SM

This is an summary of an experiment to measure engagment  on Facebook from large public radio studies.   They looked at the percentage of people that commented, shared, or liked content posted on Facebook for geotargeted stories.

(1)  The first wave of sense-making from early results.  Saying so what to their data.   This resulted in a preliminary hypothesis.

But early on in the project, we noticed something that’s probably familiar to any news organization with a Facebook page — certain stories took off, accumulating hundreds of shares, likes, and comments on Facebook and jolting the Chartbeat meter. Other stories fell flat.

So rather than geotargeting just any news story that a station creates, we are selective and calculated with the types of local stories we post. Content must have compelling headlines. It must be locally relevant and meaningful. And locals should be likely to share it, like it, and comment on it. The editors with whom we’re working closely with at KPLU, KQED, KUT, WBUR, and KPCC are terrific at identifying and creating content that meets these standards.

(2)   Created categories of content types

We looked at every story we geotargeted during the months of July, August, and September 2012, focusing on the ones that the localized NPR Facebook following liked, shared, and commented on at a high rate. From this group of successful stories, we identified similarities which allowed us to create nine distinct content categories. We then dissected each successful story to decide which category it fell into.

To identify a story’s category, we asked a series of questions. Why did people share this story? What reaction did people have when they shared it? What is the story actually delivering to people — an explanation, a video, a hard news story?

We repeated this exercise several times for each piece of content until we were confident placing it into a category.