As a product mover, there are many friends who do product operation, and I hope that the small partners who are interested in product operation can be helpful to the small partners who are interested in product operation.
Capabilities required for product operation:
User behavior analysis is to analyze the user's behavior on the product and the data behind the behavior, and make product decisions by building user behavior models and user portraits, so as to achieve refined operations and guide business growth. Specifically, it can be summarized as 5w2h, namely:who, when, where, what, why, how, how much.
But beware:
User behavior analysis is not formalized, not analysis for the sake of analysis, even if it is the needs put forward by core users, it must be verified by data, and no one can represent the real user;The needs of users are constantly changing, and we should not rely too much on past experience, which is not reliable, only data is the most reliable.
SoUser behavior analysis should be centered on user experience and data-oriented.
The iteration of my 2024 product features must be upgraded around the user experience, with the aim of retaining old users and attracting new ones. Generally, there are five main levels of products: strategic layer, scope layer, structure layer, framework layer and perception layer.
Strategic Layer:Updates to product positioning. Taking WeChat as an example, at the beginning, WeChat was designed as a simple communication app, but later it gradually had a series of comprehensive functions such as social networking and life, and WeChat positioning was repeatedly updated to meet the changing needs of users.
Range Layer:It is the scope of product functions, which things can be done by the product and which cannot be done. How can the process be forced to force product operations to consider potential conflicts and rough functional design in the product, and determine what can be solved now and what can be solved in the future? After determining the scope, according to the current development stage of the company, concentrate superior resources to meet the most important functional iterations, launch a version first, and then carry out version iterations and continuous optimization, on the premise of ensuring the user experience of the most important functions.
Structural Layer:Summarize and combine functions and information in a structured manner to determine the presentation style in the interaction. For example, Meituan can order takeout, buy medicine, and call a taxi, so how should the proportion of these functions be distributed to optimize the user experience?Meituan didn't have the function of buying medicine and calling a car at the beginning, and the addition of these functions is not simply superimposed, but it is necessary to constantly find better combinations, otherwise the functions are too difficult to find, and the entire user experience will be pulled down.
Frame layer:To put it simply, this layer is to let users intuitively know what functions the product has. For example, the discovery of the WeChat interface is an inductive functional combination, in which the combination of the explicit distribution structure of the secondary inductive combination such as Moments and Scan is at the framework level. The iteration of the framework layer is relatively an extension of the structural layer, but it is more detailed and complex than the structural layer.
Perception Layer:It mainly solves how the product display can meet the user's usage habits and expectations. In addition to visual effects such as UI, the interaction and feedback of function buttons also belong to this layer. The iterative focus of this layer is the user side, how to facilitate perception, how to guide user operation, are all interactions with the user's habits and subconscious. Remember not to bring in too many personal habits.
This is not to say that we need to implement channel delivery specifically, but more about the adaptability analysis and effect tracking of channels and products. How to track the channel, how to measure the effect of the channel, and in turn guide and iterate the next round of delivery. After the channel is delivered, it is mapped to the entire process of the product and the last purchase occurs. Understand the core indicators of channel operation effect measurement, and use data to attribute. In general, attribution models must clearly demonstrate:
Which marketing channels drive conversions?
What is the contribution rate of each channel?
What are the user paths to the conversion?
How to combine different channels to get more value?
For products, what channel users enter the product through is a question that every product must figure out, and channel attribution has become an important method of data analysis. Common attribution models are:Last Interaction Attribution Models, Initial Engagement Attribution Models, Linear Attribution Models, Time Decay Attribution Models, Location-Based Attribution Models, Custom Models, and more. Each model has its own advantages and disadvantages, and the specific use of which model requires a comprehensive judgment of product operation combined with the product itself and business needs, and a systematic understanding of the whole process.
In short, the above three capabilities require you to have a clear understanding of what the "business data" inside the product is, and understand the relationship between them, so that you can have the opportunity to use them well, and you can also stand in the perspective of "data" to have a dialogue with different collaborators and continuously improve the product.
However, most of the operators with 1-5 years of experience in operations and products will encounter the following dilemmas about "data application + data analysis" in their work:
I have read a bunch of materials about data analysis, learned Python SQL and other tools, but I still don't know what data to look at, how to use the data well, and still can't form a clear way about "data analysis" in my mind
In the face of specific business problems, such as how to evaluate the effect of a product function after the launch of an activity, I still don't know how to disassemble, evaluate and verify, I don't know what the complete application process of data analysis is in my work, and I don't know how to build a data indicator monitoring system or evaluate the effect for specific business problems
I am often said to have "low data sensitivity", but I don't know what "data sensitivity" is, and I don't know how to train my data thinking + improve data sensitivity
You may know what data you need to see now, but you won't plan in advance based on the current data** what the future may look like
I want to go further, but I don't know what kind of data application capabilities advanced operations need to have, and what is my current level of data capabilities.
In short, it is the most effective to get rid of the status quo of 80% of miscellaneous work, become a master of 20% in the industry, and cultivate your own internal strength.