analytics program. These steps involve creating good evaluation metrics and incorporating them into your programs.

Up Next: Learning Analytics Wrap-Up

Our next post will officially wrap up our Learning Analytics series, which we hope you've enjoyed! Stay tuned for our spring series that kicks off later this month. Don't want to miss out? Subscribe to Watershed Insights to receive L&D industry updates, helpful advice, and more!