gg product

Q4 – pa

怎么样决定google express该用多少promotion？promotion该持续多长时间？怎样evaluate这个promotion有没有效？

google movie要发布一个coupon，问怎么run test去看这个coupon是20% off还是10% off产生的利润大，以及coupon发行几天比较好这样子，这个题就是考AB testing的，让你说一个流程然后模型用什么说说(anova这些）

We have two variables discount and duration. For each variable we have two options, so in total we have 4 combinations. We should test all these 4 combinations, because two variable would have interaction.

In this case, we should use Multivariate Test instead of A/B test, because A/B test only test one variable, but MV tests multiple variables to determine which combination performs the best out of all the possible combinations.

 

Q8 – pa

假设你是一家new app公司的analyst，马上要去给ceo作汇报了，你会怎么做？主要就是围绕汇报些什么key metrics，用哪些graph，怎么解释等等。然后问了一些A/B test相关的问题，比如说这个app可以track location进而提醒用户要不要check in，怎么样衡量这个feature好不好？

We should report the KPIs in 4 dimensions, including the acquisition, activation, retention and revenue. For acquisition, the metric we should check is number of downloads.

https://www.oreilly.com/learning/analytics-frameworks-for-the-startup-lifecycle

A/B test for notification

Objective is to increase engagement.

Metrics: average no of check-ins; average no of sessions; check-in probability is the proportion of total unique active users who checked in

It would make more sense to consider new and existing users separately. Not only does this make comparing magnitudes more appropriate, be it also lets you test for novelty effects. Users familiar with the app might try out a new feature just because it’s new, temporarily boosting their overall engagement. For new users, the feature isn’t “new,” so they’re much less likely to use it just because it’s different.

Q17 – pa

如果要evaluate user's happiness你会怎么做 然后我就扯一堆potential ways to measure user engagement as a proxy

What product do we evaluate? gmail, map

Q19 – pa

假设产品组在**国家launch了一个compaign， 目标比如说是增加 xx metric。 你怎么判断xx 的增加是由于这compaign导致的？

转载于:https://www.cnblogs.com/ffeng0312/p/10031490.html