Editing chart data
Whenever you’re ready to see your own data on the chart, just edit the data source on a sample data series. In this example there’s a sample bar series on the left axis, so we’ll click on it and choose “Set Data Source”.
This allows you to choose to show any type of data from scorecard items, initiative items, or dataset fields. We’ll choose to show the KPI Value for Product Revenue.
Now our chart is showing real data. All that’s changed is the height of the bars.
Adding chart data
You can also add data series to your chart one at a time by clicking the “Add Data” button. Some chart types will ask what kind of data series you want. In this example the options are a new Line, Bar, or Area data series. We’ll choose Bar.
Next, choose where in Impact to get your data. You can choose a scorecard item like a KPI, an initiative, or low-level data from a dataset. We’ll choose Scorecard Item.
From here you can add bars directly to the chart. Every time you click the add button, it adds a bar for that scorecard item.
After two clicks of the Add button, we have a chart that looks like this.
Adding data to the right axis is exactly the same process. Here we’ll add a line for a scorecard item’s score.
This 0-10 score line is now showing on the right axis while the three bars’ $100k+ Revenue are graphed on the left axis.
Reordering with drag and drop
You can reorder anything on your chart by dragging and dropping. In this example, the Training Revenue bar comes before the Product Revenue bar.
After moving Training Revenue to the bottom of the list, its bar is now on the right.
The order also affects the order above and below other items. Here we’ve moved the Book Revenue line to the bottom so that it’s underneath the bars.
You can even drag items to the other axis. Here we’ve moved Book Revenue to the right axis so that it has its own scale.
Setting data series names
Data series names are used in various places like chart labels, tooltips, and legends. Spider Impact chooses a default name for each data series, but you can override this by choosing Set Name in the edit tooltip.
Here we’ve decided to change the Product Revenue scorecard item’s name on the chart to All Revenue.
This new name is now used everywhere for that data series.
By default, chart data series use automatically assigned colors. You can also choose to manually change any automatically assigned color.
You can choose a different way to color the data series, however. In this example we’re going to click on Bar Colors in the Other panel.
And we’ll change from Manual to Score.
This changes the bars to be colored based on each scorecard item’s score for that period.
You can also force all the bars for all data series to a single color by choosing Specific Color. This is the same as manually setting all data series to the same color individually.
X and Y axes
You can configure a chart’s axes through the X and Y Axes box in the Other panel.
The X and Y axis labels default to on, but you can turn one or both off here.
You can also set the Y axis range.
This opens a dialog where you can choose the chart’s maximum and minimum Y axis values. By default, they’re automatically set, but here we’re overriding the maximum value to 50.
As you can see, the chart now shows a maximum value of 50, regardless of what data is being graphed.
Most chart types have labels that you can configure. In this pie chart example, you can see a Pie Slice Labels option in the Other menu.
When we click it, we can see the various label options, including the ability to show percentages and abbreviated chart values.
Repeating left & right axes
When you’re graphing scorecard items, you’ll always be able to set your calendar period range in the Repeat panel. Here we’re showing 6 periods of data for three KPIs. We can edit the calendar period range by clicking on it and choosing “Edit Range”.
This is the standard date range selector where you can choose either a range based on the current calendar, or choose a specific calendar and choose a relative or date range.
You can even show data for multiple ranges. Here we’re showing the data for the current month as well as the data for the month one year earlier.
While scorecard items have a built-in repeating by calendar period, datasets and initiatives do not. In this dataset example, we have a single bar showing the total sales dollars for all time.
Repeating values aren’t required for datasets, but they are very useful. Here we’re going to the Repeat panel and choosing to repeat by the Sales Employee field.
We now have a separate bar showing the all-time sales totals for every employee. Whenever the dataset is updated to include new employees, they’ll automatically show up in this chart.
You can choose to repeat your data series a second time. To do this, click on the “Repeat again by…” button and choose a field. Here we’re going to choose Sale Date.
The chart is now showing the sales totals for the last two months for all employees.
Repeating non-axis charts
Repeating works the same for non-axis charts. In this example we’re repeating three KPIs for four calendar periods. By default, each calendar period is its own pie chart, but you can change this by clicking on the calendar period range and choosing “Edit Range”.
For non-axis charts there’s a “Treat Range As” toggle, allowing you to show one chart for the entire range.
The result is this single pie chart that shows four periods of data.
Grouping similar date ranges
In addition to graphing standard date ranges like all the months in 2021, you can also graph data like months of the year or days of the week. In this example we have two identical data series for Product Revenue, and we’ll choose “Edit Range”.
We’ll change to “Group Similar Date Ranges” and then choose to group yearly data by monthly.
When we’re done, we see a completely different kind of graph. As you can see, it now lists the months across the X axis but no years. Our chart now shows product revenue for the current year compared against product revenue for the previous year.
Every scorecard data series has a “Set Period” menu item. This only shows up when you’ve chosen to “Group Similar Date Ranges” and it allows you to choose which period to use for each data series. That’s how we choose Product Revenue for this year vs. last year.
There’s no “Set Period” menu item for datasets. Instead, you can just choose which date range you want as a filter in the “Set Data Source” menu.
Bar and area stacks
You can create bar or area stacks by choosing them from the “Add Analytics” menu on either the left or right axis.
This adds an empty stack to the axis.
All you need to do is drag and drop data series into the stack. Here we’ve added book and training revenue to the stack while product revenue is its own bar. This allows you to have multiple stacks and non-stacked bars at the same time.
You can configure the stack by clicking on it. By default, 100% height is off, and you can see how the Y axis goes up to $1M.
When we turn on the 100% Height toggle, the Y axis changes to percentages and all repeating stacks become full height.
You can also change between Bar and Area stacks.
You can add a trend line from each axis’ Add Analytics menu.
We now see a trend line object in the left axis panel. There’s also a trend line showing each month’s average of the three series.
You can change the line’s color.
You can also set its data source. Here we’re changing it to Product Revenue rather than All Data.
The chart now looks like this.
Trend lines have an optional fill above and below. Here we’re filling red above the yellow trend line.
Here we’ve turned off the display of the line and are showing a red fill above the trend and a green fill below.
Reference lines and bands
You can add reference lines and bands from the Add Analytics menu for an axis. There are several pre-configured lines and bands to choose from, but in this example we’ll choose a blank Reference Lines and Bands item.
This adds a Reference Lines and Bands item to the axis.
We’ll click “Add Line” and then set the data source. First, we’ll choose to show each scorecard item’s goal.
The chart now looks like this. There’s a goal line on every bar that we’ve made green, and we’ve chosen “Goal” for the line’s label.
Let’s see what a different data source looks like for the line. Here we’ll choose a constant number of 300,000.
After changing the line color and label, it now looks like this.
You can add as many lines as you want, each with its own data source. Here we’ve added a second orange line, this one at 500,000.
There are optional fills above, below, and between lines. Here we’re setting the middle fill to orange. A fill between two lines is also called a band.
You can even turn off the display of the lines to just show the fill.
Finally, we’ll change the line to show the average of all data series.
By default, the scope is the Entire Chart, so you’ll see a single line across the entire chart.
When we change the Scope to “Calendar Period”, however, you’ll see the chart is now only averaging the series inside of each calendar period, with a separate red line for each. Notice how the red line jumps slightly from period to period.
Line data series have a “Show Forecast” toggle.
When forecasting is turned on, Spider Impact will show predictions based on historical values. The area around the predicted line is the confidence interval.
By default, the confidence interval is 95%, meaning that based on the data provided, the line has a 95% chance of being in that shaded region in the future. You can change this to 90%, 99%, or turn it off all-together.
Here’s an example of Spider Impact detecting a seasonal trend.
Here’s a non-seasonal positive trend example.
Here’s an example of no trend.
You can tweak the forecast settings by choosing “Set Forecast Model”.
The default forecast model is Auto, and it’s often all you’ll need. You can also choose to ignore recent days, which is helpful for data sources where recent data is still in flux.
When the model is set to Auto, Spider Impact tries out several algorithms and chooses the best fit. If it doesn’t detect a trend, it uses Simple Exponential Smoothing. If it detects a trend but no seasonality, it uses Holt’s linear trend (also known as Double Exponential Smoothing). If it detects seasonality, it uses the Holt-Winters model (also known as Triple Exponential Smoothing). Both trend and seasonality are additive, as opposed to multiplicative.
If you prefer to choose the algorithms yourself, you can definitely do that. Auto Without Seasonality just means it prevents Spider Impact from detecting seasonality.
When you choose a Custom model, you can choose Ignore or Additive for trend and season. If you choose Additive for season, you can also choose if your seasonality is quarterly, yearly, etc.
Chart data table
To add a data table to a chart, turn on the “Show Data Table” switch in the chart’s Other panel.
On dashboards, the data table is separately configurable with options to adjust the font and margin sizes.