Decipher the Patterns: Cycle Time Scatter Plot

The utilization of the Agile Kanban method allows the evaluation of important metrics associated with the process under consideration. Among the main metrics usually employed through the Kanban method, one of the most relevant metrics is the process cycle time, and with it comes the cycle time scatter plot.

In the framework of Agile Kanban method, the metric cycle time permits teams to better understand their overall capability of delivering tasks in their projects, by controlling the start and finish times of the tasks. In this case, cycle time represents the amount of time the work team has spent working on a defined task.

There are many different techniques and particular tools which can be used to analyze the metric cycle time, depending on the type of industry being considered. One of such a tool that finds huge applications is the scatter plot.

What is a Scatter Plot?

In a general view, a scatter plot consists of a type of graph which shows the data set values for usually two variables (say X and Y) by using cartesian coordinates. This type of graph is also known by many other names, as for example, scatter diagram, scattergram, scatter chart or even scatter graph.

Scatter plots are typically used for assessing the relationship between two variables. The data set values serve as the x- and y-coordinates for plotting each observation or point. By using a scatter plot, one aims at interpreting possible existing trends in a data set.

Next, we take a look at what a cycle time scatter plot is about.

What is a Cycle Time Scatter Plot?

In the Agile Kanban method, the cycle time scatter plot basically shows the cycle time of all finished project tasks over a specified period of time.

The horizontal axis of the cycle time scatter plot shown in Figure 2 consists in the variable “time”, which in turn represents the selected time frame by dates. On the other hand, the vertical axis represents the variable “cycle time” of the finished project´s tasks.

Each circular marker shown in Figure 2 represents a finished task associated with a specific Kanban card. The positions of each circular marker are defined by the amount of time required for the Kanban card to move from the “In- progress” column to the “Done” column (in a Kanban board), as well as by the date of task completion.

In essence, the cycle time scatter plot shown in Figure 2 is analogous to that of a classical scatter plot shown in Figure 1.

The utilization of cycle time scatter plots helps analyze and identify possible patterns derived from your process, thus allowing the work team to avoid potential problems in the future.

Now you might be wondering which are the most common types of patterns that can be found through cycle time scatter plots. That’s exactly what we show you next.

Cycle Time Scatter Plot Patterns

When analyzing your data set values through a cycle time scatter plot, it is important to pay attention to possible patterns that could arise based on relationships between the two variables being considered (i.e. “cycle time” and “time frame”). The relationships between both variables reflect important particularities of your process

The first possible relationship that can be detected when using a cycle time scatter plot is the so-called “positive correlation.” A positive correlation in a cycle time scatter plot signals that there is an increasing pattern in your cycle time data set, so that they are getting higher over time. In other words, as one variable increases (e.g. variable “time”), the other variable (e.g. “cycle time”) also increases proportionately.

It is important to distinguish between two possible patterns when a positive correlation does exist: 1) weak positive correlation (Figure 3-a) and 2) strong positive correlation (Figure 3-b). In a weak positive correlation, the cycle time data set shows considerably high variation, whereas in the strong positive correlation, the data set values of the cycle time are closer to each other. High variation should be avoided when possible, as this hinders accurate process predictions.

From an Agile Kanban method point of view, the patterns shown in Figure 3 can be caused, for example, by the fact that the rate of new project tasks entering the process is higher than the rate of completed project tasks. One possible solution to that problem would be, for example, to lower your work in progress (WIP) limits, so that no tasks in your project team would remain static or without appropriate action.

A second conceivable relationship which can be observed through a cycle time scatter plot is the “negative correlation”. In the negative correlation the cycle time scatter plot indicates that there is a decreasing trend in your cycle times, so that they are getting reduced values over time. Therefore, in this type of correlation, as the variable “time” gets higher, the variable “cycle time” proportionately reduces its values.

In the same manner as noted earlier for the positive correlation, here it is also possible to distinguish between either a weak or a strong negative correlation, Figure 4-a and Figure 4-b, respectively. Nonetheless, the high variation pattern noted for the weak negative correlation shows that there is room for process improvements.

Both patterns shown in Figure 4 can represent, for example, a process which is going in the right direction in terms of delivering faster tasks or items over time. In this case, the rate of new project tasks entering the process is lower than the rate of completed project tasks, so the process tends to follow an appropriate path and good workflow over time.

A third possible relationship between the two considered variables which can also be noted through a cycle time scatter plot is the “null correlation”. In this case, the data set values of both considered variables do not seem to match neither an increasing trend nor a decreasing tendency over time, but are rather randomly distributed.

The main goal of using the cycle time scatter plot should be to improve your process in order to get less variability in the cycle time data set, aiming at better process predictability and suitable forecasts.

Take Better Control of Your Productivity

Productivity requires more than just looking at the numbers. It also requires that you take action to improve your processes based on the data you have gathered about your organization’s overall process and productivity. With better visibility, you can improve your team’s productivity. Check out Kanban Zone and keep track of your team’s regular productivity through the cycle time scatter plot metrics available on the board.