Benchmark: Mobile vs Desktop on Facebook
With Placements, advertisers can choose where on Facebook to show their ads. For obvious reasons this opens up a few questions in regards to where ads perform best. We’ve conducted a small study, benchmarking delivery of page post ads in the News feed on desktop with News feed on mobile.
We advertised one of our own page posts containing a short text and a link to one of our blog posts:
It was delivered to people with relevant interests that are friends of fans of our Facebook page and the bid type was set to Optimized CPM (optimize for reach). The advertised page post was targeted to 16 countries and split up into 16 campaigns - one for each country. We split it up this way in order to be able to control budgets for different markets, as our reach varies with country.
The ads ran for 7 days with an average frequency of one, meaning that each person in our target audience that saw the ad, on average saw it once.
The long term objective for our Facebook marketing activities is to attract new customers, and one of the ways we go about doing this is via our company blog. As we publish new content we market it on Facebook through page posts that we turn into page post ads.
Attracting new customers this way is a process that takes time, as a big part of the people that sign up for our service don’t do it during their first interaction with us. Therefore, measuring the results from a campaign like this one is not primarily a matter of new signups attributed to clicks on the page post ads.
There are a few interesting metrics we like to look at, such as link clicks (clicks on a link in the post), page likes (likes of our Facebook page), page post likes (likes of the post), and page post comments (comments on the post). These are all summarized as actions. The value of link clicks is quite self-explanatory, page likes increases our reach, and page post likes and comments improve our EdgeRank (and reach for sponsored stories related to the post).
When evaluating how successful an ad is in terms of actions, I have been debating which metric to use in order to best capture the real performance. I am starting to settle on the actions to reach ratio, since this describes how many people engaged with our content out of the amount of people that saw it.
This is in essence a conversion rate, which usually is defined as conversions over clicks. In the case of actions however, a click is likely to be an action (e.g., likes of the post) why basing it on clicks would risk skewing the data. For instance, if you’re running a page like sponsored story and you receive likes for your page directly from the sponsored story, those likes will be reported both as actions and clicks.
Although the dataset was rather small, it gave some interesting results.
- 2.5X greater reach on desktop
- 3X greater CTR on mobile
- 1.5X greater actions to reach ratio on mobile
The actions to reach metric seems low and deserves a comment. Clicks were about 10.5X the amount of link clicks reported, and Google Analytics tells me that the amount of visitors from the page posts were about 10X of the link clicks reported. Since most clicks on a page post ad are actions per definition the actions metric is understated.
That said, what’s most interesting here is the difference between the two placements, why the understated actions to reach ratio is not really an issue in the case of this study.
We saw a substantially higher response rate on mobile than desktop judging from CTR and the amount of actions taken relative to the reach, which in turn drives the cost per action down for ads running on mobile.
The 2.5X reach on desktop is not very surprising, given that people still tend to access Facebook via the desktop to a greater extent than via mobile in most of our target markets. The greater reach resulted in almost 2X more actions in absolute measures, despite the lower actions to reach ratio.
The main takeaways from this small study depend on what kind of business you’re running and what your objectives are.
When to target Mobile
Based on these results the main reasons for allocating your budget towards mobile would be:
- Your primary objective is engagement on Facebook (likes, comments, shares)
And in more general terms:
- You want to reach mobile users specifically (e.g., you deliver a mobile service or want to reach people on the go)
When to target Desktop
Based on the results, the main reason for allocating budget towards desktop is:
- Your primary objective is reach
And in more general terms:
- You want to reach desktop users specifically (e.g., your post contains a link and your landing page is not optimized for mobile)
The general conclusions (reaching mobile or desktop users specifically) might seem obvious. The reason for including them here is that I see quite a few examples of brands that don’t seem to think about this but rather just fire away.
A specific and recent example is a post from a company promoting a mobile app which reached me on my desktop. It lacked a link to follow for installing it, or a description for how to go about. It’s probably fair to say that they would have received better results by targeting mobile devices specifically.
In our specific case, the takeaways are as follows:
Based on this study, ads running on mobile have a higher engagement rate and are a more effective way for improving our EdgeRank. This lowers our advertising costs and at the same time increases our reach for sponsored stories related to the posts.
However, even though the mobile placements came with a higher actions to reach ratio, that is not enough for making us run our campaigns on mobile only. Here’s why:
People are more likely to share our content when visiting us on their desktops (thus creating more incoming traffic), and the vast majority of new signups come from desktop users. Focusing only on mobile placements would therefore not optimize for the value generated by the advertising, even though the actions to reach ratio speaks in the favor of mobile.
Also, reach is important to us and with a 2.5X higher reach on desktop it makes sense to include those placements. Going beyond the reach on mobile and including desktop placements will therefore be part of the setup for future campaigns.
Will this change how we conduct our paid Facebook marketing?
In the case of page post ads with links to our website, no. We will continue to include mobile and desktop placements to combine the engagement wins on mobile with the greater reach on desktop.
In the case of running page post ads without links (e.g., text statuses, photos) and on tight budgets, we will focus on mobile placements to benefit from the higher action to reach ratio.
Connect with Sven on Twitter: @svenhamberg