What data do you look at every day as a product manager?

Source: Internet
Author: User

What data do we need to observe every day as a product manager ? This article and everyone together to grilled a steak, I hope to be helpful to everyone.

just entered the time do not know how to look at the data, every day is looking at the background report in the UV , PV rose a little bit / fell a little bit, then there is no then. Until then the chance coincidence did a period of time user growth, channel-related work, just slowly learn how to look at the data.

Then I thought about why the first time I looked at the data, the whole person was ignorant? In the end, I think the root cause still lacks the goal. Do not know why look at the data, all the data is an isolated number, and then began to look with the goal, to find the answer to the data, slowly learning to find the link between the data, find hidden in the data under the image of the information.

in my opinion, all the specific goals are summed up as two points: Find a problem , Find a Chance . Find the product there is no hidden problems, find the design of the logic is not in line with user behavior, to find there is no potential opportunity to help products on a step.

To get to the point, I usually need to focus on the data there are roughly four categories:

1. operational data for the product, including scale data and quality data

2. user behavior data for product core scenarios

3. feedback Data After new function is launched

4. Industry Data

which is suitable for every day to see the main is Operational Data and the behavioral data for the core scenario ; Feedback Data is in a new function, or in order to verify some hypothesis of the experiment after the study; Industry Data The basic is to look at the quarterly dimension to see it.

Operational data

The most common data is the operation of the product data, I am accustomed to the scale and quality of two points of view:

· scale data is primarily a product of some data indicators, such as: The number of new users, DAU (daily active number of users), MAU (monthly active users), e-commerce products include orders, income, and so on.

· quality indicators are data that reflects the health of the product's business, such as the next day of new users, the user's start-up frequency and startup time, and so on.

Look at the data first to understand how the data is defined, the acquisition and the calculation process. Differences in understanding can lead to forced matching of a number of different data, leading to a significant error in the conclusion, especially when confronted with different products and different company data.

Figuring out the definition of the indicator, I used to use it when I looked at the data. Compare and decomposition of these two basic methods:

· contrast, is through shape factor and longitudinal ratio of the way to look at the data, shape factor is compared with similar products, and their own experience data (such as experience in the summer and winter vacation is the high season of video, but the product corresponding to the data did fall, this need to go further to find the reason); another aspect ratio is the past self in the timeline.

· decomposition, according to different dimensions to decompose the data, for example: can be viewed from the dimensions of the channel, from the geographical dimension, through the different dimension decomposition will be the difference between the comparison of the level of locking, convenient to find reasons, do the user growth, I will focus on different types of Top Channel.

The use of contrast / decomposition Dafa Basic can develop better data sense , and finally three data interpretation of the more common errors:

1.  over-focusing on the causes of the falling data, ignoring the reasons for the rise altogether, or entirely attributable to the improvement of the business. When I was doing a browser, there was a core indicator, "number of searches per person", which is "total search number / search users "calculated, so the increase in per capita search number may be the total number of search growth, may be the number of search users fell, it is necessary to further analysis, simply think that the rise is good, the downside is bad is problematic.

2. Cause and effect attribution error, the relationship is wrong to think causality, or ignore the key factors; for example: a leads to the occurrence of B and C , The analysis ignores a , and directly considers that B and C have causal relations.

3. survivor bias, ignoring most of the silence, is a mistake that is easy to make when looking at sampled data, and a detailed definition of survivor bias can be turned over to Baidu Encyclopedia.

User behavior Data

There is also a suggestion to look at the log data of user behavior every day , this data is a bit like " Baidu statistics "Inside the funnel model, but he is more detailed than the funnel model, he can not only explain the user has not walked to the funnel, but also can see further, the user in the funnel path, and how to jump out of the user how to jump out."

The user's behavior data are some raw data, the data volume is large, needs to have certain processing.

1, find the core scene of the product

Not all users ' behavior logs are to be seen, but to find the core scenarios that affect the user's cognitive products, which can be borrowed from ( MOT, Moments of Truth ), is the point of contact between the user and the product's service, and the experience of these points determines the user's evaluation of the product as a whole. For example, do users need to look at the new user to see the path of the service they want to find out what it looks like, is not short enough? Have you encountered any difficulties?

2, for the user classification, to find the target user

It is best to look at a class of users at a time, because looking at the user's behavior data is more energy-consuming process, not every time can have a harvest, need to constantly look, and constantly digging.

So in the new time, I will basically use the hook to set the drainage when the classification, each time to see one type of users, for example, through the video today only to see the user to accelerate the drainage, to see if these users can quickly find the corresponding video, after a few days to see through the information in the user.

when you're done, see the data and say something simple. Feedback Data and industry data , although these two data do not need to see every day, but in the work of the product manager is also very helpful.

New function Feedback Data

iteration is an important work tool for product managers, whether it's grayscale publishing, The AB test , or the regular version, needs to gather data to validate the previous assumptions, to think of continuing optimization, or to push back, and iterate to help the product manager accumulate proven knowledge.

The most important thing to look at in the feedback data is to think clearly when designing, what is the goal? What is the logic that I think this new feature will work? What data do I need to collect to verify?

The best way to do this is to write down the answers to each of these questions, by writing down the method to ensure that we have taken into account the problem of better detail in advance, to avoid two common mistakes.

· Collect data excessively, increase development workload;

· Insufficient data collection, unable to carry out conclusion analysis;

Industry data

finally say the industry data, I used to use the industry data to find new opportunities, industry data can see a large trend of user migration, if their products can borrow this trend, it is likely to be a wave of larger breakthroughs. For example: The rise of the WIFI master key, the growth of video traffic in the year; The rise of today's headlines and the number of quick-racer users.

Industry Data acquisition is more dependent on the big platform, if big companies such as BAT have large enough user samples to see this data more easily and in a timely manner. If there is no such condition, on the one hand can rely on the third-party reports such as Eric, on the other hand is to pay more attention to various ranking version of the data, such as:App Store , Application Market list, micro-blog key Hot List, Baidu index and other changes.

Overall is to think about, the user recently pay attention to what? This thing and its own business has no connection, should not be forced to contact.

look at the data this skill is also more practice, there is really "data sense " This kind of thing, often look at the data more sensitive, more easy to find information hidden in the data.

Source: Everyone is a product manager

What data do you look at every day as a product manager?

Related Article

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.