I can't help but notice your entire post focuses on click and engagement metrics without any mention of conversion metrics.<p>You don't mention a branding strategy, or how you plan to map branding efforts to revenue (arguably a hard problem), so my assumption would be you are most interested in seeing people signup as a result of this, and in fact you have some goal completions, but it isn't clear what those goals are.<p>If your LTV is several hundred dollars, and your current campaign CPA is ~$300, while likely not statistically significant at this point, that could potentially be profitable if that cohort's quality pans out. If those are just lead goals, then yeah, you might not be super successful here, but 3 conversions seems a bit on the low side to draw that conclusion.<p>But here's where the analysis failed to go deeper...<p>- Digging into conversion metrics and how people are finding you. Do you have a long sales cycle with multiple touch points? The reports you've shown all use last click attribution in a 30 day window. If you check the Conversions>>Attribution and Multi-Channel reports, you might find this was actually a big assist channel and maybe you're getting the final conversion off of Direct or Organic (or even phone calls...call tracking rocks)<p>- Spending $1k for less than a month when presumably you have a decently long sales cycle and relatively low volume/high LTV sales and saying your test was a failure seems like the test may not have been structured very well. That's a relatively tiny test budget and making a decision after 100 clicks seems like not the best idea (unless of course you're looking at post-impression metrics, but my guess is that isn't the case)<p>Overall, this comes across as a click-baity headline while the actual article seems rife with holes in terms of laying out the details of the test, what was measured and why, and a whole host of other missing information that would likely point to this test not running nearly long enough or with enough volume to draw any meaningful conclusions.<p>As someone who does this for a living, I kindly suggest you revisit your test assumptions, methodology, how other business metrics are factored in, and whether it is good for your brand to publicly jump to conclusions like this in a way that comes across as rather uninformed, or at the very least doesn't paint the entire picture.