1% in-depth community operation method!

What is a community, what makes a good community, and how can we run a community well.

Next, I will share with you the operation methods of the 1% community based on my 7 years of community experience. I will not talk about theory, but only about practice.

The operator of a community is actually the operator.

Nowadays, most communities are aimed at sales, or to achieve fission as a result. They implement community strategies with a very obvious purpose and result-oriented approach, such as attracting people → reaching a certain number of people → links/activities → red envelopes/bait → letting users in the group attract people to fission or make purchases → advertising group/dead group.

If your community is in the above situation, you need to look further down, because this kind of community is a pathological community. Its ultimate life cycle and results are generally not good. This is generally because it only stays at the task-based community organization rather than being people-centered.

So how can we put people at the core?

Step 1: Analyze user portraits and label them

Professional terms have been used. To put it simply, it means understanding your users. So how do you understand your users? This depends on the source channels of your users.

The sources of community traffic are as follows:

1. The traffic coming in from the public domain (Ali, Tencent, Baidu, Today), etc., is a traffic unit obtained through content output, bait output, activity output, and advertising output. Each public domain has subdivided sections, and the traffic portrait of each section is different.
2. The traffic of the company’s own platforms (mini programs, e-commerce stores, offline stores, WeChat accounts, apps, self-media accounts), etc. The portraits of these users must be analyzed clearly and labeled with factual labels, strategy labels, and algorithm labels accordingly. Labels are an indispensable part of community operations.
3. Import bursty traffic obtained through user fission.

We all know where the traffic and users come from, but how do we create profiles and labels for them?

1) Analyze user portrait data through the combination of DMP and CDP. The DMP platforms on the market are already very mature, so we will not analyze them here. The CDP platform is a mature integrated solution platform for user analysis (behavior analysis, payment analysis, channel traffic analysis), etc. If you don’t like to use other people’s platforms, you can discuss with your technical team how to do user portrait embedding and testing, data analysis, etc.

2) Set up manual consultation paths and automated customer service paths. That is, when traffic or users come in, there must be communication and interaction. Do not create a group directly when they come. You must chat. You can set up a manual FAQ template for the chat. At the same time, you can combine it based on your product business to make an automated path template. Capture user preferences, consumption habits, cultural attributes, etc. through keywords. The automated path can be implemented with social tools.

3) Labels are divided into fact labels, algorithm labels, and strategy labels

• We can use factual tags to mark user notes, and more often we should use user preferences, occupations, residential areas, hometowns, whether they are married, whether they have children, etc. as combined tags and records;
• Algorithm labels and strategy labels are completed at the technical level, but we can use artificial algorithms to implement algorithm labels, that is, his payment behavior, purchase behavior (competitive products or other products), entertainment behavior, etc.
• You may wonder if users will be so happy to follow these steps to obtain information through chatting, questionnaires, interesting interactions, games, etc. That depends on your understanding and grasp of human nature. You can refer to the tricks of pig-slaughtering schemes and tea girls, as well as the gameplay of blind box draws and pyramid points systems. You must combine your business and user portraits to make paths and chat scripts.

Step 2: Create different communities

(1) When we have accumulated a certain number of user tags, when a certain tag appears multiple times, it becomes a symbol of a group , such as: white-collar tag, Shenzhen tag, cat-keeping tag, etc. How many tags are considered a group symbol? We consider 500 users with 100 overlapping tags as a group symbol.

(2) When the group symbol of the label reaches the number we specify, we can form this precise community . So what should we do when we form this community? We must definitely base all community behaviors on the label attributes of this group, such as daily content, group activities, value output, topic creation, fission, etc.

(3) When a group with the same group symbol is formed, it must be based on a hot spot to ignite the emotions of users in the group , and this point of igniting emotions is the MVP of the community. What is the community MVP?

For example: for the otaku group of the second dimension, what they love most are figures and a popular COSPLAY; the community MVP of the shoe otaku is a limited edition shoe. The community MVP can also be spiritual, such as RMB, local spa services, etc. How to set up the acquisition and distribution of MVP? This requires the formulation of game rules for community operations . It can be a tiered design of purchase amount, membership level, distribution head, etc. to complete the acquisition. It must be achievable and usable, and there are some cases for everyone to see (design multiple successful acquisition cases).

The community MVP is liked by most members of the group, wanted by all members, and cannot be refused by all members. The way to obtain the MVP is to advance the game and set up a task-based acquisition method; the community MVP can also be in the form of multiple commodities, and the community MVP can be given with incremental task-based rewards.

(4) Use the theme of the group to operate the group. The theme of the group can be divided into membership group, activity group, fission group, lecture group, exchange group, product promotion group, etc.

We can formulate the operation timeline of the group according to the value of the group:

• For example, the membership group has a relatively long timeline, especially the paid membership group of training courses, which is generally calculated on a monthly, yearly, or quarterly basis; while the activity group, fission group, and lecture group are more often calculated on a daily or weekly basis, and the group will be disbanded when the time is up.
• For example: a group for a flash sale event will pull users into the group for the flash sale, and then disband the group after the event is over (it is not easy to set up a group, and every event must include fission and new user acquisition methods to let old users bring new users, and add customer service WeChat or other WeChat accounts to retain contact).

(5) Through RFM model analysis , users who have not made purchases recently can be grouped together for corresponding promotional activities, or users with low average order value can be grouped together for a high average order value activity;

It is necessary to obtain the user's consumption portrait based on model analysis, create different groups based on this, and use agents and activities to improve the user's funnel conversion. This conversion is the most important link in cultivating user consumption behavior. Use the mentality of getting a bargain to encourage users to consume, and use the method of making decisions in a short time to reduce the user's purchase cost.

(6) Therefore, when we form each group, we must think clearly about the following points:

1. What extends the life cycle of this group, and how to combine its own business with group members to carry out in-depth content development and value output, these are things that members of this group cannot refuse and are all interested in.
2. In terms of matching tags and portraits, we combine the quantification of our business with the precise points of group members to form a marketing form that makes group members feel comfortable for output.
3. Each output node must be indexed. The actions of each node can be divided into funnels, and funnel management must be index-based control. So don’t just output. You must know how to set funnels and indicators, and optimize subsequent outputs through analysis.
4. Each group is a script, and this script can be performed for a certain number of episodes. There must be a complete set of scripts, and the setting of the script is based on the early sample collection → script trial and error → indicator control → path optimization → script optimization → different theme group scripts are different. Don’t use the logic of using one script to cover all situations. This is a high-cost trial and error.
5. When we have a group of users with the same tag attributes, how can we increase the number of members in this group? We must combine the characteristics of the tags in group fission, and do a good job of management and control (set thresholds). The consistency of tags, topicality and vitality are consistent. The path of adding members to the group is worth thinking about. They must be screened by the staff before entering the group and then supervised.
6. Group rules, regular and irregular activities and content, how to develop topics around tags to accurately target users, how to activate group members to open the group every day, week, and month, and how to set up group member advancement.
7. The labels of the group must be localized, such as localized pet-raising groups, localized programmer groups, localized makeup groups, etc. The regional labels must be used as hard labels. Local regional labels + interest labels become a group, local regional labels + group labels become a group, local regional labels + life labels become a group, etc. Localized labels can guide offline face-to-face transactions, and then return to online operations will have very good operational results, not to mention the conversion effect. When you get to this step, you will find that everyone is already friends, not netizens. Strengthening trust will help improve buyer-seller relationships.
8. Remember one key point: accurately meet the needs of group members before creating a group.

Step 3: Lifecycle Management

1. The importance of group member labels and portraits

The conversion rate of groups with precise user portraits is not low. For example, a friend of mine is running a group for wild animal (captive) foodies. There are only about 100 people in the group, but their purchasing power is extraordinary. The average customer spending is as high as 500-3000 yuan. Each transaction has more than 50 orders, with sales once a week, 48 times a year, and annual sales of more than 5 million yuan.

This is just the sales amount of a group. Where are these customers selected from? They are selected from clubs, individual bosses, and financial markets. They are precise users obtained through layers of screening and labeling.

Previously, he also operated localized club groups, localized individual boss exchange groups, localized stock exchange groups, etc. Then, through content testing of wild game, he found that some bosses in these groups were more interested. Then, he conducted offline exchanges and tastings with these interested bosses through localized attributes, and then screened users who had purchased once as the entry threshold, thus establishing one value group after another.

The stickiness and life cycle of this type are very long. Based on the user's tags and hobbies, layer-by-layer screening is carried out to accurately build groups. The user experience, especially the product experience, cannot be ignored in this link. There must be a set of settings that exceed expectations. (Note: Some wild animals can be raised in captivity)

2. Human life cycle management

First of all, it should be made clear that what you manage in group operations is not the group, but the people , so this life cycle revolves around people.

Once we have labeled population segmentation, we can then explore the life cycle based on this labeling (it will take a long time, but you must convince your boss). The coverage of this life cycle management is based on demand, which is not limited to business, but a comprehensive exploration of people's daily life, eating, drinking, playing, time arrangements, emotional needs, etc.

For example, for a group of programmers, we have an anti-hair loss product. So, if we want to localize this group of programmers, we must first clearly explore some of the needs of this labeled group, such as refreshing tea or coffee for overtime work, complaints about superiors and unreasonable demands, common sub-health problems, massage and health care needs on weekends, etc.

Based on these needs, we can dig out topics and content, and then combine them with other products to launch sales. We can tell you what uncertainties unreasonable overtime and late nights without washing your hair will cause your body, especially your hair, and combine them with some marketing outputs (horror marketing, a combination of symptoms). For example, Mr. A next door has a lot of scalp because he works overtime, and he is losing hair. Then he lost his girlfriend, and he made fun of his friends and relatives...

This combination of demand scenarios is multifaceted, and must be created and output with a combination of multiple elements, so that it can be persuasive and achieve the effect of silence being better than words;

If you want to make a comprehensive combination, you can develop a long-term plan, such as: 28-day hair loss and dandruff solution training camp, through long-term activities combined with online and offline supervision methods to combine, by selling products to enhance the service value, thereby increasing the entire conversion cycle and effect, forming a highly sticky active group.

The formula for creation: labeling users + scenario shaping + case output + extraction of other life scenarios + offline meetings (letting the agents speak in person will have a better effect) + long-term labeling content and value creation + optimization of transaction indicators.

3. AB testing and path optimization

In terms of operating users and communities, our business and user portraits are different, so we must adopt different words, paths, interaction methods, marketing content, output content, etc. Each output plan must be subject to low-cost trial and error, following the AB testing optimization method.

AB testing is not unfamiliar in platform testing. It is to determine the orientation of the plan in a better way. This testing logic can also be followed in user operation and community operation. A better way can be selected as the implementation plan for trial. During the trial, continue to observe the actual situation of the indicators to achieve a more optimized path.

Refining formula: extract the number of users with the same attributes → propose 2 or more sets of plans → disassemble the plans to form paths → set corresponding observation indicators in each step of the path → funnel the indicators and observe the conversion of each layer of the funnel → optimize the funnel conversion, indicators, and path results → form a better plan → optimize and improve in the application.

The principle of path setting is to break down the path: add WeChat (speech, add words) → sort out the speech after passing or automatically reply → chat content and interaction method → ​​manual consultation and automated information collection → match the corresponding communication plan → label sorting → match the community of interest → observe the interaction within the group → sort out and locate the role → conduct in-depth optimization and form a personal database

If you just want to form a short group, then you can ignore the above. Attract people → make deals → fission → make deals. This is a path that can’t be simpler, and it is also a universal formula now.

The above is part of the 1% deep community. Next, the author will conduct a detailed analysis and explanation of the operational strategy output, targeted solution output, case discussion, user value mining funnel, etc. in the community. Let’s stop here.

Thank you for watching.

Author: Late night operation enthusiast

Source: Late night operation enthusiast