Posted by Augustine Fou on May 29th, 2009
As a scientist, I like to run experiments. And I like to make stuff. So my team and I made a few Facebook apps that solved needs that we had (a few samples listed below) and shared them publicly on Facebook to see if they were also useful to other people too.
I beta tested some apps with a few friends by inviting them directly. Then to get it out to a larger number of people, we decided to try Facebook advertising, the much-hyped, holy grail of display advertising on one of the largest and most active social networks.
- visual discovery, share, and queue management interface for Netflix
- visual discovery and sampling interface for music (Amazon backend)
- create and send photo or video e-cards by drag and drop (Flickr and YouTube backend)
- visual display of your friends (closest ones have the most recent status updates)
- social commerce – I’ll buy what he bought; things I have, things I want
But what I found was eye-opening to say the least. Despite the potential of social ads where the social actions of your circle of friends could make the ads more targeted, none of the anticipated positive effects were observed. Despite the promise of mass reach, there was not the corresponding attention or clicks. And despite the use of demographics-based targeting, there was no statistically significant difference between different targets nor the control sample, running during the same time period.
What we saw were click-through rates of 0.01 – 0.05% — and the 0.01% often seemed like rounding because they did not report more than 2 decimal places. As a result of these click rates the effective CPMs turned out to be $0.01 – $0.19 and average CPCs ranged from $0.05 – $0.25. I’ve been running these Facebook ads for more than 12 months; and millions of impresisons later, there is no observable improvements to CTRs and thus CPMs and CPCs. But since I set up the campaigns to only pay when there is a click (CPC basis), I can let these run indefinitely because I am getting so few clicks, it’s not even making a dent on my credit card (which I use to pay for the ads).
In the spirit of openness, as an advertiser who wants to continue using Facebook advertising, perhaps there are a few things they can do to improve the effectiveness of Facebook display ads.
1. reduce the number of ads per page to 1 — displaying multiple ads artificially depresses click-through rates because users can only click on 1 thing at a time, even if they liked more than one of them. Displaying 3 on a page simply increases the denominator while the numerator does not increase — in the click-through rate equation: clicks / impressions.
2. make ads sharable – in the rare instance a user views an ad, it may or may not be relevant to her, but she may know that it is relevant and timely for a friend. By making ads sharable, she can click and send to a friend, who is very likely to find it useful and valuable, especially having been sent by a friend.
3. let users opt-in to ads in specific topic categories – when users are in the market for specific things, they are more likely to subscribe to pertinent news feeds, offers, etc. related to that topic or category. By giving users more power over what they want to see, it will also give advertisers more targeted and engaged prospects to target.
4. expand search-based advertising – when users search they are looking for something and are open to discovering something they didn’t know to ask for. So ads served up in response to a search is usually a lot more effective than ads served up simply when a page is loaded (display advertising). Facebook can serve display ads based on pertinent search queries.
Earth to Facebook… anyone listening?
By
Dr. Augustine Fou. Dr. Fou is Group Chief Digital Officer at Healthcare Consultancy Group a group of agencies within the Omnicom family specializing in pharma and healthcare. He helps clients develop digital marketing programs or improve the efficiency and cost-effectiveness existing campaigns via advanced analytics, social marketing, and digital strategy. You can read more of his writing on digital marketing on this blog and follow him on
twitter @acfou.
This entry is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed.
Posted by Augustine Fou on February 19th, 2009
a full digital audit of a brand’s marketing efforts in the digital space consists of assessments of their online advertising and search (paid and organic) activities. It also includes a competitive scan to reveal competitors they would not normally associate with their brand (the usual suspects). Their marketing activities are compared against social media benchmarks and other digital performance benchmarks like click rates of other advertisers in the same industry vertical or similar site size.
more details on digital audits are here…
http://mktsci.com/digital_audit.html
This entry is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed.
Posted by Augustine Fou on January 28th, 2009
Many clients have asked about social media benchmarks or social marketing benchmarks. They ask things like how many fans should they have on Facebook? They are concerned if they cannot project 1 million fans on their fan page.
But that is because most clients are coming from a reach and frequency background. Some have moved to unique visitors, pageviews, and time on site. But what is more important today is not that people get to the site or the time they spend, but what they do … so social intensity is a benchmark which captures the quantity and frequency of social actions like sharing, discussing, commenting, voting, etc. All of these actions lead to value that accumulates for other future visitors to the site.
This entry is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed.