Make Your Metrics Meaningful: 3 Must-Have Rules for Any Report

If you’ve ever presented the results of your SEO or SEM efforts to a client of yours then this scenario will be familiar:

You spend hours (or days) compiling the data. You conduct your analysis, create your graphs and package it up. It’s been a great month and you make sure to convey that in your email, though the charts speak for themselves. You send it off.

Questions like these are the ones that come back:

"What does this chart mean?"
"Why did our numbers go down here?"
"What caused this increase?"

Now, if you’re fortunate enough to have clients who actually read your reports, consider yourself lucky. An informed client is more likely to be a champion for you and your future efforts than one who never opens your attachments.

However, that also means you have a responsibility to anticipate your client’s questions and take the necessary steps to make your metrics meaningful.

How to do this? While there are many paths to the creation of meaning, below are 3 Must-Have Rules for any SEO or SEM report:

1 – Show Everything in Context

In the words of Edward Tufte, context is king. Without context, your data runs the risk of being meaningless or, worse, misunderstood.

Consider the graph below. We see a nice upward trend in Organic Traffic for most of the year, but there’s a noticeable dip in February that begs the question: why did the traffic decrease?

As the creator of this graph, it is your responsibility to answer that question before your client even asks it. Consider the benefit of adding this simple callout box to bring some context to the data:

Now your client will have the answer before the question is formed. You have provided the necessary context to make this chart meaningful.

2 – When In Doubt, Annotate

A close cousin to contextualization is annotation. When there is the potential that the data could be unclear to the reader, even a minor annotation can make the difference between misinterpretation and clear understanding.

Take, for example, the line graph below. A client taking a quick look might think there is cause for alarm because the average keyword ranking decreased precipitously in February (noted here with a red arrow).

However, the purpose of this graph is really to compare the current month’s average keyword rank to the average keyword rank at the beginning of the project. In that case, the fact that February’s average rank is higher than at the start is, in fact, a positive result – an observation that is called out in the yellow annotation box below.

A caveat: it is important to note when too much annotation is required to clarify a chart or graph. If you find yourself spending a lot of time explaining a graph, it’s possible that a different type of visual display would convey the same information in a more meaningful way.

3 – Put Labels in Close Proximity to the Data

Edward Tufte (can you tell he’s a hero of mine?) is also a proponent of using "sidenotes" instead of footnotes to clarify information in any document, especially one that contains charts and graphs.

The need for these "sidenotes" is clear in the example below.

In this pie chart the client must look back and forth from the labels to the chart to glean meaning, such as discerning whether it was the DMV group or the SAN group that received 13% of the clicks in March.

Your client shouldn’t have to work that hard to understand your charts.

Instead, bring the labels in close proximity to the data. By simply adding the labels near the percentages around the chart you suddenly have data that’s much easier to compare and therefore much more meaningful.

Now don’t get me wrong; there are certainly several other factors that contribute to meaningful metrics: pulling relevant data, showing trends over time, using accurate tracking methods – the list goes on and one.

But without incorporating these three Must-Have Rules of contextualization, annotation and label proximity, you run the risk of having the meaning and value of your reports (and your hard work) washed away in the tide of questions about what those numbers really mean.