Sorry, we could not find the combination you entered »
Please enter your email and we will send you an email where you can pick a new password.
Reset password:


Plus Report - By Thomas Baekdal - May 2012

Initial Test of Facebook's Promoted Posts

Shared by Plus subscriber
Avinash Kaushik
This is Baekdal Plus content. It is shared with you for free by a member. Please reshare it.

What actually happens when you promote a post using Facebook's new 'Promoted Posts'? Is it worth using? Is it worth paying for?

Last week, Facebook came out with Promoted Posts, which is initially only available to a relatively small group. And, one of my pages happens to be included (I have no idea why).

I wrote about my initial reaction to this new business model in "Facebook EdgeRank, Promoted Posts, And The Connection". The short version is that I think the whole concept is a scam. It's Facebook first taking our own fans away from us by filtering them out with EdgeRank, and then forcing us to pay Facebook to get them back. To me, that's like asking for a ransom. If you have not read my previous article about it yet, please do so before reading this one.

Some have suggested that it's not bad at all. They say that what EdgeRank is really doing, is just to filter out all the people who have no value and all the posts that people don't care about. That is an interesting theory, but I can debunk it with just one sentence:

If EdgeRank was truly filtering out all the people who are not valuable, why do Facebook think it is a good business model to force you to buy them back?

Clearly, Facebook knows that EdgeRank is filtering out a valuable chunk of our fans, otherwise this whole business model of promoted posts would be a failure. It's just dishonest.

The test

To see just how much an impact promoted posts have, I decided to do a small initial test, and it has been running for the past 4 days. The numbers are in, so let's get cracking.

This is just a small test. I posted a new article to my design site, 42Concepts, about something that I thought would have a good chance of going viral. The article was about a bathrobe that looks like it is censoring out parts of your body. It's a brilliant design.

I then posted this to Facebook, and promoted it using promoted posts.

When you do this, Facebook will promote the post to all the people it has filtered out with Facebook EdgeRank. It is essentially a way for you to disable the filter, so that all your fans can now see your post. It's not advertising, because you are only reaching people who have already decided to become fans, and thus 'expect' to see what you post.

It also increases the potential viral effect. When one of your fans engages with your posts (Facebook calls this the 'people talking about' your post), that person's friends are more likely to see it too. But only for the fans who engage with you.

The first thing that caught my attention was the 'estimated reach' (as seen below). I only have 2,254 likes on this page, so how can I reach 3,100 people? That's more than 100% of my total audience, of which a big group of my fans will see the post naturally.

One of my readers said, "maybe they are showing it to more people than just fans? Maybe it will also turn into a sponsored story?" But no, that's not the case.

What I think is happening here is that Facebook is saying that maybe, someday, if the temperature is just right, some of your fans might potentially like and share your posts, and that will possibly create a plausible viral effect that might lead to 3,100 people seeing this post.

I decided to pay $5, for which Facebook claimed it would give me an estimated reach of 2,500 people.

This was the result:

The first thing you will notice is that I didn't reach 2,500 people. I only reached 30% of Facebook's estimate.

It generated 19 link clicks, which gives us a click-through rate of 2.5%, compared with 46 total clicks. Meaning that the promotion accounted for 41% of the clicks generated by this post.

I then compared this to all the traffic coming to only this article from Facebook. The number is 343 unique visitors, which means the promoted post only generated 5.5% of my total Facebook traffic. The rest must have come from the organic views (fans seeing it naturally), and people sharing the post without connecting with my page.

It ties into something I have been saying for a long time. The real value of social is not what you post yourself. It is to create something that people want to share by themselves. 87% of my Facebook traffic comes from people who never saw this post.

We can also compare click-through rates to the reported rate of organic vs paid-for fans. The reason for this is to learn if paid-for fans have a higher or lower click-through rate than 'normal fans'. That would be very interesting to know.

I had expected that paid-for traffic would have a lower impact, but the result is almost the same click-through rates for both groups, with slightly higher numbers for paid-for traffic. Looking at fans alone, the click-through rate is 3.6% for organic fans, and 3.7% for paid-for fans. Looking at the total reach, the click-through rate is 2.3% for organic fans and 2.6% for paid-for people.

Or, in other words, the same people that Facebook filters out with EdgeRank is actually slightly more likely to click on your links. At least in this test.

It also generated 5 post likes, out of a total of 13. That's good, because then maybe those five people won't be filtered out with EdgeRank the next time I post something ...but just maybe. Again, if I compare the like rate, it is exactly the same for both groups, this time though, slightly biased towards viral effect from your organic fans (by about 0.002%).

Think about this for a moment. What this tells us is that the people that Facebook filters out with EdgeRank are just as valuable to us as the people they include.

It also created one page like, meaning that because of the viral effect of a friend seeing it, a fan has engaged with a post. So, I got one new fan.

Another way to look at this is to compare the traffic, like this:

We can also look at only our fans, and see is this:

For those of you who are good at math, you will realize that something doesn't add up.

This post is reported to have a total reach of 1,663 people, of which 1,191 is organic, 350 is viral, and 746 is paid. But if you add those numbers together, you get a total reach of 2,287 people - 37% more than the total.

Note: What's actually happening here is that Facebook has three separate data points, causing one person to show up in more than one category. For example, on day one he sees it as an organic post, the next day he sees the same post as a promoted post, later that day he sees it again because a friend liked it. It's a common problem in analytics. Each data point is correct in itself, it's just useless as a whole.

Another example: The total reported paid-for reach amounts to 746 people, 518 people (23% of 2.252) is by fans, reaching a reported paid-for reach of 753 people. This leaves us with 235 ...or 228 people ...caused by the viral effect.

That just makes no sense. How can we have two numbers for the same thing? And they are not even the same numbers.

Okay, let's look at another one. We know that 1,663 unique people saw this post. We also know that 56% of our fans saw it (1,261 people), this leaves us with 402 people as a result of the viral effect. But the reported viral effect is 350 people.

What the heck is going on here Facebook? It's just a mess.

What can learn from this?

As I started out saying, this was just a small initial test of Facebook's promoted posts. Considering that I only tested one post, for one brand, linked to one site, I cannot really make any conclusions. But that was not the point either.

The point is to illustrate what you need to look for when you do the test yourself, for your product, your brand, and your fans.

There are, however, a number of intriguing indicators that I'm fascinated about:

Over a four day period, only 56% of my fans saw this post - even when I promoted it. To me, this indicates that the rest (44%) did not visit Facebook during that time span. I find it to be highly unlikely that only half my fans see my posts, simply because they don't sign in every day, or even every week.

I'm also fascinated to learn, once again, that the value of a fan is not solely the result of engagement. The organic fans (the ones Facebook has determined to be valuable using EdgeRank) are just as likely to click on a link as those Facebook filters out.

For a while I have been saying that a 'listener' is just as important as any other fan. In fact, since there are often more listeners than engaging fans, the listener actually creates more sales overall. I have tested this so many times, with different things, and every time I see the same pattern - including in this small test.

Do not give up on your listeners! And Facebook, stop filtering them out with EdgeRank just because they don't engage.

Note: Coming next week, I have a much more detailed 'Plus' article about why this is. So stay tuned!

It also fascinates me, once again, to see how little effect my own posts have to the overall traffic from Facebook as a whole. 87% of Facebook traffic came from somewhere else other than my Facebook page. I have seen this so many times, with so many brands. The real effect is not what you do on your profile, but how interesting and sharable you make your products.

Is it worth it?

Well ...yes and no. Personally, I'm pissed off that I have to pay to reach my own fans just because Facebook has filtered them out with EdgeRank. Especially since I, in this test, can see that they are just as valuable.

They have taken my own fans hostage based on some misguided belief that unless they constantly engage with me, they are not valuable to me. They are valuable, so leave my fans alone!

But if I forget that for a moment and just look at the financials, then depends.

$5 is arguably not that much money, but as a publisher, it was like throwing money out the window. The cost of making the article was $6, hosting it was another $1, and promoting it was $5. Total cost: $12.

For that I reached 753 people, and gained 19 clicks, which provided me with a total advertising income of only 5 cents (with a CPA of $0.26). That's not a good ROI...

On the other hand, let's say you have a brand and an average conversion rate of 6%. Then this little stunt would have provided you with a single sale. Arguably that's not much, but it would more than cover the cost of promoting the post in the first place.

And, at least in theory, it should scale quite well. So if you are H&M with 11 million fans, paying to reach your valuable listeners will probably produce a sizable volume of sales for a relatively low amount of money. Remember that Facebook advertising has an average click-through rate of just 0.05%, while this had a click-through rate of 2.5% to 3.5%. You are not reaching strangers, you are reaching an audience who are already interested in you.

I hate the idea of promoted posts because Facebook is stealing the connection I have with my own fans. But from a financial point of view, it makes sense to 'give in' and pay the ransom.

I think it is unethical for Facebook to do this. They are basically lying to us. But after the IPO, this might just be how it's going to be.


The Baekdal Plus Newsletter is the best way to be notified about the latest media reports, but it also comes with extra insights.

Get the newsletter

Thomas Baekdal

Founder, media analyst, author, and publisher. Follow on Twitter

"Thomas Baekdal is one of Scandinavia's most sought-after experts in the digitization of media companies. He has made ​​himself known for his analysis of how digitization has changed the way we consume media."
Swedish business magazine, Resumé


—   analytics   —


Creating a propensity model for publishers


How my focus on analytics has changed as an independent publisher


How can publishers measure trust and other editorial metrics?


A guide to analytics for independent journalists


Why producing less news leads to a boost in subscriptions


GDPR: How publishers can track things without tracking people