2 cases to teach you the correct approach to user operations!

Everyone knows about user operation , but most people don’t really understand what user operation is? In this article, the author will share with you some of his understanding and views on narrow user operations, and use two cases to further help you understand.

Many people who work in operations in companies are busy organizing various activities, catching hot spots, and doing special topics every day. Those who are more advanced look at data and build systems every day. Anyway, they use all kinds of high-end, cool and fancy words to promote themselves.

But if you observe carefully, you will find that the user operations of most companies are simply terrible. This situation exists even in many unicorn companies.

Let’s first look at the recruitment requirements for [User Operation] of a certain company (the company is a well-known education company):

Responsible for the establishment, maintenance, and improvement of the user management system, as well as user retention and life cycle management. Responsible for the basic operations of online activities, able to complete customer content planning and basic execution, responsible for community atmosphere/content creation. Collect and analyze user behavior data and user feedback, provide suggestions for event operations, content operations, etc., and collaborate to improve various indicators such as new additions, activity, retention, and transaction volume. Responsible for using personalized and refined operational methods to enhance user experience and increase new user conversion investment rate.

Can you tell what user operations are by looking at the recruitment requirements?

Most people are talking about user operations, but in fact most people don’t really understand what user operations are.

If we define user operation in a broad sense, traffic operation, content operation, and product operation will ultimately be attributed to user operation, because all operation systems ultimately serve users.

Therefore, the user operation referred to in this article is a narrow definition of user operation, which makes it easier to focus on most of our on-site work.

1. What does user operation in a narrow sense do?

Simply put, user operation means: through different stages of the project, customers are divided into different categories, important customer groups are identified, and then key categories of users are guided for conversion.

So, you see, even in the same company, the work done by user operations is not static. For example, for Didi Chuxing, the work of user operations in the early stage of seizing market share and the later stage of expanding GMV may be completely different.

If we want to briefly explain what operations actually do, it is roughly divided into the following three parts, of which 2 and 3 are usually cyclical:

New users/downloads/registration retention/conversion/paid activity/stickiness/repeat purchase

Many times, the scope of 1's work will overlap greatly with that of market and marketing. Even the growth hacking we have been talking about is part of 1's job content.

Therefore, the working logic of 1 is usually different from 2 and 3 to a certain extent. User operations are mainly concentrated in parts 2 and 3.

Point 2: User operations in a narrow sense involve three things:

1. Classify users into categories

Many operators are dismissive. Isn’t user classification an easy task?

User portrait, user management system, user classification system, you know there are too many professional terms used to describe the user classification.

I have seen many operations generally divide users into: new users, interested users, paying users, active users, etc. It seems reasonable, but if it is really divided in this way, it is too general and rough. It's easy to get lost while you're working.

Users are living individuals, and their behaviors are expressed as pieces of data, but users are by no means pieces of data! In fact, user behavior will be affected by a large number of factors such as time, place, environment, and social events. Therefore, operators should build a large number of different user models for their products.

For example, I once made the following user classification for a certain product:

Those who paid more than 5 times before the 2.0 version update (this may involve major version updates) Those who did not pay before the 2.0 version update but converted to pay (this classification is not scientific for products with low average order value, it is suitable for products with high average order value) Customers who did not convert and lost within 3 months of registration (customer portrait and characteristics) Customers who did not convert within 3 months of registration but have been logged in (customer portrait and characteristics) Users with real comments of more than 30 words...

Users of a product can be divided into about 20-30 customer types. Moreover, some user types may only exist in the operation system for one week, and then quickly become antique data recorded in the operation log.

There are actually many methodologies on how to classify users. We will have the opportunity to discuss 1-2 of them in detail in the future!

Once your product can classify users in this way, you can then think of solutions for each type of customer.

1. What did these key users do?

When you have classified users into nearly dozens of methods according to the above method, you will quickly find that there are serious problems with one or more categories of users in this classification.

For example: At different stages of a project, you may focus on different data.

Sometimes, you may be more concerned with overall conversion users. Sometimes, you will pay more attention to repeat purchases, daily active users, and high-stickiness users.

If you want to efficiently increase operational data, it will be easy as long as these key types of users can achieve a certain degree of change in quantity.

And you can definitely use statistical software (each company’s statistical software and methods may be different, so I won’t go into details) to find their key landing pages, bounce pages, and other similar behavioral characteristics.

Sounds simple, right? Believe me, this step is definitely not easy. Just try it and see.

1. Interrupt and convert these customers

Once you find one or two key types of users, you can optimize activities, pictures, prices and other information on their landing pages and pop-up pages that are obviously different from other users.

Here are some examples:

When you find that users with an average conversion rate higher than 3% have visited a certain page, you should try to push these pages to people with low conversion rates and conduct corresponding data tests. When you find that a considerable number of churned users have visited the after-sales policy page before churn, you should quickly check whether there are any major loopholes in these after-sales policies that you have not noticed.

I believe that by now, these will no longer be difficult for you.

As for how to interfere with these users? I believe that the opportunity to show your talents in event operations and content operations has come!

Case 1:

Previously, there was a startup company that was a gas station APP invested by a certain chemical company. Users who downloaded this app to refuel could enjoy certain subsidies or discounts. The discounts were quite good. Filling up a tank of gas could basically save them between ten and several dozen yuan.

During the initial promotion, we utilized offline gas stations for ground promotion, as well as the guidance of gas station salespersons. In addition, with certain preferential subsidies, the APP achieved good results in downloads and order volumes.

The situation seemed very good at the beginning, but once these industries stopped, the data immediately declined. How serious was it? The data is not even a fraction of the original amount!

After we carefully analyzed the characteristics of the user groups, we found a very typical customer group:

This type of customer is the key group that can most efficiently increase data.

We further investigated the behavioral characteristics of this type of users and found that compared with other users, they open the “Daily Specials” page significantly more frequently than the “Buddhist” user group (who often use the APP to refuel, and use discounts if there are any during the checkout process, and don’t care if there are no discounts, just to earn points).

At this time, they added small elements of development games to the subsidy activity page. For specific types, you can refer to the gameplay of Alipay’s Forest and Chicken. Just follow theirs.

Users originally wanted to take advantage of the app, but were eventually attracted to the app by the need to develop their own products. And were converted into target customers.

Case 2:

A friend of mine owns a company that provides online training. They put different types of quality-oriented education business courses on an APP and website, and each institution is independent of each other. Although the data is not too bad, we feel after discussion that there is still the possibility of further improvement.

Here are the reasons:

The proportion of users who sign up for a certain type of customer individually is as high as 90%. Users' needs in learning calligraphy and taekwondo are almost highly overlapping (because they want to go). About 3% of users were attracted by packaged courses and signed up for cross-disciplinary courses. In theory, we can transform the need to learn into the need to show off to some extent. (For example: turning the need to keep fit into the need for users to post their works on WeChat Moments) The proportion of calligraphy users posting their works in the "learning circle" is much greater than that of other learning users.

So, they added user honors, relying on a medal that was both civil and military (in fact, this medal system was very complex, and there were also physical manifestations, such as mailing it to schools for users, etc.), and carried out targeted guidance in the learning circle, converting a considerable proportion of calligraphy students into Taekwondo students.

This means that while keeping the original data unchanged, the unit price of user payment may double, which is also a good example of user operation.

Then, classify your current users as complexly as possible, and then find their common points (behavioral characteristics). On the key paths of these behavioral characteristics, conduct operational interference to increase data...

It is the methodology of user operation in a narrow sense. Do you understand it?

Source: Barley