We recently ran a two-month campaign with the goal of driving a new branding message for a client that is in an industry with many strongly held opinions. Because we weren’t driving a lead or some sort of conversion, we couldn’t turn to our usual Pixel Ads and instead used Facebook’s Awareness ad type. The main goal of this ad type is to generate Ad Lift, which is the number of people that will remember your ad message after two days, implying that people are receptive to your message. This is highly effective for general branding as well as driving leads into funnels with long sales cycles.
Testing with Polling
Facebook claims that they test their estimation algorithm by polling thousands of people to verify their accuracy. Of course, we couldn’t take that at face value and felt we had to test this ourselves so we could confidently report our numbers to our clients. We called in a friend, who has run polling for presidential campaigns, to set up a test and control group with rolling data throughout the two month campaign. We measured changes in opinion on multiple levels and disguised what information we were truly after by including many unrelated questions in the surveys. Once we had everything dialed in, we were off to the races.
Ad Recall Lift Rate Results
Our Ad Recall Lift Rate reported by Facebook averaged around 18% (after a few weeks of optimization of course) which is very high based on the research that we have seen and the type of creative that we were using. We were anticipating around an 8% Lift Rate prior to starting the campaign based on reports from the industry. We were thrilled to see such a high rate, but the true test would come after reviewing our own polling.
As a side note, we were able to achieve a Frequency of 3.5 over the course of the two months among our target demographic, which certainly helped contribute to our high rate. This is just a friendly reminder to not stretch your budget too thin or you will certainly see lower performance.
Are Facebook’s Metrics Accurate?
The short answer is that the Facebook metrics are very accurate. We compared the metrics to our own polling questions on a number of different levels, from remembering specific messaging to changes in general opinion about related subject matter. The metrics for most of the ads were within one or two percentage points with only one being off by three percentage points. This was far more accurate than we anticipated simply because of the variability in types of messaging.
We also continued to measure the lift rate for four weeks after the ads ended. There was a fall-off in the lift rate of about six percentage points over the first two weeks, but then it held fairly steady over the course of the second two weeks. This is highly dependent on your messaging and audience, but has potentially powerful implications for how you run your brand awareness campaigns over the long term.
The Facebook Ad Recall Lift Rate proved to be a very accurate measurement of how effective your ads are. We feel very confident reporting this number to our clients as another tool in our box to understand how consumers are behaving. And while this metric is focused on recall lift after two-days, the rate remains high even after a few weeks. This, of course, depends on your messaging and audience, but this reinforces the idea that it is much easier to keep a customer than it is to win a new one.
As a parting note, while many of you are hyper focused on generating leads with your ads (and rightly so), just remember that the stronger your brand, the easier it will be for you to convert on every other marketing channel.