Your use of metrics to track performance and ability/desire to experiment clearly had an impact as well. In these kinds of situations where you can get good quality metrics relatively easily, and experiment so quickly, it is still surprising to see folks not collecting those metrics.

Of course, it's not ALL about metrics, and in some ways I reject the line of thinking that you can't improve what you don't measure (qualitative research/understanding can go a long way). But with quantitative research/data so easily available... it only makes sense to make use of it.

Thanks for sharing your insight!