In the wake of the Cambridge Analytica scandal, Facebook is running its own advertising campaign. The ads proclaim how terrible Facebook was, how it compromised your data, and how it’s going to win your trust back.
In theory, that’s good. Improving privacy is a win for everyone, right? Well, not necessarily. It may not help Facebook’s advertisers.
Advertisers have long benefited from Facebook’s extensive targeting capabilities, which include the on-site behavior of Facebook users as well as data from third-party “partner” companies. But in late March, Facebook announced the phasing out, in six months, of both behavioral and partner data.
Partner Categories let advertisers target ads based on income, shopping behaviors, and more. It’s a good place to start if an advertiser does not have a strong audience. But the data from partners is not perfect. For example, if you created a campaign that stipulated only people with annual income of $100,000 could see your ads, there’s likely going to be users that do not earn that amount see them, too.
Some observers believe Facebook will reverse its March decision due to the importance to advertisers. But I wouldn’t bet on it.
New Options Coming?
Beyond the elimination of behavioral and Partners Categories, Facebook must also overcome its heightened consent rules, where users decide what to disclose to advertisers. Thus many advertisers, including me, believe Facebook could roll out two new advertising options: an in-depth version and a limited one. Consider it a “ads advanced” and “ads lite.”
Facebook would have two goals. First, make up for no partner categories and for the extra consent rules. Second, for those who don’t provide consent, serve a more general (and less targeted) ad. This could be interest targeting only with a few categories. My guess is the “ads advanced” version is more likely because Facebook is starting to run into a return-on-ad-spend problem.
Advertising costs on Facebook are broadly increasing for a variety of industries. Big brands are starting to notice and are considering lowering their spending levels. The culprit is advertiser competition for lookalike audiences. There are only so many people to target with a default setting of “paying customers from [X] country.”
So if you don’t hyper-target your audiences and own them, you’re losing out. For now, adding extra interests on your lookalike and updating your creative as often as you can is the best thing to do.
Finally, there’s the GDPR, the E.U.’s General Data Protection Regulation, which became effective on May 25. In a nutshell, it requires express consent from E.U. residents as to what you’ll do with their data, how you’ll protect it, and how they can manage and remove it. For Facebook, you’ll need to certify that the email addresses you upload for custom audiences have that permission. I don’t expect much immediate change from the GDPR, though the rollout of an “ads lite,” which would presumably require little to no permission, could come sooner as a result.
In short, from my perspective of an advertiser, Facebook’s data debacle has not been too disruptive. Certainly there are Facebook doomsayers and, yes, there are problems, such as too many advertisers for too few users.
Stick to the basics and closely monitor performance.
- Check targeting. Are your lookalikes audiences still performing, or have they hit a wall? Are you getting anything useful from the soon-to-be-sunset behavior targeting?
- Review costs. Check for cumulative reach and impressions, as well as frequency. If costs are rising but reach and impressions are not, you’re likely hitting the same people repeatedly. And this is a good time to remove behavior-targeting campaigns or plan a strategy to phase them out. You don’t want behavior to be driving 80 percent of your leads or sales only to have that option go away.
- Review ad creatives. Have you been using the same text and image combination for months? If so, it’s time to replace. As a rule, I try to keep two or three ads in reserve, for quick replacement options. Keeping an eye on engagement. An ad with a healthy number of likes and (positive) comments will perform better than one with no engagement.