Facebook recently announced that it would expand its Ad Delivery system by showing advertisers a relevance score that rates ads on a 1-10 scale based on how pertinent the ads are to those who viewed them. Facebook’s base currency is users’ attention. The better the platform can cater to users, the more time they’ll likely spend on the site. Therefore, adding this extra layer of transparency makes perfect sense – advertisers will be better able to provide relevant ads, users will be exposed to things that are closer to their interests and Facebook will personalize its content structure even further, potentially raising usage rates at a time when growth in terms of number of users is slowing down in its core markets.
Now that we’ve established that this is a win-win-win, let’s take a look at how this new system works.
The 10 point relevance score of a given ad is calculated by taking into account the positive and negative interactions that users have had with the ad. Positive interactions are actions taken by the user such as liking, sharing or clicking on the ad. Negative interactions are actions that users take to avoid the ads – things such as reporting, hiding. Although the actual algorithm used to calculate the score is a secret, this new system gives advertisers increased transparency into how their ads are doing. Google has already been using a similar system, which they call a Quality Score, for quite some time, with excellent results.
So how can you use the power of Facebook’s new Relevance Score?
This new system is part of a larger movement by Facebook toward increased transparency. The end-game here is that by giving marketers more data points, Facebook is allowing them to optimize their ads on a larger number of dimensions. Until now, an ad which “didn’t work” could be the result of a variety of factors. Since most of these were hidden, a marketer would judge ads by click-through and conversion metrics alone. With the addition of Relevance Scores, advertisers now have a better view of how even small groups of people interact with the ads. Therefore, you don’t need to pump large amounts of cash into ads to do A/B testing and optimization like before. Instead, you can see how even a small population reacts to your ads, regardless of conversion. Obviously, this is to be taken as complementary to existing metrics, not a replacement. Still, it’s a step in the right direction and we hope to see a lot more fine-tuning of paid content on Facebook in the future.