Operation-driven: Only those who understand the true meaning of "operation" can be considered as real Internet people

Because of my work, I have been responsible for the work of product manager and operation manager for quite a long time in the past. In this process, I have also had the honor to communicate with many veteran Internet operation experts. Through mutual communication and practice, I have deeply realized that only a person who understands the true meaning of "operation" can be considered a true Internet person.

Operations may be the job in the Internet industry with the most complicated troops, the most trivial work, the most irregular appearance, and the most difficult to learn systematically. Many times, I found that some old operations people with many years of experience also find it difficult to describe how to organize their work more clearly. Sometimes, some product managers are already doing operations but are not aware of it.

Precisely because operations work is tedious and meticulous, it is sometimes easy for front-line operations colleagues to become overwhelmed with various daily situations and spend all their time on so-called "miscellaneous" work. Instead, they lose the ability to summarize and sort out things. As a result, they lack the ability to see what operations are all about from a higher level and are unable to examine operations work from an overall planning perspective.

However, when we are able to step away from the details of operations and look at the entire operations process, I was surprised to find that almost all work on the Internet should be driven by operations. Operations are like glue that connects products, technology, business, market, marketing , etc. together to ensure the smooth operation of each module.

Perhaps, the Internet should be driven by operations to update and iterate products.

Let’s first look at how the operation came about.

Operations are not exclusive to Internet companies; any industry needs operations. For example, you have used China Mobile for so many years, and its professional term is - operator.

Let’s take an interesting example.

Suppose your uncle is a laid-off worker, but he has a good skill - he makes delicious noodles. So my uncle borrowed 50,000 yuan from relatives and friends and prepared to open a small noodle restaurant to make a living. My uncle rented a storefront, asked someone to help with a simple design, painted the walls, and simply decorated the kitchen and dining room. A small noodle restaurant of about 20 square meters opened on a busy street in your hometown. The first thing the uncle needs to consider at this time is where to find the first person to come and eat noodles? So he went to the factory next door and told the foreman that his noodle shop had just opened and that the workers in the factory would get a 20% discount on noodles this month. Then he put up a small blackboard at the door, on which was written that a certain type of noodle was the main dish every day, and that this noodle was limited to 50 portions that day, and each portion was 40% off. Gradually, the business of the small shop picked up, and my uncle began to consider whether to open a bigger shop in the city, so that the two stores could be chained. So my uncle started preparing for a new store.

I wonder if you operations students feel familiar with this example, as if it is exactly the same as what you do every day. Where to acquire users, how to make users active, whether to organize some activities, and after that, it feels like more users can be attracted.

If we break down the word "operation", it is nothing more than "operation + management" . "Operation" is to ensure the organic combination and coherent operation of various functional links; "management" is to ensure a continuous source of customers and ultimately make money.

Why are operations responsible?

From the story of the noodle shop above, we can see that operations are only responsible for traffic .

The essence of the Internet is connection, and among all connections, the users involved are the most critical. Whether it is To C or To B, there is a possibility of subsequent profits only if someone uses it. Baidu connects users with website pages, Alibaba connects users with sellers’ products, Tencent directly connects users with each other, Toutiao connects users with content of interest, and so on. When more and more users participate, various different demands will continue to accumulate, and then the product will have room for expansion and iteration. Just like the Yellow River, although the upper reaches are surging, they are far less wide than the lower reaches. The reason is that wherever the Yellow River passes, various rivers can be seen converging into the Yellow River, and the flow in the lower reaches is far greater than that in the upper reaches.

In my opinion, there are actually only two things in operations: where the traffic comes from and how to keep the traffic . These two issues need to be considered simultaneously at all times. If only one of them is considered, the situation will become dangerous.

Let us give some interesting examples.

In 2014, an app called " Zuji " suddenly became popular. From a later review, the Zuji team at the time did not expect that this app would become well-known because of the effect of a movie filter, so they just treated it as an ordinary function when developing it. However, with the influx of massive users, the technology and functions of the entire product were not perfect. The river channel could not carry such a large amount of traffic, and there was no way to retain users, which eventually led to the product disappearing after being popular for a month.

Another similar case is Faceu, which also had a surge in user base but no subsequent retention ability, which eventually led to the loss of all users. Perhaps because they learned from previous lessons, FaceU developed by the FaceU team was a huge success. The reason is simple. Subsequent event operations , social content operations , and social function operations are all necessary means to ensure traffic retention.

From the perspective of acquiring traffic: channel operation helps to obtain traffic sources, such as QQ groups, WeChat groups , forums, Tieba, partner fan groups, Weibo, etc.; event operation obtains traffic through the channel itself through targeted activities; good content produced by content operation can be directly delivered to appropriate channels to attract new users ; good brand operation can directly form word-of-mouth influence, thereby attracting new users, etc.

If you want to retain users , the use of any function, consumption of content, participation in activities, and solving of any problems are all for the purpose of retaining users. When we define retention, what we actually define is "users actually use our products, or solve some problems through our products." We can call the related operations of content, functions, activities, and user systems involved in the product product operations . Product operation is a comprehensive concept that ensures user retention.

Target-oriented operations

When operations are clearly responsible for traffic, we can clearly understand that each specific operational work is aimed at a specific goal related to traffic .

A traffic goal can be described by breaking down into: what to do, and how much to achieve, at what time . For example, it is expected that the number of users will reach 1 million in 6 months; it is expected that the daily active users will reach 30% in 3 months; it is expected that PV/UV will reach 5 times the current level in 1 month, and so on. To put it in a bigger picture, we expect to achieve a new market coverage rate of 20% in 6 months, and product services to cover 5 new cities in 3 months.

Here I will give a familiar example to see how operations generally break down goals.

Suppose we decide to add 10,000 new followers to the official account , and the time requirement is 30 days. How should such work be broken down?

Before deciding to start the work, it is important to first identify who is responsible.

The person responsible must be a single individual, not a team . If it is a team responsibility, then no one is responsible for the results. As the saying goes, “Every man has his share of responsibility for the fate of his country.” If everyone is responsible, then in fact no one is responsible. Because no one feels that this matter has anything to do with them, no one will really put all their efforts into taking responsibility for this matter. It is difficult to implement work that has a way out. Military orders and deadlines are often the primary productive forces.

Determine the source of users, that is, the channel.

Any place where you can get users is a channel. Generally speaking, user channels can include WeChat groups, QQ groups, forums, Tieba, Weibo, etc. We need to list all the channels we can think of at the beginning. This process requires both daily accumulation and a certain degree of brainstorming.

Describe the objective situation of each channel.

The subject of description is the channel, and the method of description is the degree of association with the target . For example, when we describe a QQ group, we need to understand how many related users there are in the group, how active the group is, whether the group owner is one of us, and so on. The description of these objective situations can help us sort out the importance of each channel more clearly in the subsequent analysis. Look for key issues across channels.

Key issues refer to the problems that need to be solved to achieve the goals . For example, there are about 3,000 people in a certain forum who are our target users , and the key issues to open up this channel include: dealing with small forum owners, ensuring that each content released is relevant to the forum theme, operating the celebrities in the forum, etc. The descriptions of these issues need to be clear enough and must be actionable.

Develop solutions to various key issues.

When we have a good understanding of the objective conditions of each channel and have identified the key issues that need to be addressed, we then need a real brainstorming session to find solutions to the key issues.

ps: Let me add one more thing here. I think brainstorming is meaningful only when everyone has some understanding of the background knowledge, objective conditions, and key issues, and has given some thoughts, and the person in charge has given sufficient thought to the issues that need to be brainstormed. If no one has thought about it and then everyone is brought together to brainstorm, the results will definitely be one-sided. If you think independently for 5-10 hours and have a clear understanding of all aspects of the situation, and then explain the situation to everyone during a brainstorming session, then everyone's discussion is based on your thinking and the brainstorming session will be meaningful. Many companies love brainstorming, but the results are minimal. The reason is that everyone just expresses their opinions, not real brainstorming.

At this time, the discussion about the solution does not need to consider too much the difficulty of implementation, but just collect the methods first. For example, if you need to attract new members in a forum, possible methods include content operation, event operation, KOL operation, etc. What kind of articles to write, what kind of activities to organize, and who are the KOLs, these things need to be determined clearly in advance.

Based on all the information above, evaluate the conversion rate of each channel.

The estimation of conversion rate depends on the objective situation and the conversion path we design .

Evaluating conversion rates depends on our work capabilities, objective channel conversion rate estimates, and conversion cycles. In addition, in order to ensure that the conversion rate will not be lowered by our work itself, we need to design a clear conversion path. The key nodes in the user conversion process need to be clearly defined, including which step is to import the follow-up page, how to describe the official account, how to prevent unfollowing after following, etc. With these evaluations and designs, we can roughly estimate that through the forum mentioned above, if the moderator is on board, then through content operation, we estimate that the conversion rate can reach 50% within a week, which means that 1,500 people can be attracted.

In the design of the conversion path, we need to plan the conversion cycle. For example, if we want to attract new users from Tieba through content operation, then what should be the average reading volume of the articles we need to write, how many articles should we need to write, what should be the approximate interval between them, how long should it take to write the articles, and whether we can execute it effectively. The evaluation of conversion rate must be objective and accurate, and not made at random. If a WeChat group channel can only achieve a 20% conversion rate if the articles have an average reading volume of over 10,000, and our past articles have an average reading volume of only 5,000, then we should objectively evaluate that the conversion rate of this channel should be around 10%.

Make reasonable priorities based on all the information above.

After completing the above disassembly process, the next thing to do is to set the work priority. The basis for our priority setting is "channel volume × conversion difficulty". If a channel has a large volume and is easy to convert, then its priority must be the highest. Those that appear to have a large volume but are difficult to convert should have their priority lowered. Because we are responsible for the target results, the number of fans who ultimately stay is the final KPI. Otherwise, if we write a lot of articles and do a lot of activities but there is no conversion rate, it will be all in vain.

Plan resources in advance and establish a review mechanism.

With priorities and conversion rate assessments for each channel in place, the final step is to develop tactical strategies. If we regard the entire goal as a war, then the release of an article and the execution of an activity can be regarded as a battle . Before going to war, you need to plan in advance which combat teams will be involved, what resources will be needed to support them, how much support will be needed, etc. These things need to be planned clearly with other teammates in advance so that everyone can focus on the target.

It is necessary to establish a review mechanism and set key time nodes based on the review mechanism . For example, the expected time for reaching the 100th fan, when the 1000th fan will be reached, and when 5000 fans will be reached. With these time nodes, the next step is to further refine each task based on these time nodes. The best operations are those that can break down work into daily tasks. Time-driven work, although very hard, has proven to be the most effective.

Predict in advance and keep it under control

Operations is actually a job with routines . Once all these things are planned in advance, all we have to do is try our best to get all the details right. There is a rule of thumb that if 70% of things are under your control, then the possibility of accomplishing it is very high. I remember that when I watched "Nirvana in Fire", what impressed me most was Mei Changsu's plan to rescue Wei Zheng. In order to save people and ensure that no accidents occurred, Mei Changsu rehearsed the entire process over and over again. Every detail needed to be communicated to the executor to ensure what actions to take in any unexpected situation. During this process, Mei Changsu can almost ensure that 99% of the situation is under his control. It is difficult for such a plan to fail. The same goes for operations. Before things start, all objective conditions, execution processes, and user paths need to be deduced. The more detailed the deduction, the more controllable it can be .

Operation drives products to keep users active and continuous

At the end of the article, I would like to tell a little story.

In 2014, I joined the Microsoft XiaoIce team and became one of the first product managers of Microsoft's second-generation XiaoIce. In May of that year, XiaoIce infiltrated WeChat group chats as a ChatBot and took WeChat by storm in three days. More than 100,000 people had direct conversations with XiaoIce until XiaoIce was officially banned by WeChat three days later. A month later, we made a completely new design for XiaoIce and resurrected it on Weibo. This revival was called the second-generation exclusive XiaoIce, and each user was required to adopt his or her own exclusive XiaoIce. This model has continued to this day.

Thanks to the residual heat brought by XiaoIce's popularity on WeChat, it brought a large amount of user traffic when it was revived on Weibo. At one point, more than 100,000 people adopted XiaoIce every day. When everything was going well, we considered a serious question: How can we continuously keep users active ?

We did two things at the time: "persistent in continuous iteration of chat quality" and "released a new feature update every week". The iteration of chat quality requires the participation of all algorithm engineers and relevant product managers, while the weekly update of a new feature requires the participation of all product managers. Looking back today, I think that insisting on weekly updates is the most critical factor in XiaoIce's continued activity to this day .

Our definition of weekly updates at the time was that it had to be a new interactive feature, had to have a strong sense of participation, and had to be released at a fixed time . The weekly update was later scheduled to be released every Wednesday afternoon. This time point gradually became the time point that Xiaobing's die-hard fans would stick to every week. If an update was not released by then, many users would even go to Xiaobing's Weibo to urge her to do so.

Why are weekly updates important? There is only one reason - to attract new and active users by continuously strengthening XiaoIce's attribute positioning . If we regard weekly updates as a project, then we should start thinking about next week’s updates at least every Wednesday night. We later divided the updated functions according to the development difficulty. If the development cycle is short, it will be completed that week. If the development cycle is long, it will be planned at least two weeks in advance. The functions that can be planned for weekly updates are all discussed and communicated repeatedly. Only good products will be included in the weekly update releases, and the others can only be released silently.

Looking back at weekly updates, they have two characteristics. One is that they maintain the continuity of user activity, and the other is that they are operations-driven products . XiaoIce's subsequent products were almost entirely driven by weekly updates, and even some key partner releases were disseminated through weekly updates. It is extremely commendable that an operations job eventually becomes a driving force for the execution of the entire team's product, R&D, and even marketing work.

All this seems incredible, but it is actually reasonable, because only operations are responsible for the goals and only operations are responsible for the results. When the entire team is responsible for the traffic results, operations-driven becomes the first choice, and products and technology cooperate with the rhythm of operations to ensure that users are always active and retained .

Mobile application product promotion service: APP promotion service Qinggua Media advertising

The author of this article @帅帅的帅 is compiled and published by (Qinggua Media). Please indicate the author information and source when reprinting! Site Map