I wouldn't put much stock in that. If the results were 2/17 vs 7/20 that would not be a statistically significant result; in any case, there could be other maj

I wouldn't put much stock in that. If the results were 2/17 vs 7/20 that would not be a statistically significant result; in any case, there could be other major factors such as time-of-day and experience-building which are not discussed. The job applications limit and 2 week period of this trial suggests that the sample sizes were not any bigger than that. It's likely just a coincidence.