Can you give a sense of what size budget you tested with and how much data you collected? Tests that are too small to deliver statistically significant results

Can you give a sense of what size budget you tested with and how much data you collected? Tests that are too small to deliver statistically significant results are something I see quite often. For FB, think of it as a place to connect with your audience. You probably aren't going to see a lot of click-based performance, but there might be a strong view-through correlation since it is very similar to display in that regard. So if you are trying to raise awareness of HR people for example, you can probably target them pretty easily on FB, and if you give them shareable content, that can help be a force multiplier since you won't pay for the shares. You just have to be clever about your targeting, and even more clever about how you are measuring things from an attribution standpoint. If you are expecting it to perform well on a last-click basis, you'll likely be disappointed.