So you're saying the sale was basically yours to lose? I would think that if the results of your tests say it was significant (assuming you ran them correctly),

So you're saying the sale was basically yours to lose? I would think that if the results of your tests say it was significant (assuming you ran them correctly), then it probably was. If you have data, then you should use it to your advantage. What exactly makes you not trust it? It sounds like the question you mean to be asking is more like: Do I have bigger fish to fry than playing with the color of the buttons on the site?