Menu
How to make complex data work for you

07/16/2020

How to make complex data work for you

Like it or not, data is unavoidable. Whether you deal with data every day, every month or once in a blue moon, making sense of the numbers isn’t always straightforward.

If you’re unsure about how to handle a complex data set, fear not! The Traction® Tools team has put together our best tips for making sense of the numbers so you can have effective, productive meetings

Ask yourself, “What is the data telling me?”

Anyone can report numbers without analysis. But to make data useful, you need to tie it back to a key performance indicator (KPI) or goal. When sorting through data, look for patterns, and identify the overarching message—do the numbers show that performance is down? Does it look like your budget is realistic? Do the numbers indicate that you should be shifting focus areas? 

If you have some historical knowledge of the data, apply that to your first sweep of the numbers. If you’re going in blind, consider using a chart or graph to help you visualize changes. 

This habit doesn’t just help you understand complex data; it also helps you present your findings to other people. When it comes to data, try not to assume everyone has your analytical skills. Most people need the story told to them in simple, clear terms. When in doubt, consider your leadership teams’ management styles for useful insight into their preferred communication methods. 

Focus on what you need

Once you know what you’re trying to say, run with it. Muddying up your data set with unrelated metrics won’t just confuse you, it’ll also confuse everyone else. 

Instead, focus on what you already have. Ask yourself questions like:

  • Does this data tell the full story? 
  • What real-world events might have affected this data? 
  • Is this similar to what we saw last month? Last year? Why or why not? 

The idea here is to deepen your knowledge and understanding of the most important metrics in your data set. The rest of the data might come in handy later—but for now, focus on the key takeaways. 

Take randomness and outliers into account

Data is rarely predictable. You can guess at trends or speculate patterns beforehand, but in the end, the numbers have the final say. While this can be useful in creating a data story, be careful not to look at every piece of data too closely. Randomness and outliers are unavoidable; They could be representative of problems with your data collection, anomalies worth looking into or just a reminder that data is never perfect. 

Because of this, it’s best to communicate your data using degrees of confidence. Use words like “usually,” “rarely,” and “expect” rather than “always,” “never” or “will.” 

When in doubt, look at overarching trends, and take surprising metrics with a grain of salt. They could offer useful information…or they could be distracting you from the big picture. Consider using these instances to encourage meeting participation through discussion. Sometimes, you’ll find outliers that are just plain unexplainable—and that’s okay, too.

Use different types of visualizations

One of the best ways to conceptualize and communicate data is through visualizations. Pie charts, tables, line graphs—these are all very useful for conveying the big picture to your team. But each type of graph  works for different types of data. 

There are tons of graphs to choose from, but these four can generally be applied to many data sets: 

  • Use a line graph to communicate trends over time
  • Use a pie chart to compare parts of a whole
  • Use a bar graph to compare different groups over time 
  • Use a scatter plot to show the relationship between different variables 

If you’re struggling to organize your data, consider creating a chart using Excel or Google Sheets to easily identify trends. 

Remember: Data is only useful if you make it useful

Complex data sets are, well…complex. Simplifying and streamlining the process is key to finding useful insights, important trends and identifying next steps. Plus, it will help keep you sane. So, when in doubt:

  1. Look for the overarching message
  2. Focus on the most important metrics 
  3. Take outliers into account
  4. Recognize that randomness is unavoidable
  5. Use visualizations to communicate your findings