How to analyze a internet product? (the second part)

分类:望尽天涯路 | 作者:观尔腾 | 发表于2018/06/28

These recent years, as the internet users became saturated and increase more and more slowly, the network traffic is much more costly than ever before, only making a user-friendly internet product satisfying the market demand is not enough to pursue a success. We need to make sure every click counts.

Here we have AARRR startup metrics, which is divided up to five sections – acquisition, activation, retention, revenue, referral, to analyze a internet product. However, actually things do not always happen in the exact order, the last three sections can be process in different priority base on the purpose your product in different stage. For instance, if your product is still in MVP stage, the level of retention is your first priority. Because you need to make sure whether customers buy your ideas. Until you get adequate data convincing your thought, try not to pay much effort on other section.   


To analyze the product, it would be much easier if you have a reliable data platform for you to trace the performance of all the traffic flows. Many analysts tend to assess the product in noddle, driving them to crazy with nothing convictive found. As a debatable conclusion will not lead to a helpful insight, making sure you can trace the detail of your funnel from different channel will save you some unnecessary arguments.

If it was a web product, using the designed trace code attaching the link will be a good idea. Instead of building your own data platform, you can make use of some third-party platform service like google analysis to trace your link.

If it was an app, using codes to mark your Android install package for different traffic flows. When it comes to iOS app, which people can only download from APP store, we need users leave something unique for us to match the traffic flow and then trace. For example, requiring users to register to get coins for reward, we attach the register information to the specific traffic flow.

In order to make the most out of our data, we should try to make enough tags for our users for Muti-dimension analysis. I will carry it out in my next article.


日志信息 »

目前盖楼 (0)层:

发表评论 »