Slide 21 I found particularly interesting in that they are testing micro-innovations by releasing them to specific user segments.
I'll say upfront that this is a great idea, but as a longtime Netflix subscriber I have to say the "satisfaction" polling I occasionally receive seems a bit too simple for such a purpose. For example they ask you how the quality of your stream was, and on HD content I do have to say "unwatchable", but if they are using it to triage software improvements, I wish they was also a chance to say "but I know my broadband is beyond sucky and I didn't really expect anything better than that".
In other words what matters is not just satisfaction, but satisfaction as compared to expectation.
I have no direct knowledge of how we use that survey, but I suspect that an outlier such as yourself would not affect the results of the survey. If there is a wide-spread drop in quality, I think the survey would pick that up. It's hard to design the perfect survey though, but I'll see if the powers-that-be can improve the survey.
BTW one of the nice things about netflix is how well it treats its outliers :-) For example for ages I patiently filled in the "when did your DVD get to you" survey not expecting anything better being out in the boondocks (Hawaii) and them to my astonishment, Netflix opened a shipping depot in Honolulu.
And then there was the "multiple queues" incident where the company agreed to keep the feature even though only a relative small percentage of subscribers really used it.
4 comments
[ 3.0 ms ] story [ 20.6 ms ] threadI'll say upfront that this is a great idea, but as a longtime Netflix subscriber I have to say the "satisfaction" polling I occasionally receive seems a bit too simple for such a purpose. For example they ask you how the quality of your stream was, and on HD content I do have to say "unwatchable", but if they are using it to triage software improvements, I wish they was also a chance to say "but I know my broadband is beyond sucky and I didn't really expect anything better than that".
In other words what matters is not just satisfaction, but satisfaction as compared to expectation.
And then there was the "multiple queues" incident where the company agreed to keep the feature even though only a relative small percentage of subscribers really used it.
It's a good way of doing business.