One of the first articles I ever wrote as an engineering journalist was about a massive failure at a power station. The cause was an elementary design fault. There was a switch linked to an actuator, and a light which was supposed to show whether the actuator was on or off. Unfortunately, the light was connected to the switch, rather than the actuator. If things were operating normally, that would be fine. But if the link from the switch to the actuator failed (which it did), the light was showing the status of the wrong thing.
This slightly tortured analogy shows the importance of measuring what happens, not what you intended to happen. I often find that people running Google AdWords advertising place too much emphasis on reports from the AdWords end (the input) rather than from their website analytics (the output).
To take this to its logical extreme, what would happen if your website was down? The AdWords end would say everything was fine, and indeed, you would be spending money sending people to your website. However, the website analytics would have told you that nothing was actually arriving.
This is why when we report back to clients on their Google AdWords advertising, the reports we produce come solely from Google Analytics. The number of visitors being sent to the site may determine how much the advertising costs. But all that matters is counting how many arrive, and seeing what they do.
I hope you’re judging all of your advertising by the response it generates, and not by how much it’s costing you. This is the mistake which a lot of advertising media hopes its customers will continue to make. Option A is not better, just because the arbitrarily-set price is double that of option B.