How to conduct user behavior analysis? What is the value?

Many people may not know that the meaning of the four words "behavioral analysis" has represented the degree of sophistication of data analysis needs of some fast-growing companies in the past two years. At the same time, it has also spawned a new generation of data analysis tools and methods. In the past, everyone was paying attention to PV, UV, bounce rate, visit depth, length of stay, and heat maps that many people could not let go of. Generally speaking, these indicators are statistical indicators that reflect the overall situation of the product. The value of data lies not only in reflecting the current situation, but more importantly in its application. Statistics are the result of data aggregation and collation. Without analysis, how can they be used to guide business growth?

Of course, maybe I have been guided. For example, for a reading product, the user's preference for an article can be judged by the length of time the user stays on the article details page. However, the user may be scrolling the screen to read the article or typing a comment in the comment section. Therefore, rather than summarizing the time the user spends on the page, it is more instructive for us to know whether the user is reading the article or commenting and what the comment is about.

Technology is developing and the times are calling, so you need to understand behavioral analysis and know how to apply it.

1. What is User Behavior Analysis?

In a nutshell: Based on the user's behavior on Internet products and the time frequency of the people behind the behavior, the user's usage scenarios can be deeply restored and business growth can be guided.

Slow down: make key additions to the user model. By supplementing behavioral data, a detailed and complete user portrait can be constructed. There is no one behind the data of traditional statistical tools, so there is no user model (portrait). A complete, multi-dimensional, and accurate user portrait = user behavior data + user attribute data

2. The value of user behavior analysis in applications

Three dimensions that influence data results and judgments:

1. Product-market fit (PMF) means that the product meets the market demand. For example, if you have built a community e-commerce platform and launched a demo product to the market, you may not look at the revenue at this stage, but how many people visit, how many people view the product link, and how many people actually pay. At this stage, user recognition may be a more decisive factor in whether you decide to do this project and whether this project has a market.

2. User source refers to user quality. The core population of all products is not for everyone (at least not at the beginning). There must be a core population, so the quality of the user source channel determines the quality of the data results, which may affect your judgment and decision on the project and product; 3. The third point of the product itself is the product. The product carries your business model . The product’s functional design, user experience, iteration, and operational strategy will affect data, retention , and the life and death of the project.

3. The product itself The third point is about the product. The product carries your business model. The product’s functional design, user experience, iteration, and operational strategy will affect data, retention, and the life and death of the project.

Based on the above understanding, we see the importance of behavioral analysis. Based on behavioral analysis, the CEO needs to have a more accurate judgment on the market that the product is entering, marketing personnel need to finely evaluate channel quality, product designers need to accurately evaluate user behavior path conversion, product revision quality, and the impact of a new function on the product, and operations personnel need to do precision marketing and evaluate marketing results.

So, where is the value? List 4 points

1. Custom retention analysis

In the past: We usually consider a user as an active user for the day as long as he opens the app, and a new user added today is considered a retained user as long as he opens the app once tomorrow;

Now: Based on user behavior, we can do refined retention evaluation. Customize user retention based on product features. For example, for a reading product, we can count users who browse or view an article after opening the app as active users today if they have viewed at least one article.

Significance: Retention is a high-level indicator that reflects the health of any product. Regardless of the quality of operations or the design of product functions, retention can ultimately be used to measure the overall situation. Measuring retention in line with the product business itself can help refine the assessment of product health and make retention data more valuable and instructive.

2. Refined channel quality assessment

In the past: In the era of traffic, the evaluation was based on how many visitors and registrations the channels brought.

Now: The demographic dividend period has passed, product promotion channels are increasing, products are becoming more and more vertical, and homogeneous competition has directly led to higher customer acquisition costs. Both from the perspective of market executors and companies are evaluating ROI. In addition to looking at traffic, quality is also important. How to evaluate quality? It needs to be based on user behavior and in line with the business. For a financial product, among the users brought in through different channels, how many of them actually viewed the details of the financial product, how many of them successfully invested , which channel of users had invitation behavior, and which channel had the highest retention? Then, combined with the channel investment, calculate the ROI.

Significance: Reduce channel costs and improve channel conversion.

3. Product analysis (path conversion/funnel analysis, finding key points for product improvement, finding relevant factors that promote core conversion, Aha moment, etc.)

Path conversion/funnel analysis: focus on core conversions, such as registration conversions and purchase conversions, to optimize churn nodes and behavior path design;

Find the key points of product improvement: use data to quantify the core functions of the product, make product iteration scheduling more scientific, and make departmental cooperation more efficient;

Find relevant factors that promote core conversions: dig out key behaviors that promote users to trigger core behaviors, such as certain factors that can promote users to make purchases, and find relevant behaviors for optimization or operation incentives;

Aha moment: Executing quickly and exceeding user expectations to grow the product exponentially.

4. Precision marketing (user grouping, user stratification, activity quality assessment)

In the past: marketing and services were provided to all users or based on user attributes. Attributes include: registered account, mobile phone number, gender, age, region, points and some tags.

Now: attributes + behaviors, infinitely close to real users. Through the user's actions such as clicking to view product details, searching, clicking the follow button for a certain financial product, purchasing something, etc., as well as the person, time, and frequency that triggered the behavior, we can know what the user has been paying attention to recently, what types of products he is interested in, what types of articles he is interested in, and what kind of financial management preferences he has.

Significance: Use behavioral dimensions and attribute dimensions together to identify the user’s role in the product life cycle and real-life scenarios. Marketing is more precise and user experience is better.

Each of the above 4 points can be expanded to talk about value, methods, and cases. The space is limited and there is still a lot of time, so let's talk slowly~

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

The author of this article @朱葛io compiled and published by ( Qinggua Media ). Please indicate the author information and source when reprinting!