4-step method for product data operation

Product managers do a lot of things from 0 to 1, while product operations are constantly adding 0s after this 1, from 1 to 10, 100, and 1 billion. I believe that every product manager hopes that his or her product can grow from 1 to infinity. This article mainly shares the role that data can play in the product operation process and how data empowers operations from the perspective of data operation.

Why is data so important to products?

With the development of the Internet, data-based management has penetrated into the management philosophy of every company, from making decisions based on intuition to using data to speak. I believe that the first thing many product and operations people do after work almost every day is to open the report to check the KPI performance; when reporting to the boss or getting a promotion, or even in an interview, the boss not only wants to hear what you have done, but also wants to hear about the product data you have produced, or what you have done to improve the data.

The birth and development of each product has clear goals. In order to monitor the achievement of these goals, a series of KPIs will be formulated, which will generate a variety of data. It can be said that data is a tool used to measure the development of product business. With it, the understanding and judgment of the business by different roles such as bosses, business, products, and operations will have a unified yardstick.

For example, when two product managers, A and B, reported the value of the Qingming special event project to their boss at the same time, Product Manager A said: The event was very successful, a large number of users participated in the event, and many orders and revenue were generated, so we can increase the subsidy for the event. Product B said: In the exposure stage, 10,000 users were deployed in channel A, which brought 20,000 people to activate and convert. The online CAC (average online customer acquisition cost) was 50 yuan, and the activity ROI (return on investment, revenue/investment) was 1.2. The intensity of the activity can be further increased.

If you were the boss, which product manager's reporting style would you prefer? The answer is obviously the latter. Since data is so important, how should product managers work around data during the product operation phase? Today, I will share with you the four-step method of data operation.

1. Establish an indicator system

In the user growth world, there is a special term "North Star Indicator" or the First Key Indicator Method (OMTM: one metric that matters) when determining the product indicator system. It is the only most critical indicator for measuring business performance in a stage. It guides the product forward like the North Star and can reflect the product manager's pursuit of the core value of the product.

One thing to note here is that, first, within a stage, that is, the product is constantly moving forward, the market is also changing, and the indicators in each stage may be the same, such as focusing on user scale in the initial stage and focusing on revenue in the middle and late stages; second, "the most critical" is the key indicator that can truly measure the health of the business, rather than a vanity indicator. In the early days, Facebook used the number of registered users as a key indicator when defining KPIs. The product design process revolved around the optimization and guidance of the registration process. The operation team used various operational methods to stimulate user registration and gradually discovered that although the number of registered users continued to grow, many users had registered for two or three years, but had never visited again or had lost access and no longer visited. For a social application, users who have been inactive for a long time are of no value. Therefore, using the number of registered users as the North Star indicator will lead to deviations in product and operation strategies. Later, Facebook used the number of active users as the North Star indicator.

After the North Star Indicator is determined, it is necessary to define and break down the indicator system. Because in actual work, multiple teams often work together to achieve the North Star Indicator of the product. This also requires the North Star Indicator to be objective, simple, and easy to understand and break down.

For example, the North Star indicator of a video website is the number of content subscribers. How can it be broken down into execution indicators for each team?

First, let's look at the factors that influence the achievement of this North Star Indicator. From the perspective of the application of subscription attributes, the content supply side largely influences user subscription behavior. To increase the number of content subscription users, it can generally be broken down into three aspects: the first is to increase the number of user subscriptions, the second is to increase the attractiveness of the subscription content, and the third is to extend the user subscription cycle.

These three aspects can actually be further subdivided. Taking increasing the number of content subscribers as an example, we can break it down into three aspects: the first is the activation of new users, the second is the recall of old users, and the third is the conversion of users who experience or use the content.

An analysis like this will determine the work goals for product operations. Those responsible for attracting new users and promotion will know how many new users they need to attract every day, and they must ensure the quality of new users so that they have more user retention. In this way, there are both clear small goals in the short term to guide the implementation of work, and large, global North Star indicators to make the work direction clear and meaningful.

It can be seen that the decomposition of the above process is not a strict quantitative decomposition of KPI from top to bottom, but more about formulating relevant product strategies around the factors that affect the achievement of the North Star indicators.

Another method is to split based on the indicator calculation formulas. For example, based on the relationship between DAU and MAU, DAU=MAU*visit days/30 (number of days in the month) = (newly activated MAU in the month + historical return MAU)*visit days/days = [(newly activated UV in the month*retention rate*visit days) + (historical old user return UV*retention rate*visit days)]*visit days/days.

This disassembly method has been introduced in many articles online, so I will not elaborate on it here. If you need to communicate, you can follow the author and leave a message.

2. Find the right operation method

After the indicator system is broken down, we need to find matching operation methods from different data dimensions.

Taking the number of paid members of Tencent Video as an example, the North Star indicator is the number of paid members. Starting from different data indicators, different operating methods are derived: for example, to increase the number of new users, you have to rely on channel promotion and various user growth methods; and if you want to increase the attractiveness of user subscription content, students in content operations positions need to find ways to provide users with more matching and valuable content through cooperation, screening, theme operations or algorithm recommendations; for example, to extend the user payment cycle, commercial operations are needed, including pricing strategies, guiding member activation strategies, and even event operations, that is, setting up some promotional hot activities within a certain period of time, thereby extending the user's subscription cycle.

In general, there is no absolute standard for dividing operating methods. We often base our divisions on the core goals of the business and break down specific execution data indicators. Then, we identify the focus of current operations based on the business development stage and team characteristics, thereby determining specific operating methods.

At the same time, it should be pointed out that the operational division of labor is also constantly evolving. For example, in the past, attracting new users was more of a channel promotion job. But now, in order to integrate attracting new users, user activation, user retention, and user recall, many companies have introduced the concept of user growth. Around this goal, product planners, product operators, and even technical developers have formed a dedicated small team to independently advance the process.

3. Analyze and improve the completion of data indicators

After the indicators are clear, there will be a series of product iterations or operational activities to achieve the goals. So how to evaluate the quality of the strategy? The two most commonly used data methods are funnel analysis and A/B Testing.

Funnel analysis is to abstract the user behavior path. For example, for e-commerce transaction products, users will go through multiple process links from visiting to placing an order. For example, the order conversion rate in the takeaway product channel is 8%. As a product manager, your KPI is to increase to 12%. Your boss asks you, what are you going to do? How should you answer? Look at the competition?

Using funnel analysis, we split the ordering process into the home page, list page, detail page, bill of lading page, and payment page, and analyzed the user conversion in each link. We found that only 40% of users entered the list page from the home page, and only 30% of users successfully paid from bill of lading. The churn rate in these two links is the highest. What are the possible reasons? Can we start with these two links?

At this point, you can answer your boss's soul-searching question like this: We broke down the conversion of each core node and found that there are two links that performed poorly and need to be improved: one is the traffic distribution from the homepage to the list page, and we think the possible reasons are as follows...; the second is the bill of lading to payment, and we think the possible reasons are as follows...Therefore, we plan to improve the operational results from these aspects. Isn’t this analysis well-reasoned, clear and easy to understand?

Next, let’s look at the second method, A/B Test, which is to verify the optimal solution by comparing different versions and let the data speak for itself.

For example. Carry out an operational activity to distribute red envelopes and attract new users. Users can receive red envelopes ranging from 1 to 10 yuan through the activity page. Users will be guided to withdraw money on the APP at the bottom of the page, thereby achieving the goal of attracting new and active users. As a product manager, you are not sure which copy is more effective. At this time, you can conduct a small traffic test through the system to see which group has a higher click-through rate. The results show that the click-through rate of the "Withdraw Now" button is 25% higher than that of the "Receive Cash" button. This means that the words "Withdraw cash immediately" are more attractive, so you can put more traffic, or even all the traffic, on this "Withdraw cash immediately" activity page.

4. Do a good job in data-oriented summary and optimization

The last step is data summary, review and iterative optimization. The goal of all product and operational work is to achieve business data indicators. After breaking down the data indicators, formulate an execution strategy and look at the results of the data analysis. If the results are good, then think about how to expand the results; if the results are not good, then identify the causes and formulate a new strategy. The entire process of looking at data, finding problems, and positioning optimization is a process of summarizing, reviewing, and iterating optimization.

Summarize

There are four main steps in digital operation:

1. Define and break down data indicators. The North Star indicator is the most critical indicator for a product at a certain stage, and it guides the product forward like the North Star. This indicator needs to be objective, simple, easy for the team to understand, and can be broken down and completed by different teams.

2. The team finds matching operation methods based on different data indicator dimensions to achieve the indicators;

3. During the operation execution process, two very practical data analysis methods, funnel analysis and A/B Test, are used to analyze and test the operation effect;

4. If the operation effect is good, the product team should summarize and review in time to expand the benefits; if the operation effect is not good, they should find the reasons and formulate new strategies.

Author: Data enthusiast

Source: Data Maker