Standard deviation is a concept that is often beyond the regular toolkit of the average digital marketer. However, it isn’t difficult to understand. Widely used in analysis, it measures the range of variation that a data set has from its average value. In other words, it simply tells you how spread out the data is from the average.
To understand standard deviation, let’s first get a definition of what the mean is. The mean is simply the average of the data set. For example, let’s say your 5-channel campaign has respective CPCs of 1, 2, 3, 4, and 5 for each channel. Adding these up and dividing by the number gives you the overall mean CPC of the campaign. Written out, this would be (1+2+3+4+5)/5 = 3.
Now, let’s say your next campaign has CPCs of 2, 4, 6, 8, and 10. Following the same calculation as above, the mean of this campaign is (2+4+6+8+10)/5 = 6. So very simply, the mean CPC of this campaign is higher than the first campaign’s. That is to say, on average, the CPCs in the second campaign are higher than those in the first campaign.
However, the mean only tells us a very specific part of the data story. All it has given us is the central point of the data. It doesn’t tell us anything beyond that, and nothing about how spread out the data is. Worry not though, because this is where the standard deviation helps to augment your understanding of the data.
So how exactly is this measurement calculated? The formula used is admittedly a bit complicated, and Excel will have a handy formula for you, but for the curious ones, let’s look at it step by step.
- First, we have to calculate the mean. Add up the data point values and then divide by the number of points we have like in the examples above.
- Then we need to find the difference between the mean value, and each respective point of our data set. To do this, just subtract the mean from the value of each data point. Don’t worry if some of the results are negative.
- Next for each data point difference that we just calculated, we need to square and add them all up together. This is why we don’t need to be concerned about having negative results. Squaring a negative number always gives a positive number.
- The next step is dividing by the total number of data points we have. Effectively, we are getting the mean of the squares.
- Finally, we then take the square root of the result and the final outcome is the standard deviation.
Interpretation of this figure is quite straightforward. The higher the number, the wider the range of values in the data set. A low standard deviation just means that the data set is very consistent, as the data points are likely to be grouped around the mean.
How would standard deviation be relevant in a marketing case, then? Well, it would help to indicate how consistent your data is. As another example, it could help to inform you whether or not your CPCs (cost per click) are within the expected range, or if you’re experiencing sudden wider ranges of impressions or traffic numbers. If anything, it’s a single number that gives you an indication of the spread of the data, and this may prove useful as you’re trying to find explanations or justifications for the data.
Have you tried using standard deviation yet? How easy was it to use? Did it prove useful, or did your audience really understand it? Let us know!