Detailed measurement of the key indicators of the product

I was fortunate enough to have worked in several fast-growing and interesting companies, such as Linkedin,facebook and Twitter. All I've learned is to look at the data. Don't just stay on the surface--such as PV number, registration account, etc. these things, we have to focus on the real users. In particular, see what they're doing, and why they're doing it. In LinkedIn we don't talk about total PV numbers, but instead emphasize "browsing the user data"-How many people are looking for someone with LinkedIn, and how many people are browsing the data. On Twitter, it has a lot of PV, but we're still concerned about how many people are looking at their message flow and how many are reading or pushing.

When I look at other new companies/products, I always hear them say "We have 10 million independent visitors per month, 30 million PV" or "We have 25% daily/monthly active users, 2% arpdau". All this sounds pretty scary at first blush. But a deep scrutiny will find that the so-called 10 million independent visitors are from the search engine, they randomly read two or three pages, they slip away and never come back. or the percentage of active users is 25%, in fact, because the app has recently been wildly popularized, and then quickly attracted new users, who did not stay long.

But these fascinating data, after all, are, er, just good-looking data. I don't think they can say that this product is really popular. When I hear this data, I always ask the same question:

How many people are using your product?

You need a molding measure to answer such questions. It may be "how many people have searched 3 times in the past week", or "how many people have visited my site 9 times in the past one months", or "how many people have had at least one purchase behavior on my site in the last 90 days". But no matter how it is, it should be able to show that users are using your products as you expect them to, and use them very often, and you have reason to believe that they will follow up with your product.

Once you've identified a measure that answers these questions, you can actually track the growth of your product every day, weekly, and monthly. And from then on, you can identify the key metrics that explain how users will use your product more frequently or rely on the product.

On Twitter, for example, we found that if a user visited Twitter at least 7 times on an average of one months, he would probably continue to use Twitter next month, next month, next month. So we think this indicator has been very good for users to "really use" this product.