What is user growth operation and what does it do?

From data products to user growth operations , here are my thoughts:

1. What is user growth operation and what does it do?

To explain user growth operations, first of all, we need to clarify what operations are and how to do user growth. Regarding these two issues, I would like to quote the opinions of two industry insiders with whom I agree.

Operation is to deliver product value, build content ecology, create new gameplay through users, content, brands and other operations, better connect products and users, and achieve the ultimate goal of the product - Han Xu

" Growth hackers " are a group of people who use data to drive marketing , market to guide products, and use technical means to implement growth goals. This requires them to not only understand technology and be able to write code, but also understand human nature and be able to capture the users' psychological feelings and real needs. Most importantly, they can often come up with sudden ideas, use their creativity, and use their imagination to acquire more users and revenue with small investment. – Growth Hacking

So in my opinion, the work of user growth operations is to break down growth points and growth formulas through business analysis , and through experimental testing and data analysis , drive rapid product iteration , find growth directions, deliver product value to users, and achieve product growth goals. At Didi , the city growth team I work in will collaborate with product BI R&D and other departments in the form of project FT, and guide business development through methods such as AB testing and data agile analysis, so as to quickly iterate products and achieve growth goals.

In order to achieve user growth, in addition to having an in-depth understanding of the business, breaking down the growth formula throughout the entire product process, and looking for growth space, you also need to continue learning and find new growth directions through new dimensions. For example, data analysis is the most important skill for user growth. At the beginning, I only knew how to use Excel pivot tables, vlookup and other functions to process large amounts of data in college. At NetEase, I learned to use agile BI tools such as NetEase Youshu to perform data visualization analysis, and found growth space through dimension drilling and user segmentation. When Didi switched to Tableau for visualization analysis, I learned to use SQL to run data for quick analysis directly for the new decomposition dimension and started to get involved in algorithm tuning. For user growth, every new experimental method and new data analysis dimension is a possible new growth point.

As for why user growth operations are needed, in this era when the mobile Internet traffic dividend is no longer there and even offline traffic is in fierce competition, it is even more necessary to achieve product growth goals through refined operations . The role of user growth operations is particularly prominent.

For example, for content operations , traditional operations mainly rely on long-term accumulated experience to grasp the content needs of target users , and select topics and produce content in a model similar to that of traditional media editors. As for user growth, we prefer to shift from experience-driven to methodology-driven , relying on user group segmentation and rapid experiments, data-driven iterations, and further finding optimization space based on experience. Only in this way, driven by methodology, can we achieve the distribution of targeted content and tailor it to each individual, rather than relying on the editor-in-chief's experience and personal inclinations.

For example, for e-commerce operations, if the company's revenue indicators are broken down into traffic conversion rate and average order value, then the main focus of traditional operations is channel operations to attract more traffic, and to improve conversion and average order value through the skillful use of various resources and tools of category operations. The idea of ​​user growth may be more based on the specific data of the channel, and the corresponding division of user groups and exploration of operation methods should be carried out. Through data-driven, different operation strategies can be matched for different users to achieve growth indicators.

Some people may think that user growth is simply a change from relying on experience to relying on experimental data verification, and it is not a brainstorming-style bottom-level innovation. But in fact, brainstorming-style innovation is an unsustainable innovation that relies on inspiration, while rapid iteration of experimental and data-driven operations is a sustainable innovation approach . Even if only small improvements can be made each time, continuing will definitely yield greater results. Ultimately, when the entire product process has found the direction of growth, this is a great innovation.

Relying on user growth as a single indicator seems to be able to call the shots in the primary market and make VC/PE lick your boots, but in the secondary market, there must be several growth factors that can be used for maneuvering, otherwise you will be slapped in the face in minutes.

Excellent managers are masters of rhythm. They know when to adjust troops among several growth indicators. They will not wait until a single indicator is almost "killed" before looking for reinforcements. They know how to rest and recuperate. For example, Facebook let ad load rest for several quarters until Instagram began large-scale commercialization and then reactivated this indicator.

A greater operator is one who can rewrite the company’s revenue formula and expand its territory. This is a magnificent adventure and the most exciting journey. For example: I never laughed at Boss Ding raising pigs. NetEase’s e-commerce revenue from Yanxuan and Kaola accounted for 20% of its revenue in Q2 2017. If this proportion continues to increase, NetEase’s revenue formula will be successfully rewritten. -<The Baton of Growth> Zhu Shiyu

Finally, as mentioned above, the ultimate goal of user growth may be to find every possible growth point, sort out the product's growth formula, and reasonably allocate and balance growth indicators according to growth needs, find the appropriate growth rhythm, and even rewrite the product's growth formula by exploring new growth points.

2. “Operation is a basket, everything goes into it”

When it comes to operations, it seems to be an industry with a low threshold. Compared to "everyone is a product manager ", it seems that "everyone is doing operations", but when it comes to the actual work content, they seem to be quite different. Those who are engaged in community operations spend every day in the core user group to keep the group active; those who are engaged in user operations are busy doing data analysis and processing user feedback; those who are engaged in new media operations , in addition to writing for public accounts , they also grab hot comments on Weibo.

Behind the jokes are the complex and diverse types of operational functions : user operation, content operation, event operation , content operation, community operation, e-commerce operation, channel operation, and new media operation.

Even the specific positions and work contents of different companies at different stages are not the same. For example, the user growth operation of Didi Hitch , where I work, should be classified as user operation. Since it faces hundreds of millions of users, it is more inclined to strategic operation. It divides the corresponding user groups according to user attributes and formulates corresponding strategies to achieve user growth through attracting new users, activating, retaining , converting, and spreading. However, in medium-sized products, the main work of user operation may be to promote the activity and conversion of core user groups and motivate users to produce high-quality content. In small-scale startups , due to the small scale, user operation may be more about processing user feedback and operating user groups.

In the early stages of a product, community operations may focus more on the introduction of angel users and the preparation of community content to achieve a cold start . In mature communities, the focus may shift to user content distribution mechanisms and incentive systems, and the construction of a community ecosystem. For example, new media operations are the work of the marketing department at Didi.

Therefore, faced with a large and complex operating system, rather than rushing to achieve " full-stack operations ", I prefer to delve deeper from the perspective of user growth operations, trying to summarize a set of operating methodologies that are suitable for myself in practice, and try to apply them to expand into other operating areas, gradually moving towards true "full-stack operations".

3. There is not necessarily a clear distinction between operations and product positions

At Didi, I joined the user growth team which belongs to the operations group. I thought I would have to go through a period of adaptation from NetEase's data products to Didi's user growth operations. But looking back now, from business data sorting and analysis, to data and experiment-oriented business decisions, to project promotion and team communication, there are actually many similarities in the daily work of both sides. It's just that the perspective of thinking may have shifted from the needs and logic that the product focuses on, to the processes and details that operations focus on.

Of course, this is also related to the particularity of the user growth direction of the operation I am in and the company's departmental structure.

When I was at NetEase, the operations of the business unit I was in was mainly content operations (editing), and the ability to control demand was limited. After the demand was put forward, the product needed to re-research and evaluate to clarify the demand, produce product solutions and deliver development. It was the product that took the lead in promoting the progress of the project. After the plan review, the operations department had less involvement in the project. Basically, apart from regular progress communication and joint exploration of project problems encountered, they directly waited for delivery and acceptance.

In Didi Hitch's user growth operations group, for operational activities such as Car Owner Day, the preliminary research and planning of the project, the planning of activity models, the division and promotion of project nodes, and the data analysis and optimization iteration after completion are all operations work. The product is more involved in the interactive output of prototypes and communication with development schedules. It is the operations team that takes the lead in the project and assumes the role of project manager, who is responsible for the entire process from project goals to project outputs, and this tests the operations team's ability to control projects. (I originally thought that as an operator here, I might not need to draw a prototype, but in fact, in order to communicate and plan activities with the market, product, and design, in addition to the basic activity process and prototype, even the abnormal flow is sorted out, and if it is further refined, it can be regarded as a PRD.)

So I am very lucky that this time I can experience a relatively "painless" transformation while gaining a deeper understanding of and participating in the actual business. This is also my biggest expectation for this transformation. Compared with the previous data analysis, which could only be used to monitor and review operational results and make optimization suggestions for the business, here I can deeply participate in the business, refine operational strategies through user group segmentation, verify possible growth points in various business directions through AB test data analysis, and guide business decisions based on experimental conclusions. In my opinion, this valuable opportunity to shift from theoretical data analysis to using data to make decisions is invaluable in deepening business understanding and promoting personal growth.

4. Why are products and operations incompatible?

When it comes to products and operations, although they need to work closely with each other, there seems to be a subtlety in the relationship. Operational needs always seem to be not given priority, and the revenue from planning activities for new features goes to the products; products also seem to always be faced with demands from operations that have no product thinking at all.

So when I first switched from product to operation, I was still a little worried about whether I could maintain efficient communication and collaboration with the product team while considering issues from the perspective of operational goals as an operator. But maybe because I have previous product experience, I am familiar with the common details that need to be confirmed for product operational needs, and I prepare in advance. So as an operation here, it is relatively smooth when I raise requirements with the product. However, in actual communication, I also realized why there may be miscommunication between operational products. In my opinion, one important reason may be the difference in thinking perspectives .

For operations, the responsibility is to go from goals to results, focusing on process, refinement and data-based thinking.

For a large-scale event, the focus of operations tends to be on breaking down the event from its goals, clarifying the indicators to be achieved in each link of the process, and the division of team responsibilities, ultimately achieving project results. For example, the indicator of active car owner growth can be broken down into "active car owner growth = new car owner active growth + old car owner active growth" , and the active growth of old car owners can be broken down into "active growth of old car owners = active car owner retention improvement + silent car owner activation". The action of activating silent car owners can be broken down into target user selection, selection of user contact copy contact method, guiding post-contact conversion and other processes, and dividing the growth indicators into each specific link, paying attention to multiple indicators such as contact rate, click-through rate , conversion rate, etc., and finding growth space through continuous experiments. In terms of division of team responsibilities, new car owners may be more closely related to the brand building of the marketing department. In addition to improving through operational actions, the car owner retention rate is more closely related to product mechanisms and experience, and requires joint responsibility of the product department. For operations, when facing the final KPI, only by breaking it down into each process link and clarifying each time node and responsible department can we always have clear goals during the project cycle, effectively advance progress, and complete the task together.

For products. Responsible for the process from demand to solution, focusing on logic, details and input-output ratio.

Similarly, for a large-scale event, the product may pay more attention to the logic of the needs, whether it truly meets the user's needs in the usage scenario, and whether the logic and details of the needs are clear and feasible. After all, only by coming up with a complete process and preparing for possible questions in every detail can it pass the technical review smoothly without being asked questions, and the time estimate of the technical schedule is relatively accurate. At the same time, facing an endless pool of needs, the product also needs to pay attention to the input-output ratio of functional development, and may reject some operational needs with relatively low returns.

5. How to communicate efficiently between product and operations

In fact, whether it is product or operation, the common goal is to make the product better and more popular with users. It is just that due to different positions, perspectives, and statuses, the opinions may be different. In order to achieve efficient communication, we must put ourselves in other people's shoes and find common ground .

Regarding the logic of demand, whether it is product or operation, sensitivity to user demand scenarios and empathy from the user's perspective are basic conditions. As the person closest to the user, the operation should have a more precise control over user needs. In terms of demand logic, as long as we communicate patiently, clarify user needs and specific scenarios, and have data to support it, we should be able to reach a consensus.

As for the details of the requirements, they are actually similar to the refinement of operations. We will segment users and perform different actions, break down the processes, and conduct ab tests and data analysis at each link to find the direction of growth. When making requirements for products, we should also clarify the details of the products to avoid unnecessary rework. Of course, the sorting out of details should be done together with the product. The product will be more professional if we sort out the logic from demand to product logic and eliminate some common technical pitfalls in advance.

Finally, the input-output ratio may be the most controversial issue. For operations, if a small product change can save a lot of operational energy and improve the effectiveness of operational activities, then of course people will think that the input-output ratio is very high and should be given higher priority. However, for products, there may be large-scale version requirements waiting for scheduling at the same time. These requirements involve improving the experience of all users and naturally have to be given priority. In comparison, operational requirements naturally do not have a high priority.

So, in the face of such cognitive differences, how can we synchronize the cognition of both parties in product operations ? The best way is to abandon subjective judgments of importance and speak with objective data.

6. How to speak with data

How to speak with data can be divided into the following situations according to the situation of the project:

1. Have clear data to prove

For functions that have already been launched, analysis can be performed based on existing data to prove the importance of the demand. For example, with the development of business and the expansion of user base, the original user operation strategy may no longer be applicable to all users, and different strategies need to be applied to different user groups. In this case, the previous data can be analyzed in layers. When the stratification results show that different types of users have obvious stratification in their acceptance of the same strategy, the product naturally needs to be changed accordingly.

2. There are similar strategies for reference

In the absence of clear data, in order to objectively illustrate the problem, you can make estimates based on the data of previous activity strategies. For example, based on the cheating interception situation of previous activities, explain the necessity of anti-voucher swiping and anti-cheating strategies for this activity. Some data can also refer to the situation of competitors. For large-scale strategies that can evaluate the effects by observing competitor data, it also has certain reference significance. However, it does not mean that you must catch up with a strategy just because the competitor has adopted it. Instead, after analyzing the product user group and verifying that it is the real need of users, refer to the data to make expectations on the effects.

3. MVP verification of pure new strategy

What should you do if you have never used a similar strategy and don’t even have a competitor to refer to? This is a common problem for user growth operations. In order to avoid wasting manpower and material resources on developing functions, it is common to use more traditional methods to conduct MVP experiments, manually complete the functions expected to be achieved through product mechanisms, and use data analysis to decide the importance of functions. For example, for the activation of silent users, in addition to the common regular event touch and reward touch, if you want to increase the activation rate through emotional and personalized touch such as birthday touch, it will definitely require the support of product mechanisms in the long run. However, in the stage of demand verification, it can be achieved by manually running a batch of birthday user touch every day. With a certain amount of data as support, it is possible to verify the importance of demand.

7. How do products and operations collaborate?

After narrowing the cognitive gap by using data, how should product and operations collaborate in the entire process of product launch? The following is my summary of this period:

First of all, after the boss makes a decision on the direction of product development and the team has a unified understanding of the project background, goals, and direction, product and operations should begin to collaborate to analyze the users, needs, scenarios, solutions, market competitors, operating models, etc. involved in the product functions, and clarify the product function framework, operation plan, key project time points, etc.

Then, according to the plan, product and operations began to advance their work separately. Product researched users, planned prototypes, communicated technology, and operations prepared traffic introduction, user education, and content filling plans. We will also regularly synchronize work progress with possible problems and risks, and jointly explore and improve solutions. Before entering into development, we will conduct a comprehensive review of the plan to identify possible problems. After confirmation, we will enter into development. The product department will follow up on development issues, the operations department will prepare operation plans , and the project progress will be synchronized regularly.

Next, before going online, we will jointly accept the product plan to confirm whether it meets product expectations, confirm the operation plan and preparation status, summarize the problems encountered in the project, discuss the second phase plan, and confirm the function online.

After going online, the product focuses on user feedback on functions, bug situations and user acceptance. While operations are advancing the operation strategy as planned, it also pays attention to whether the product data meets expectations and analyzes the reasons, looking for problems encountered in operations and room for product optimization. After verification, they are converted into product demand, and a new round of cycle begins.

8. Why “Everyone wants to be a product manager”

Although product and operation are both important roles in Internet companies, compared to the popularity of "everyone is a product manager", operation seems to be a little "neglected". Even when they heard that I switched from product to operation, some classmates who used to work on product with me expressed a trace of regret.

Of course, in actual working situations, I don’t think operations has a disadvantage compared to product. They both collaborate for product goals. Product is responsible for the design and follow-up development of functions, and operations is responsible for achieving product goals and delivering product value through operational means. Although there are differences in the content and direction of the work, it cannot be said that there is a distinction between high and low. In the campus recruitment of companies such as Tencent, product planning and operations are recruited in a unified manner, which also indirectly shows that in terms of basic capabilities such as demand analysis, data analysis, and project execution, products and operations have the same standards.

So why do people prefer to work on products rather than operations for non-technical positions, and even many operations people want to switch to product development?

I think it may be mainly due to the enlightenment effect of "everyone is a product manager". People have a stronger understanding of product managers than operations, and have a stronger tendency to look for internships and jobs . In the recruitment process, companies also define more general positions as product managers (assistants) in response to this situation. Even data labeling and processing user feedback, which are basic jobs without much technicality, are defined as product assistants, which further lowers the threshold for products and attracts more newcomers. (In fact, there may be a certain gap between having a sense of product and being able to make a good product. For example, new students without a technical background may not fully understand the possible problems in technical implementation when scheduling technical reviews. They may be easily stumped or give difficult-to-implement solutions. In fact, it is not as good as imagined.)

In addition, the content of operational work in different directions is different. Content operation, new media operation and other directions require a certain foundation in writing skills and have a relatively high threshold. Many operational internship positions with lower thresholds actually involve basic work such as uploading product content, collecting user feedback, and establishing user group notifications. This also makes students who hope to engage in operations confused about the future space for operations. (When I first switched to operations, I wanted to find a few classmates to exchange ideas and learn from their experiences, but after chatting, I found that although they were all operations, the work content of channel operations, new media operations, and user operations was quite different. Although ideas can be used as reference, in terms of actual working methods, I still have to do more and think more about the methodology that suits me.)

Finally, influential public figures such as Robin, Zhou Hongyi, and Zhang Xiaolong have also defined themselves as product managers, which further idealized product managers. While attracting talent, companies have also paid more attention to product managers, further promoting the popularity of product positions.

But in fact, as a novice product manager (assistant), of course you have to do a lot of basic work such as following up trivial needs and handling user feedback. Product operation work also exercises product thinking, data analysis and collaborative communication skills. It cannot be said that there is a difference between high and low. You should still make a trade-off based on your own interests and abilities, the company's culture and job content.

9. Tell it to your former self

After two months of internship, in addition to the accumulation of work experience, I have also summarized some work methods, which I would like to share with you here, and also tell my original self.

1. Start from scratch

No matter how much experience you have accumulated before, when you go to a new company, you should have a zero-based mentality and start from scratch. Every company and every business line has a set of accumulated rules and methods. If we use the old ideas accumulated before to apply them to new and unfamiliar scenarios, we will inevitably encounter problems. The experience accumulated in the past will certainly be helpful in our work, but different project directions, different boss styles, different colleagues' personalities, and different department relationships all require us to adapt to the environment again. Only by having a solid understanding of the project goals, business status, and work content can you better integrate into the new environment and find a new rhythm.

2. Time management and execution

As an operator, it is inevitable to encounter problems such as many parallel tasks, many project docking people, and a lot of temporary inserted work. If you cannot manage your time efficiently, you may end up with a busy day but the project does not progress. How to handle things efficiently and improve execution? You can start from two aspects: self-management and project management:

For self-management, you can sort out tasks in advance before work every day or after work the day before, arrange meetings and discussions at specific times, and then arrange work priorities according to importance and urgency. It is inevitable to interrupt the work at hand in the face of temporary needs, but sorting out tasks can help us prioritize temporary needs and avoid missing the deadlines of truly important tasks, which would not be worth the loss.

For project management, the most important thing is to identify the project contact person and the estimated completion time. For projects that are difficult to determine immediately, at least the time for the next communication should be clarified to avoid a situation where you can only passively accept notifications. You can actively control the progress and clarify responsibilities. In addition, a better project management awareness is to regard each task as a project that you lead, and think about the project's goals, resources, plans, data, and progress. This also helps us think carefully, improve, and do every small task well.

3. Manage Upward

Upward management is a complete theoretical system. I was fortunate to have some basic understanding of it through a sharing session. I would like to share my understanding here. The most basic upward management is to think from the boss's perspective, understand the boss's goals, provide the boss with the information he needs most when he needs to make decisions, seek the boss's resources, and report project results after the project is completed.

You have to understand that after your boss has communicated the goal to you, he is asking you to execute it. He hired you to save his time and energy to serve the team more. He does not understand the specific methods and details of execution as well as you do. Therefore, when facing problems at work, you have to package them into a "product" and output the best possible result, changing from question-and-answer questions to multiple-choice questions, and from "I can't do this" to "this matter requires you to decide the direction and provide some resources" to reduce your boss's time cost.

Instead of directly seeking help from your boss, it is better to synchronize the background and problems of the matter and provide several potential solutions for selection to help your boss make decisions and request corresponding resources. While solving problems quickly, this also helps save the boss's time, allows the boss to understand the progress of the project, and encourages the boss to make decisions.

Back to the topic, "A product without BI is not a good operation." Although products, operations, and BI have different focuses in specific work, the requirements for basic product sense, operational skills, and data analysis capabilities should be common. Using experiments and data to drive product iteration and achieve user growth should also be everyone's common goal.

The above are some of my thoughts. While I have systematically sorted them out and written them down, I also hope that they can be helpful to everyone.

Let us encourage each other on the road to growth.