While receiving a static analysis of your software development is great, it can sometimes be difficult to determine what to do with that information. The primary challenge is knowing how to change your patterns, habits and other factors to positively impact the numbers you're seeing. Pinpoint helps by also surfacing the relationship among performance signals, through our use of Related Signals.

An Example

One day you notice that the Cycle Time (the average number of days required to close an issue) has been growing for a team over the past few months. That's potentially problematic. What's the root cause, and how do you reverse course?

According to our math, the two signals with the highest impact on cycle times are Innovation Rate and Strategic Issues. This makes sense, because these values are linked to the development of new features, and new features generally take longer to implement than a bug fix – more focus on new features equates to a higher cycle time. If this is the case, all is well.

However, if your Innovation Rate and Strategic Issues are both low, but your Cycle Time is still high, something is off.

So, how do we determine these correlations?

Spearman's Correlation

The Spearman Rank Correlation Coefficient is a standard statistical method for measuring the strength of the associations, or correlation, between two variables. By plotting two variables on a graph, we can determine whether they are positively associated, negatively (or inversely) associated, or not at all associated. We can also determine how closely associated these values are and assign the relationship a value.

A simplified example of the types of correlation is presented below:

correlation-type

The value for a Spearman coefficient can range from -1 to 1. A full -1 or 1 implies the variables match up exactly, while a value of 0 shows there is no correlation. A negative number means that the two values are inverse from one another – that is, as one signal rises, the other falls.

We ran all of our anonymized historical data through the Spearman process to determine the relatedness of our primary signals. Going back to our example, we saw that Cycle Time has a correlation value of 0.59 and 0.42 for Innovation Rate and Strategic Issues, respectively. These numbers suggest a strong correlation.

Interestingly, but not surprisingly, a high Innovation Rate and Strategic Issues also strongly correlate with a higher Cost.

correlation-chart

Like with any statistical analysis, the more data we can run through this process, the better the output will be. Our data science team is committed to regularly performing a Spearman coefficient analysis and making the updated results available to our customers.

Where to Find This Information

Every signal detail screen in the Pinpoint app will show you the most highly related signals, their period value, and their change over prior period:

Screen Shot 2019-01-22 at 4.42.37 PM

Get the data science behind high-performance teams.