FAQ

In the previous tutorial How to create a SCORE metric we showed how you can create an effective SCORE calculation on a metric. However, there is limited value in creating a score on a single metric since we already report the status of every metric by default. The score method is best used as a factor in more advanced calculations. In this tutorial, we’ll show how you can bubble a bad status score to the top of your blocks tree, maximising the visibility of problem areas so they can be acted upon quickly.

First, we need to start with a cascading metric. In this example we have a metric called Expenses. Here’s a table where all the leaf blocks are displayed with their expenses metric in the Single Metric column:

This example is contrived to clearly illustrate the problem and solution. Every block has expenses of $10 and a target of $20, except for the Support block, which has expenses of $40 and the same $20 target. The Bad status shows up easily in this table, but what happens when we remove the filters and look at the blocks tree? Everything looks really good! Expenses are well below targets:

Sure the overall numbers are good, but you don’t want a particular department’s overspending to become a problem later on down the line, so you want to tackle these issues immediately when they come up. Let’s create a MIN_SCORE cascading metric. The calculations are pretty much the same as the first SCORE example, except we’ll use the MIN_SCORE this time. Here’s the Actuals:

And the Target calculation can be exactly the same as before. (As a reminder, this formula just assures us we’re using the same dates as the Actuals series):

Now lets see what this looks like in the blocks tree. Our new metric is called “Find Bad Expenses” and if we display it in the tree:

Now we have a series of Bad blocks from the Support block all the way up to the company vision block. This allows someone to go from the root block and easily drill down to any problems blocks as they appear.

Note: The instruction below is reliant on metrics currently existing within your company to calculate the score on.

Go to Monitor -> Metrics. Click the New button.

With SCORE, we will only have values in the set [1, 0, -1]. The tolerance (yellow) range is the range between the start and the end, but it does not include the values themselves. So if we set the Start/End values both to 0, you would never have any “Warning” points. If you set Start / End to 1 / -1 you would have the warning region filling most of the chart. A way to balance this on the chart is to go halfway: 0.5, -0.5, which results in a chart that looks like this:

So you can decide how you want to fill that range.

Now you may save your metric.

This metric started off badly, but the target has diminished over time resulting in an upward trend into the “Good” range. Let’s create an actual calculation based on this metric. Go to your new metric’s “Actuals” menu, You should see an option to add Manual Actuals or create a calculation:

Click on the Calculate button to start the calculation editor:

Next we want to tweak our target so that it’s always 0, which forces the tolerance we created before to mark -1 values as BAD(red) and +1 values and GOOD(blue) and 0 values as WARNING(yellow). There’s many ways to do this. The first is to not enter any Targets, which gives us this because it’s historic and there are no historic targets:

Since our first value is in 2012, let’s add a 0 there and see what happens:

That works, somewhat. Since we’re basing this on a calculation, if anything changes to the source data we many need to tweak the target. For instance if they remove the 2012 data from the source metric, we’ll have a dangling target to change. The most reliable option is to make the target another calculation that always resolves to 0. There are many ways to do this, but the most reliable is to make the same calculation as the Actuals, but then multiple the result by 0. Here’s what that looks like under the target menu:

The reason this works best is that it aligns itself with all the same values that appear in the metrics series. This means you don’t get dangling future / past values showing up where you don’t want them, and it’s fully dynamic based on the source metric.

Now you can compare your two metrics in the table view and their x-axis lines up perfectly: