This is the subhead for the blog post

Some folks recently said that there was no correlation between amplifying a post and the number of organic impressions. The analysis had 5,000 posts from 1,500 pages, which should be a decent sample size.

But there are a bunch of arguments to make against this position; I’ll outline just a few of them here.

1)     Paid posts drive fan growth in the long run, which means that over time, the audience base grows. All else equal, you’re getting more impressions from a larger base.

How this works outside of Facebook: If your email list grows over time and has decent quality, aren’t you going to get more clicks?

2)     Correlation is not causation.  Brands tend to select sales-oriented posts to promote, which get less engagement. A new product announcement is less engaging than a funny picture– so if anything, we should see a negative correlation.

How this works outside of Facebook: Patients who take medicine for an illness are more likely to die than people who do not take medicine. Well, only ill people will be taking medicine. Only posts that brands want to assist get promoted.

3)     Micro-targeting drives the right audience. Certainly boosting a post is ridiculous, unless you sell a product that everyone uses (soap and other CPGs).

How this works outside of Facebook: Just because a few newbie drivers crash their cars into a wall doesn’t mean that cars are bad.


A proper study means that we randomly divide posts into those we promote vs. those we do not. (Imagine that a drug manufacturer could choose which patients to give his drugs to vs. who to give the placebo.)

We have a long way to measure the impact of Facebook ads.  It’s more than impressions, for sure. Any single metric can be gamed.  So if you see a “study” like this that doesn’t have a counter-balancing metric, look for the loopholes.

In this case, we have selection bias, not looking over the long-run at the impact, and not measuring multiple metrics.

Imagine if Red Bull, who has produced some amazingly awesome content, measured their sales of beverages only within the 24 hours following Felix’s epic jump from space?

Would they conclude it was an unsuccessful campaign?