Facebook and its advertising system is always the source of doubt and controversy. The reasons for that include the system’s size, complexity, and the varying levels of user education. But let’s not forget one of the most important reasons: sometimes the social media giant doesn’t give users enough of an explanation. Because of this, one of the mysteries that marketers struggle with is the PTAT (People Talking About This; storytellers) metric.

What is PTAT?

Facebook doesn’t offer anything more than an ambiguous definition of this indicator. In order to learn what it actually means, we need to use data from other sources. Basic information can be found in  Facebook’s Insights API documentation. That’s where we find the  ‘page_storytellers’ indicator, which is explained as follows:

The number of people sharing stories about your page (‘People Talking About This’ / PTAT). These stories include liking your Page, posting to your Page’s Timeline, liking, commenting on or sharing one of your Page posts, answering a Question you posted, RSVPing to one of your events, mentioning your Page, phototagging your Page or checking in at your Place.

So you can see the PTAT has a much wider definition than just engagement. On top of that, it can also include results that occurred after a page sponsored, for example, events or posts (dark/unpublished page posts). So it wouldn’t be accurate to compare PTAT and engagement directly. But that’s not the only challenge.

Does PTAT change every time I reload a Page?

It’s possible. The number of storytellers refreshes itself in real time (almost; it’s actually every 15 minutes). Each time it refreshes, the metric will refer to the last 7 days.

So how does PTAT work in practice? Let’s look at an example. A metric registered on Wednesday at 12PM will contain all the people who ‘created a story’ no earlier than 168h prior (7x24h). That means it won’t contain people who performed the same action on the same Wednesday, but at 11:25 AM, for example.

But I heard Facebook has a deprecated PTAT metric

Facebook has informed users about the deprecating PTAT, but only in Insights. To replace it, they’ve decided to show us more detailed indicators:

To better see how people interact with a pages content, the PTAT metrics have been split into separate elements: Page Likes, People Engaged (the number of unique people who have clicked on, liked, commented on, or shared your posts), Page tags and mentions, Page checkins and other interactions on a Page.

On most Pages PTAT is still visible for everyone under ‘Likes’ tab:

ptat, storytellers, people talking about this, facebook, facebook metric

People Talking About This

If there are so many problems, why bother showing Storytellers?

In the new version of our Trends reports, next to Engagement, we’ve decided to show the number of Storytellers. We even added them to our app! If you’re still asking why, let us explain.

The PTAT metric demonstrates how people interact with a Page across a much wider context. According to the definition, the only connection between the data from Facebook Insights and public data is the number of Storytellers itself (which we publish in Trends reports). In other words, the data from non-public brands, which is unavailable to the wider public, is somewhat represented by PTAT metric.

Is this data reliable?

We wouldn’t be Sotrender if we didn’t check! But.. how does one do that? As it turns out, this indicator is strongly correlated with Page reach and is the only one that better demonstrates a correlation with paid reach.

ptat, storytellers, people talking about this, facebook, facebook metric

Relationship between weekly PTAT median and sum of impressions for Facebook pages

FROM THE CHART:

  • Weekly PTAT and sum of views on Facebook correlation.
  • PTAT median.
  • Variables have undergone a logarithmic transformation. The size of the points depends on the median of the paid reach. The historical data for 2,481 Brand Pages was analyzed.

Of course, the relationship mentioned above is not straightforward and far from a simple linear correlation (although we also observe a high coefficient regarding Facebook data). However, we can see quite a high rank correlation coefficient.

Start analyzing your Facebook profiles with Sotrender!

Let’s translate this analytical slang into normal English: the higher the PTAT metric, the greater the reach. But this positive effect is not clearly defined. So we can’t say how much reach will increase if PTAT has increased by 10 points. However, the probability that it has increased is greater than the likelihood that it has not.

The PTAT metric is correlated not only with reach, but also with paid reach. And that is something unique. It is the only indicator that in some way (unfortunately still with a margin of error) it gives us the number of paid views.  But most importantly, this indicator is moderately correlated with views, as long as they did not have the effect of clicking on an ad, making a story, etc.

To present all these relationships in chart form isn’t easy. Especially with data this noisy and full of outlier observations (which can not be eliminated for various reasons), as both variables were logarithmic.

However, we can definitely say that the higher weekly median of storytellers per month, the greater the total number of views (including paid).

How does Sotrender calculate PTAT for Trends Reports?

PTAT is a very atypical metric. It’s a challenge for us to represent it in our Trends reports. So how do we go about doing it?

Throughout each and every day, we ask Facebook’s API for the number of Storytellers for each Page in our database (taking place every hour on average). At the end of each month, we compile the hourly results and calculate the median (we treat it as a daily value). Based on those, we calculate the median for each profile (treated as a monthly value).

Facebook Trends United Kingdom, July 2017

Fashion in UK’s Facebook Trends, July 2017

Summary

The PTAT metric is definitely not a favourite of admins and Facebook Page owners, mainly because of it’s complexity. But after long consideration, we’ve decided that Trends reports are significantly more interesting after including them. It helps us to better show how UK’s Facebook is doing – especially since ads are an integral part of it.

Would you like to know more about your competitors paid activities? Order our Paid Communication Analysis.

Analyze your social media efforts with Sotrender!

Author

Aleksander Zawalich

Data Science Team Leader

Data Scientist with social psychology background. Former competitive swimmer addicted to H₂O, podcasts and MOOCs.