How to build a data analysis system suitable for your own community? The authors point out five key steps. Following the previous article, the author of this article continues to explain the next three steps, let's take a look.
Today we continue to talk about the content of community operation data analysis, in the previous article we said that the key to community operation data analysis is to build a set of data analysis system suitable for your own community, which can be roughly divided into five steps: designing user path maps, sorting out key data indicators, data access, building data analysis frameworks and decision support.
In the previous article, we talked about how to design a user path map and how to sort out the key data indicators on key nodes based on this road map. So today, let's talk about the remaining three links of the community operation data system.
Step 3: Data access
Once we've designed a user roadmap and sorted out the key data metrics, we need to determine how we're going to get that data. That is, data access.
This step is actually relatively simple, which is to determine whether your data statistics are manual or use third-party tools, such as WeCompanion Assistant, etc. In general, it's best to use tools that will make the whole operation a little easier. Moreover, the current tools can basically meet the real-time monitoring of all paths by community operations, and the data can also be exported to its own database, whether it is basic data or conversion data, and even user behavior analysis, heat maps, etc. The so-called data analysis, you have to have data first, right?
Okay, so at this point, we already have all the data in the community operation, and we have classified the data according to the user path map and each key node. The next step is to get into the analysis phase. That is, we need to build an analytical framework.
Step 4: Build a data analysis framework
Why build an analytical framework instead of analyzing the data directly? It's because there are so many ways to analyze data. If you use different data analysis methods to analyze a certain data, you may often come to different conclusions. And with the continuous expansion of the size of the community, the data in the database, whether it is the type and quantity, will reach a relatively large scale, at this time, we must learn to find out the key data from these massive data, analyze it, and get the data related to our initial operational indicators.
Therefore, for data analysis methods, you don't necessarily have to master all the analysis methods, you just need to grasp the most important ones and know which link to use which set of analysis methods. Moreover, the benefits of building an analysis framework can improve our work efficiency on the one hand, and on the other hand, it can allow us to develop a good data analysis habit.
So how should the specific analysis framework be built?
First of all, let's still use the user path map, we continue to use the case mentioned in the previous article, the entire user path can be roughly divided into: advertising exposure, official account, customer service WeChat, community and paid order.
For this path, we can choose to use the funnel analysis method to analyze the entire user path.
1. Method 1: Funnel analysis method
It is a set of process-based data analysis, which can scientifically reflect the user's behavior status and the important analysis method of the conversion rate at each stage from the starting point to the end point.
From the very beginning of the user to see our advertising content, to the final payment order, in the whole process, the user through a key node, what is his conversion rate, so that we know, which step of the user conversion rate is low, then we can make adjustments for this link.
2. Method 2: Churn analysis method
The counterpart to the funnel analysis method is user churn analysis. Through this method, we can locate which type of users are churned, mainly in which link, and further analyze why they are churned.
The above two methods are generally an analysis of the entire community operation, from drainage to final payment and ordering, and even the subsequent sharing link.
We can also take different analytical approaches for different operational segments.
3. Method 3: Channel analysis
For example, for the part of advertising exposure in the drainage link, we can use channel analysis methods. For the active part of the community, we can use key behavioral analysis methods. Of course, with this method, you first have to define what kind of behaviors are key, such as replying to keywords in the background of the official account, participating in community topics, checking in at events, and so on.
4. Method 4: ROI Forecasting
To put it bluntly, you need to know how much you plan to spend on the overall marketing budget of the community, how much you plan to complete the sales, and what the overall input-output ratio is. This also makes it easier for us to put a reasonable price on our own community products, including paid communities.
Of course, for some communities, the purpose of operating a community is not to increase sales, but to improve the quality of customer service. Then this piece may also be counted as a brand premium.
So when it comes to ROI forecasting, the key depends on how you position your community at the beginning. Then I won't talk much about it here.
5. Method 5: User stratification
Finally, in the sharing session, I generally like to stratify users and measure user loyalty and satisfaction through the four-quadrant rule. In fact, the analysis method for this part is not necessarily in the sharing stage. Like some growing communities, this method can also be used to analyze users on a regular basis.
That method may not be like the previous methods, I believe most students have understood, so I didn't talk about it. The measurement of user loyalty and satisfaction may not be well understood, so I will briefly talk about it.
When we operate the community, the reason why we need to stratify users, including adding customer service WeChat first, and then pulling the customer service into the group, is actually to put a label on the user, so as to distinguish the proportion of different types of users. So as to design different community content, activities, etc. In order to better meet the needs of users.
Then we can analyze it through the four-quadrant rule, first of all, we can divide users into 4 types according to the strength of satisfaction and loyalty: for example, like the following diagram.
The abscissa represents satisfaction, and the further to the right, the higher the satisfaction; The vertical axis represents user loyalty, and the higher up it is, the more loyal it becomes.
In this way, we divide users into four types. They are: loyal users, demand users, wool users, and low-demand users. (Of course, you can customize the name of this tag)
However, the criteria for determining user differentiation need to be defined in advance.
Like what:
- Loyal users (high satisfaction, high loyalty): will repurchase every month, and successfully recommend friends to buy more than 1 time
- Wool users (low satisfaction, high loyalty): will buy because of the affordable price, will not make product-related recommendations, and the average order value is less than xx yuan
- Demand-oriented users (high satisfaction, low loyalty): strong demand for products, low brand loyalty, and no repurchase within 3 months
- Low-demand users (low satisfaction, low loyalty): Users who have only purchased once or even never made a purchase
We can fill in all this information through a form and then distinguish users. Then adopt different operational strategies in a targeted manner.
For example, for wool users, we can generally use combined discounts, increase the unit price of customers, or invite friends to bargain for promotions to increase activity traffic; For communities with many loyal users, they can recommend some items with high customer unit value to provide more thoughtful services. Wait a minute.
Okay, then the above is the analysis framework built for different links according to the user roadmap. These data analysis methods are more practical.
In fact, at this point, I believe that most people have a clear understanding of the data analysis work of community operation, and the whole idea must be clear. Know what to do, what data you need to master, and what kind of data analysis methods should be used for different links. But the reason why we do data analysis is not for the sake of analysis, all data analysis is ultimately to serve the operational goals of our community. Therefore, the last step in the construction of the data analysis system is to complete the decision support.
Step 5: Decision support
The so-called decision support can be roughly divided into three parts: the formulation of operational strategies, the evaluation of contributions, and the search for growth points.
The ultimate goal of all our data analysis is to be able to get feedback and help solve these problems.
1. The first is the formulation of operational strategy
Through various data analysis, we know what problems exist in our operation process, such as whether our topics can effectively improve user engagement, whether content output can satisfy users, whether community activities can stimulate user desires, and so on. Then AB testing is carried out, and finally a suitable operation strategy is formulated and precipitated. That way, when we scale up our operations, we can use it directly.
2. The second is to evaluate the degree of contribution
Mainly for operators. Through the analysis of data indicators in each link, such as community quality scoring, activity effect evaluation, operation means scoring, transaction amount in the community, etc., the weights are calculated, and the comprehensive score is finally calculated to rate the operator.
I have been in contact with many companies, and when they are doing community operations, the performance appraisal of operators is still relatively simple and rude, that is, the final transaction amount is used as the assessment content of the operation.
However, we all know that operation is actually a system engineering, and the importance and complexity of the process are often more important than the result, and if we ignore this, not only will the community operation not be able to grow, but even the community operators themselves will not be able to get satisfaction from their work, not to mention the improvement of their capabilities.
3. Finally, look for growth points
This part can be regarded as the most important for some community operations that already have a certain scale.
Especially when traffic operation enters the era of retention operation, how to find new growth points from the existing operation model, including user growth and performance growth, is the most important and critical. In addition, it is also the most important challenge for enterprises and operation teams to make reasonable use of the community operation data analysis system we have built and use various analysis methods in the framework to find business breakthroughs.
summary
OK, that's all the data analysis content about community operation, let's summarize it at the end:
The main core of the data analysis work of community operation is to build its own data analysis system. The specific construction process can be roughly divided into five steps, which are to design the user path map, sort out the key nodes and key data indicators on the nodes, data access, build the analysis framework, and finally make a decision support.
The design of the user path can be determined according to the actual operation, and it can also be determined according to the community operation model described in the previous course, such as dividing the entire operation process into three stages: user drainage, community activity, and paid conversion.
At each stage, you can sort out some key nodes, and at the same time, some core data indicators can be listed on the key nodes according to the three types of basic data, conversion data and user behavior.
Next, we need to prepare the corresponding operational tools in advance to complete the data access. In this way, we can use the established analysis framework to conduct targeted analysis of the data according to the obtained data.
The whole data analysis system, we present it with a graph, that's it!
This is more from the perspective of thinking, to explain the data analysis of community operation clearly, as for some specific details involved, it still needs to be adjusted in combination with our initial community positioning and actual operation strategy.
When we say community operation, it belongs to the category of user operation, and it is the closest place to users. And operations itself is not a position where results are the only criterion for judging. Therefore, this requires us operators to master modular tools, arm ourselves, improve professionalism through data, and make every decision better and the whole process more visible through more reasonable auxiliary judgments.
Of course, some specific analysis methods involved in it, such as channel analysis, funnel analysis, etc., are not discussed in detail and specifically because of time. But the reason why I wrote this article is to help you build an overall framework for community operation data analysis. It may not be that practical, but it is an underlying logic, and it is more like a work guide to help us do our work better.
Columnist
π Ye operates, WeChat public account: Pai Ye operation (pyyunying), everyone is a product manager columnist. A post-80s generation who independently claims to be a π master! Share operational dry goods and industry insights from time to time, and look forward to meeting more interesting souls......
This article was originally published by Everyone is a Product Manager and is prohibited from reprinting without permission.
The title image is from Unsplash and is licensed under CC0
The views in this article only represent the author's own, everyone is a product manager, and the platform only provides information storage space services.