A must-read for designers! Product Data Planning Guide

Product data is usually used to measure product health, help locate and solve problems, perform stratified operations on users, measure product revenue, and explore key points for improving product indicators. This article is a preliminary introduction to the product data system, hoping to help novice product, interaction, and experience designers master product data planning, use data to drive product iteration, and form a systematic cognition.

[[424583]]

Basic concepts of data indicators

1. What are data indicators?

Definition: The quantification of a data is generally obtained by performing some calculation on the field (such as summation or averaging). Based on the original data, it is statistically summarized and processed to form data used to characterize the quality of business activities.

Data indicators usually consist of three parts:

• Dimension refers to the perspective from which products are evaluated. The way of evaluating products from different dimensions is different.
• The calculation method refers to the statistical method. Let’s take some examples from life. We often hear or tease friends. For example, someone asks you if you have a partner. Some people directly say no, some classmates may not have one in the ecological park, or they may not have one today. These are all different statistical methods.
• Data, that is, data + calculation unit. Following the previous example, when introducing new students, we usually ask if they have a partner, and then ask how many they have, which shows that numerical value is also a very important dimension.

Characteristics of good data indicators:

• Comparability: Ability to compare performance across different time periods, user groups, and competing products.
• Simple and understandable: easy to remember and discuss
• Ratio priority: The data that products look at is preferably ratios, not absolute values.
• Change strategy: The most important criterion is to take corresponding measures as the indicators change.

Let me briefly explain the two characteristics of comparability and ratio priority. Ratios are naturally comparable, and the difference from general numerical comparison is that ratios pay more attention to process, quality, etc. Then the most important of these characteristics is that when good indicators change, the product can take corresponding measures to improve.

2. What is the indicator system?

Definition: An orderly combination of a series of indicators. Data indicators are usually scattered, and the indicator system can comprehensively, concisely and clearly reflect the business status.

A more common indicator system divides indicators into three levels. The first-level indicators are product goal-level indicators. The second-level indicators are mainly product strategy-related indicators, which can be understood as the implementation path of the first-level indicators. The third-level indicators are business execution-level indicators. The difference between the second-level indicators and the third-level indicators is that the third-level indicators break down the second-level indicators and can be implemented.

Common indicator types

The classification of indicator types has a certain subjective component. Here are three common ways of classifying indicators.

The first type is divided according to purpose, which can be divided into operation effect category, user experience category, product feature category, promotion channel category, and performance indicator category.

• Operational effect: indicators of value output, product scale, and user activity. For example: revenue, number of active users, and frequency of use.
• User experience: observe product design from the perspective of experience and interaction of each product entrance, setting method, arrangement method, etc. For example, the conversion rate of various links.
• Product features: statistics and data related to product features to promote product improvement. For example, the number of Tencent questionnaires created and answered; the song switching rate of QQ Music.
• Promotion channel category mainly measures channel effectiveness, channel conversion rate, contribution ratio, etc.
• Performance indicators are more technical-oriented, and commonly include concurrency count, interface time consumption, etc.

The second type is divided by links, which can be divided into process indicators and result indicators. If you want to see the quality of products and operational activities, you will generally look at the result indicators directly, but if you need to improve them, you will use process indicators. Some indicators, such as the number of active users, can be either process indicators or result indicators, depending on what your goal is for system construction. Then there are cumulative indicators such as the total number of users and GMV, which are usually used by companies for external publicity and public relations.

The third type, classified by attributes, can be divided into user data category, user behavior category, and product data category.

• User data refers to the basic information about users, such as new users, active users, and retention status at different granularities.
• User behavior records user behavior, visit-related PV, UV, and communication-related forwarding rate. Here we mainly talk about the K factor indicator, also known as the viral coefficient, which is used to measure the recommendation effect and the number of new users brought by each user who initiates the recommendation.

The product data category includes general revenue-related and business-related information.

How to guide business iteration

1. Understand your North Star Metric

The North Star indicator is probably the most commonly heard indicator. It is meant to guide the direction of the company and products like the North Star. When formulating it, it is best to follow the smart principle. By paying attention to the North Star indicator and limited peripheral indicators, it can effectively help business iteration planning.

Characteristics of a good North Star indicator:

• Delivering value, which can reflect the user value and commercial value of the product;
• Easy to understand, no need for too much explanation, and can be used and communicated with each other;
• User activity is related to user behavior and can reflect the user activity of the product;
• Non-lagging, leading, non-lagging indicators, reflecting the current status of product operations;
• Operability can be broken down into more detailed indicators to guide product planning and execution.

Ps: Good indicators do not necessarily mean that the product is developing well. They are more of a monitoring function and should not be used blindly to carry out work. When actually carrying out business, you also need to think about what problems the business will encounter and whether you can use indicators to warn.

2. Choose a North Star around the Life Cycle

Here we recommend a method for selecting the North Star indicator based on the product life cycle. The product life cycle can be divided into four cycles: startup, growth, maturity, and decline. Products usually need to first confirm which stage of their life cycle the product is in, and then further select the appropriate North Star indicator by understanding the focus and common operating strategies of the products in the corresponding stage.

Products in the startup phase are usually based on the business insights of the initiators. It may be a great idea, but is the demand real? We need to verify the hypothesis through an MVP version and pay attention to whether the product can meet the needs of users. The North Star indicator at this stage is usually chosen to represent the retention of target users.

When making specific choices, consider segmenting the user groups and choose the place with the greatest focus as the North Star indicator for the stage. A simple formula can be: overall user retention equals new user retention + old user retention. The difference between the three is that overall user retention is the final result, and new users are more comparative, which is reflected in the fact that new user retention in different time periods can be used to measure whether the recent product iteration has produced a good impact. Old users may show relative stability because they are already accustomed to it. Generally speaking, new users and overall user retention are more suitable as North Star indicators. Another point to consider is the impact of the business form on the granularity of the indicator, which generally depends on user habits. High-frequency businesses may prioritize daily retention, while low-frequency businesses will choose weekly retention or monthly retention.

During the growth stage, the focus of the product is to attract new users, seek growth in size, explore commercial monetization, and pay attention to the growth of product traffic. At this stage, the product activity indicator is usually chosen as the North Star indicator.

There are three common activity indicators: DAU\MAU, activity rate, and online time. Among them, the difference between activity rate and DAU and MAU is that the number of active products evaluates the market size of the product, and the activity rate evaluates the health of the product. The absolute value of the number of users of a product may be very high, but the activity rate is not healthy enough because the overall registered user volume is larger. The selection criteria are based on the market growth space of the product. Products with large growth space usually use indicators such as DAU and MAU. Online time is more suitable for killtime products, such as Douyin, Tencent Video, etc. It should be noted at this stage that activity is monitored as a North Star indicator, not to force users to poke every day, but to provide stable support for payment, referrals, etc.

In the mature stage, the product's market value and user scale have passed the growth period and entered a stable development stage. At this stage, the product usually focuses on increasing user activity, ensuring product retention, and ensuring stable growth in product paid conversions. Therefore, North Star indicators will give priority to payment-related indicators.

There is a basic formula for revenue, which is traffic multiplied by conversion rate multiplied by average order value (Arpu). Based on this basic formula, many variations can be made. When choosing the North Star, we can find points with large room for change and improvement from the formula. For example, the commercialization of a certain music product chooses the number of new members as the North Star indicator. The reason is that the paid penetration rate of the domestic music market is only 5%, while compared with the 50% paid rate of the US music market, there is nearly 10 times room for improvement.

In the decline stage, the product characteristics are that user demand gradually decreases, market competition tends to ease, and products will consider launching new products and services, allowing old products to divert traffic to new products. Common behaviors are to slow down the decline and conduct new user scenario verification, focusing on new incremental markets. The North Star indicator will give priority to indicators related to new scenarios. This is equivalent to starting a new cycle to verify whether new products and new scenarios can meet user needs.

3. Methods of disassembling indicators

This article mainly introduces three ways to decompose indicators, namely, decomposition by business strategy OSM, growth model or relationship between main business participants.

The business strategy OSM model is mainly broken down according to business goals, business strategies, and business metrics. The overall logic is that the goal is what, the business strategy is how, and the business metrics are which.

For example, we broke down the North Star indicator of a certain product. The North Star indicator of a certain product is the number of weekly active products, which is equal to the newly connected products plus the active product retention plus the silent user activation. We are currently accepting the status quo for silent users and have not taken any measures to promote activation, so the big goal is broken down into two small goals, improving new connected products and improving the retention of old active products. Strategically, for Goal 1, we will increase natural traffic and channel expansion through influence. The focus of measurement is the conversion evaluation of official website traffic, from visits, creation, access, and channel quality evaluation, including traffic distribution and conversion rate. In terms of retention of old active products, the main focus is to improve product experience to improve user retention. The main indicators to look at are the continuous tracking of the number of active users and the retention of segmented user dimensions.

The second method is to break it down according to the growth model. You should have heard of the AARRR Pirate Model more often. It mainly includes customer acquisition, activation, retention, revenue, and recommendation. This model was proposed in 2007. Over the years, as the environment has changed, the RARRA model has gradually evolved. The main reason for this change is that in the past, in the era of Internet traffic dividends, the cost of acquiring customers was relatively low, so we would choose to let users come in and then convert them. Now, with the increase in customer acquisition costs, we tend to do a good job of user retention operations first to avoid wasting new costs, and then expand the user plate. The specific model selection is more of a question of suitability. Choose a model that is relatively compatible with your business to carry out work.

For example, the revenue of an online course platform can be broken down into the number of paying users multiplied by the average order value. Paying users can be further broken down into new users plus old users. New and old users can be broken down into traffic multiplied by conversion rate, and the number of old users multiplied by renewal rate. Old users can recommend to others to form a closed loop. The four indicators at the bottom correspond to attracting new users, activation, retention, and revenue. The actual online course business may be more complicated, and each dimension may correspond to several indicators, which need to be flexibly applied in the business.

The third method is to disassemble according to the relationship between the main participants of the business. This method is usually suitable for platform-type products, where the users of the product include both suppliers and C-end users. When disassembling indicators, further disassemble according to OSM or growth model based on the edges or points of the relationship between the two parties.

For example, taxi apps are probably the most common two-sided market examples. When passengers take a taxi, they need to place an order, pay, and evaluate the driver. The driver provides pick-up and drop-off services to passengers, during which they need to pick up and drop off passengers, collect payments, and evaluate passengers. The indicator that reflects the relationship between supply and demand is the number of orders, and the driver side may affect this indicator. The driver side may affect the service time, and the passenger side is the taxi experience, which can be further broken down. The food delivery business is a typical three-sided market, involving merchants, riders, and users, and building relationships around transaction orders/GMV.

Some of the content comes from the company's internal courses and is for reference only.

Reference courses:

• Product Data Operation Planning
• Building a North Star and Excellence Index System - Decision-making
• Data Indicator System
• Growth Hacking

Author: pikarzhan TEG User Research and Experience Design Department Luban Studio Product Planning