3 ms·
100 customers is not remotely close to a large enough sample size to glean anything worth the hassle to do the test. 100,000 would be better, but then you're sp
by daimyoyo 15y ago
100 customers is not remotely close to a large enough sample size to glean anything worth the hassle to do the test. 100,000 would be better, but then you're spending a huge amount of money for results that might add nothing to your research. The whole "faster horse" theory applies here.
- sukuriant 15y agohttp://www.publicagenda.org/pages/best-estimates-guide-sample-size-and-margin-error http://www.publicagenda.org/pages/best-estimates-guide-sampl... 100,000 is overkill. 1,000 is reasonably sufficient.
- sbov 15y agoI don't see how "faster horse" theory applies to this experiment. The whole point of the faster horse theory is to look beyond your customer's assumptions to give them what they really want. Netflix wasn't looking to give their customers what they wanted, they were looking to inconvenience customers based upon internal logistics. In many ways they were violating the faster horse theory: asking us if we wanted better streaming service or better dvd service. They were focused on technical details (do you want a horse or car) rather than the bigger picture (better travel - I don't care if it's by car or horse or teleportation).