If you're like many Facebook Page administrators, you check Facebook Insights for your Page on a regular basis.
A lot of insightful data is contained within the Facebook Insights user interface (UI) and even richer data if you make use of the API or a robust social analytics tool.
There are also, however, a variety of metrics in the Facebook Insights UI that can be misleading unless you know what the data represents. These metrics relate to Reach and Talking About This and are found within the corresponding tabs in Insights.
The reason why these metrics can be misleading is because they are unique metrics, which means daily totals can't be summed to get the total for a week, month, or any other time period.
For example, you could have exactly 1,000 people talking about your Page each day in February. That doesn't mean you had 28,000 people talking about your page for the month. Those 28,000 stories in February may have come from just 5,000 unique people.
The same is true for the various Reach metrics. Reach is defined by Facebook as the unique number of people who have seen any content associated with your Page.
The challenge is that it takes a significant amount of computing power to be able to calculate these types of metrics on the fly for any date range that a Facebook Insights user defines.
So how does Facebook deal with these types of metrics? It restricts the time increments that you can view to either daily, 7-day or 28-day periods.
In the Facebook Insights UI, your view is restricted to rolling 7-day periods. This is not evident when looking at a chart such as Reach where, based on the x-axis, it appears to be showing daily values.
Hover over any point in the chart, however, and you will see it is actually showing the total for the previous 7-day period. Exporting the data or using the API will allow you to also get at the daily and 28-day values
While fairly useful when used in conjunction with a frequency distribution, or even looked at in relation to impressions, the non-unique equivalent metric, it isn't very useful if you're trying to create monthly, quarterly, or any other report that isn't based on daily, 7-day, or 28-day periods.
What are you supposed to do if these metrics are important to you and your reporting period does not align with those that Facebook provides? There are a few options, none of which are ideal. For this purpose, I will assume the desired time period is monthly.
Use the 28-day value from the last day of the month. For March, you would have the correct value for March 4-31, but would be missing March 1-3.
It's pretty clear how this doesn't provide a complete picture and can be quite inaccurate, especially if the first three days of the month were strong days. It does, however, provide a uniform method with accurate data, albeit not monthly data.
Use the 28-day value from the last day of the month and then try to create a sophisticated calculation that attempts to estimate the two to three days that aren't included in the 28-day value.
You should be able to get to a point where you have values that are closer to the true full month values than with the previous option. The obvious downside is that you're introducing estimated data points that are known to be incorrect.
Only use non-unique metrics when reporting at the Page level. In this case, Impressions would replace Reach and Stories would replace People Talking About This.
If you also report at the Post level, you could still report the unique metrics since Facebook only provides lifetime values for Post level metrics. This method would ensure that all your data is accurate and representative of the full month, but it would mean that there are no Page level metrics related to Reach or People Talking About This.
Sum all the daily values for the month with a giant asterisk indicating that the values for Reach and People Talking About This related metrics are wildly inaccurate (please don't use this option!).
There really isn't an ideal solution here. Instead, you have to pick an option that works best for your needs or come up with an alternative solution that works.
As long as you know that the look of some charts in Facebook Insights can be deceiving, you can at least explain the data to other stakeholders and, if necessary, modify how the data gets reported to meet your organizations needs.
Meet Your Favorite Search Engine Watch Contributors
Many of SEW's leading expert contributors will be at ClickZ Live, the new online and digital marketing event kicking off in New York (March 31-April 3). Hear from the likes of: Thom Craver, Josh Braaten, Lisa Barone, Simon Heseltine, Josh McCoy, Lisa Raehsler, Greg Jarboe, Dan Cristo, Joseph Kerschbaum, John Gagnon, Eric Enge and more!