with data, the production does not do well, because the data may be deceptive!
no data, the product may also do good, because you may be lucky!
read these two words seem hard to affirm, but the real data reflects the magnitude of an Internet product, but also warned us to be detailed to treat the data.
a product of the data and data analysis in the end give us what?
Here are some menial attitude,
Deadly accustomed namely eating your life!
1, allows us apt understand the product Status:
(1) Each product has a target of data, such as: the number of users, login users, member, PV, UV, Posts, Video number, etc., of course, assorted depending on the product, the elucidation of indicator data will be different; these data will tell you a whole, this product is immediately what the location;
(2) emulated at the indicator data, but also There are some functions of the data can make you a extra careful understanding of products, such as: the user's old and gender distribution,
Zou Shenglong Phantom Thunder 】 【 switch, the proportion of the user's play class, the type of delivery of posts, consumers search as what cares, PV distribution of content categories;
these data we can do is to understand the fundamentals of the product, and nought more; In fact, our products actually do help is not large, however very important.
2, allows us to understand trends:
adding to the characteristic indicator data and the data, we will change frequently used data, such as: a month or several months throughout the site change PV or change a feature of the PV, landing the number of users changes, and changes in content distribution, by varying the curve, we find that the product is moving toward healthy development, or to base!
trend data gives us a quite momentous message: product evolution is healthy, is moving in the intention we expected in the development. Through analysis of trend data is easy to detect whether the product in question, whether the deficiencies, but the trend analysis is not the problem let us know specifically where,
Workers have to look the calculator skills apt sol, and solutions.
3, can help us find the problem:
further data analysis, we will find a feature for data analysis, and a functional relationship between the upwards and downstream analysis, will empower us to more clearly to know the product, those links is important, which is problematic places, such as:
(1) new feature for the product analysis, we can observe the user to use this feature changes, to make sure whether this function renowned; observing the function of the source and destination of the page the page the user to make sure the function of bringing much value to the product,
Lacoste Arixia FD Trainers, or the various facet of the whole product, accounting for what kind of percentage. Combined with the use of trend functions, we can decide whether to mushroom its investment in, or to improve his side did not encounter our expectations;
(2) process for the product the user experience of data analysis, you will also find that users How to approach, the core content of our pages, this process of distribution, will we improve the process for reference.
this time, we have products from the understanding to use data to improve the products it! This is the core value of data analysis!
4, can help us identify the user:
this namely the core worth of file thinking.
we will find, in truth, we can tap into the user's behavior data, such as: Users favor to see what kind of content, users like what kind of interaction, the user clicks ashore the sheet of the heat analysis, the an from the deeper level to cultivate our products, from product design point of outlook, to cater a more personalized entities. Technical trouble here is the suitable way will be higher!
to speak with the data we secondhand to mention, not many people really do, and the applying of real and even fewer flexible.
give a small example, when the introduction of a new feature, feel relatively nice, while suddenly he received a grumble that this feature does not work well, how do? Directly by the user to alteration it, or bypass?
recommendations are:
1, to gather such response to penetrate how many folk? What variety of person are (background)? Why do they muse does not go well, the subsistence of a private habit alternatively experience of vulnerability?
2, to the background good look at the data, the use of this feature, there may be inspected 2 days, look at the data changes and user feedback.
Finally a referee to do, via in-depth understanding and clutch the real user data, it is not cozy to be cheated by the data.
Data thinking of the understanding of the product
Linear Mode
