That Facebook is important seems undeniable. A growing body of thought is emerging that Facebook may be in the process of becoming the de facto center of the internet universe – effectively replacing Google. That’s not inconceivable, but there are some fairly good reasons why it may not happen. Fortunately, as a measurement company, we’re much more concerned with optimizing the present than predicting the future.
As part of that “optimizing the present”, we’ve been involved as the measurement arm for several recent large Facebook campaigns by clients. None of them have been very successful.
I think it’s still unclear whether Facebook will turn out to be a really good advertising platform. The rich availability of targeting data and the impressive engagement numbers are attractive. But not every medium is conducive to advertising; Facebook is still working to find an effective format and agencies are still working to find effective targeting and creative strategies.
Understanding how immature Facebook is as an advertising platform is essential to developing a good launch strategy if you are planning a buy.
Since most Facebook efforts have a viral component, there is a tendency to heavily front-weight the ad buy and rely on viral spread to carry a program forward. But front-loading a Facebook effort necessarily eliminates or sharply reduces your ability to tune the campaign and learn from your mistakes.
The sheer size of Facebook can also work against you. Facebook can deliver so many impressions so rapidly that your entire effort can be swallowed up before you’ve had any chance to react or optimize. This also makes wallpapering strategies virtually impossible.
So Facebook is inherently a targeting platform. That’s a good thing and it plays to the strength of information that’s available. It also means that you have an opportunity to fine tune the combination of creative and demographic/interest targeting.
To optimize that combination is going to take time. It will probably take significantly more time than it will take you to optimize and equivalent Display campaign: you have more data for optimization and your agency will have less experience in using that data effectively.
Which brings me to my mine main point. If you’re planning a Facebook buy, you should think seriously about back-loading it not front-loading it. If you can (meaning if your program timeline permits it), soft-launch your effort and do extensive testing before the bulk of your buy. Give yourself and your agency more optimization time than you would for Display. If the program starts to go viral you can always ADD impressions. That’s easy. But there’s no way to get spent impressions back into your budget.
Of course, this also puts a premium on measurement. That’s as it should be. Immature mediums demand and reward careful measurement. You can’t rely on your agency’s years of expertise to get good Facebook targeting. Nobody has that expertise. By building in the time to launch your Facebook campaign slowly and measure your targeting efforts carefully, you not only give yourself a better chance of success, you give yourself a much better chance of actually learning something.
In the early days of PPC advertising, we frequently saw programs that performed very poorly at launch but that improved rapidly with optimization. If your Facebook performance is significantly better at the end of a buy than at the beginning, that should give you real hope for future campaigns. If it isn't better, even after aggressive optimization efforts, you'll have to think carefully about when and how to invest again.
Blow out your Facebook campaign in one fell swoop and not only do you risk failure, you risk ignorance. In the long run, the absence of learning is worse and more expensive than failure.
Comments