We grabbed 1,000,000 random posts from Facebook and tracked all interactions at a user level. We used the graph API, which is open to anyone. [Side note: if you'd like to learn how to pull this data, ping me.]

Highlights:

• Women tend to “like” more than “comment” on a post. Worldwide, 51% of post likes were from females, though this was 55% in the US.
• Women “Like”  comments.  In the US, 59% of likes on comments were female
• Males post on pages more frequently than females– at 51% vs 48%.  The difference is gender unknown.

For more on Facebook Analytics or to get your sneak peak social dashboard, head on over to BlitzMetrics. Or come check us out at facebook.com/blitzmetrics, where we’ll answer your questions!

This is a recovered post from 2009, but is still relevant to engagement on Facebook

Can you beat my 205.9 Facebook Post Quality Score?

Google has their Quality Score and PageRank, Yahoo! has their Quality Index, and now Facebook has a Post Quality Score. There have been a lot of people asking about what a good score is and what it means, so I thought we’d share Facebook Quality Scores from a few of them (we have about 30 pages).

Here’s what Facebook says about the Post Quality Score:

The Post Quality score measures how engaging your Posts have been to Facebook users over a rolling seven-day window.

And here’s what factors govern it:

When you create compelling content, your fans may choose to interact with the material by commenting, liking, or writing on your Wall. These fans help to spread your content virally throughout Facebook, as their engagement leads to organic stories being published in their friends’ News Feed.

Your Post Quality is determined by the percentage of your fans that engage when you post content to your Page. It is calculated on a rolling seven-day basis. The number of stars depends on how your Post Quality compares to similar Pages (for example, Pages that have a similar number of fans.)

So let’s look at a few of our pages and see how their level of interaction affects the page’s Quality Score in real life.

Keith has a Facebook Post Quality Score of 205.9, while MyEstateManager.com has a score of 55.4– Keith is 3.7 times higher. Why?

Looking at Keith Wilcox, you can see that nearly every one of his posts gets commented and usually has several comments. Plus, multiple people like his posts. On a side note, if you hit “like” on enough items, it will trigger a “suggestion” on homepages of related users. We don’t know the ratio you have to hit. Small fan base of 22, but highly interactive– 8 interactions. Do you know anyone else with a Quality Score higher than this?

Looking at MyEstateManager.com, it has a larger fan base– 164 fans– and even more interactions– 21. But the proportion of interactions is only 12.8%, which is 21/164, while Keith Wilcox is 36% (8/22). Thus, the percentage of interactions is 2.8 times higher for Keith Wilcox. However, his Quality Score is 3.7 times higher.

If Facebook Post Quality Score is measured by percentage of fans interacting, which is similar to number of interactions divided by fans– let’s calculate:

• In the last 7 days, Keith has had 5 unique fans interacting on his page. 5/22 is 22.7%.
• In the last 7 days, MyEstateManager.com has had 8 unique fans interacting. 8/164 is 4.9%.
• The ratio of 22.7% to 4.9% is 4.7 times– still don’t have the 3.7.
• If you count yourself in these interactions, then the number skews to 4.97, even further away.

This is not an exact math and I don’t think folks are going to count all the different variations of fans, commenting, and friending possible to determine the formula for the Facebook Post Quality Score. It’s also not clear how this Quality Score has practical value.

Will it effectively discount the price of PPC, as you see with the search engines (where Quality Score x max bid = AdRank)?

Will it evolve like the SEO industry, where Post Quality Score may drive search rankings? I can’t seem to figure out how Facebook is determining the order of search results– I’m guessing it’s as simple as whether the word is actually in the page’s name. For example, search for “harlem globetrotters”– you’ll have a hard time finding the official group, which is called “The Original Harlem Globetrotters”. Search “baseketball” and you’d expect to see the NBA in there, right? Nope, just a bunch of results that have the world “basketball” in them. Anyway, Facebook’s internal search is a topic for another post. I can see Facebook evolving in search, which will spawn dozen of experts trying to calculate how to game Quality Score and rankings, just like in SEO.

Here’s what Facebook has to say about how to increase your Quality Score:

To increase the number of Interactions and improve your Post Quality you may consider:

• Making sure that your posts, whether they are Status Updates, photos, links, or videos, are relevant to your fans. Posting engaging content is the best way to get people to interact with your Page.
• Posting frequently, but not posting an overwhelming amount of content that users may find spammy or burdensome to consume.
• Increasing your total fan base to generate more interactions: Promote your page with an ad

A couple things are true here. If you spam your list to generate more fans, it will hurt your Post Quality Score. Having a counterbalance to just raw friends (which is what folks do on Twitter and the old MySpace train days) is a good thing– Facebook should create incentives to encourage good behavior in the community. Also, they want you to advertise. I wonder if interactions from a paid ad count the same as a natural one? I wonder if this will spawn an industry of paid interaction folks, just as you see with paid links.

Let’s take a look at the Insights page from MyEstateManager.com– they provide free funeral planning advice, in case you’re wondering.

Of the 21 interactions in the last 7 days (I think they should change to 14 days, since 7 days isn’t much data and causes noise in the calculation for the small guys), 19 of them are likes. Liking is an easy way– and automatable way– to create interactions without having to actually come up with something clever to say. Perhaps people will generate tools to auto-like posts in order to drive up their Facebook Post Quality Score.

Fan growth jumped from 109 to 161 fans in one day, largely because we mass invited from my profile. As far as we know, you don’t get hurt mass inviting, even if you have thousands of fans. We’ve tested where we’ve invited 500 or 1,000 folks at a time– the system lets you do this, but it’s a manual process. Then again, inviting too many fans who aren’t really fans will potentially hurt your Post Quality Score. We should call that FPQS or something– quite a mouthful.

I’d think a better measure of quality would include percentage of active fans multiplied by a spamminess index (governed by how many people unsubscribe and how many folks don’t contribute real comments). What has been your experience with Facebook’s Quality Score? Have you been able to beat Keith Wilcox’s 205.9? What’s the maximum score possible?

The price of moving fast is that you have more mistakes. So you may have noticed some weird stats in your Facebook analytics. For example, today is November 26, but we have a post that is on December 1st:

And then there’s one of our client pages that reports almost 300,000 new fans in one day:

Though we all wish we did get 300,000 new fans in one day, it turns out to be a bug.

Sometimes there are holes in the data. In this case, we show a zero for one metric for one day.  The impact of this is that it looks like we lost all our fans in one day and then gained them back:

Moral of the story– watch your stats carefully. Don’t trust data blindly. If you see one stat that is off, check the related stats to corroborate. For example, if we really did get 300,000 new fans in one day, you’d see a massive increase in impressions, likes, storytellers, and other metrics.

This is a common issue with pulling data Facebook API data, as well as what you see in the web-based insights. But a strong data analyst won’t be fooled.Two of the above images come from our tool, which pulls from the various Facebook APIs, while one is a screenshot from the web insights.

What have you seen in the data?

Here are the top tips from Web 2.0 NYC:

• Be real.
• More interactions are key- responding to comments generates more of the same.
• Size and engagement are key metrics, although engagement is more important than raw size.
• Don’t buy fans– it’s poison.
• Social affects search because of a Google overweighting on social results, especially real-time.
• Don’t autopost to Facebook from twitter– Facebook is shutting it down.
• Customize your message to each platform.
• Use the data– you now have fan-level data to inform your marketing and be more personalized.
• Be real-time– don’t sit for a week.