One of the most underrated metrics to track in my opinion is the frequency metric. It gives you a good idea of how saturated your audience is and whether or not they may be experiencing creative fatigue.
The frequency metric is what I believe one of the most underrated metrics to keep an eye on when you're monitoring your Facebook and Instagram ad campaigns. So the impressions metrics that you're probably already familiar with, tells you how many times your ads have been viewed. And the frequency metric gives you an indication as to how many of those views are repeat views from the same people. And the reason why this metric is important is because we want to prevent people from just seeing your ads again and again and again, and starting to get annoyed with them, because what can happen is they could start to develop a negative association with your brand if they're just really sick of seeing your messages.
A few ways that you can keep your frequency down, if you start to notice it's getting high, is by increasing your audience size. So you could add a new interest targeting ad in, broaden your look-a-likes, you could add in new locations to target, anything you can do to increase the pool of people that Facebook can pull from to show your ads to. So they don't need to show it to the same people. Another thing that you can do, especially for warm audiences is, for example, say you have a campaign targeting your website visitors, and you notice your frequency is getting high, you might want to just decrease the budget on the campaign because there's not really any way to increase the audience, unless you launch another campaign to start sending people to your website, which is an option as well.
Lastly, something that I always recommend doing is to change up your creative often to keep your messaging fresh and prevent people from seeing the same ads again and again. The more that you keep your messaging interesting, the more people will be more likely to engage with your ads and potentially click through and convert.