As described in a previous L2 article, dark posting is an advanced Facebook marketing tactic popular among publishers and b2b marketers. Adoption at traditional brands has been slower, with only a handful of brands using the capability. To help brands have a clear understanding of results that can be derived from deploying dark posts, L2, in partnership with Naytev, conducted a review of normative performance data for over 100 publishers deploying dark posts and engaging in the dark post optimization.

Dark post optimization, often referred to as “dark testing”, is essentially A/B testing on Facebook, conducted by 1) building multiple variations of a single post by adjusting the message, thumbnail, image, etc., 2) serving these variations to different, similar audiences, and 3) measuring performance and designating a “winner.” Oftentimes, the winner is published to the advertiser’s timeline as a promoted post.

Test, Test, Test

Standard practices for post optimization include:

Test Volume: Typically, 6-12 variations of a single post are tested.

Timely: These publishers use Naytev to set up these tests in less than ten minutes, whereas using Facebook’s Power Editor is more time consuming (approx. 90 minutes per test). A winner is typically deemed within six hours of the test running.

Dynamic: Naytev leverages machine learning to dynamically allocate budget towards the high performing messages, thus maximizing overall test performance.

Targeting Old & New Users: These posts are typically targeted to a 5% lookalike audience of the publisher’s Facebook fans— thus this can be a vehicle for customer acquisition, serving ads to non-fans.


Dark Effect

For the period of January through May 2016, across 48K posts, 100 publishers saw improvement in post performance across engagement and click-through rates, as well as cost savings. Most notably, click-through rates improved a +104%, averaged across tests. Engagement rate improved +63% and average cost-per-click savings was 39%.

Within the tests that contained video, view performance benefited as well– +30% improvement in the view rate to the 10 second mark, a 24% cost savings for viewers to reach the 10 second mark, and a +16% improvement in the average percentage of the total video watched.

Dark Effect_table1

Note: Improvement calculated on highest performing message vs. lowest performing message (that had at least 250 impressions)

Click-Through Impact

With an aggregate click-through rate of 3.7% and significant variance by publisher category— from a high of 5.8% for animal-based content to a “low “ of 2.6 for Tech Newsclick-through emerged as a key area of positive impact for publishers undertaking a post optimization strategy.


To understand if 3.7% is in fact a “good” click-through rate, we looked to other forms of media– this rate is nearly double that of Google AdWords search benchmarks and significantly higher than Instagram and display, suggesting that Facebook can and should compete for its fair share of the marketing budget. This also helps explains why 20-60% of publisher’s traffic is coming from Facebook, with Google’s contribution to traffic share plateauing.


In light of Facebook’s recent announcement of an algorithmic change to favor UGC over publishers and brands (and quite reminiscent of past Google algorithm updates, #Publishergeddon), brands must become ever smarter in paying to engage audiences with optimized imagery and messaging.
In the next installment of this series we share best practices for optimizing dark posts.  

Daily Insights in Your Inbox