Which data should startup apps monitor?

Many entrepreneurs have no experience in viewing data when monitoring APP data, and they cannot find the key points in the selection of data indicators. For startup teams, which data should be monitored? I would like to recommend a good article to you. Let’s take a look at how this O2O merchant platform product in the early stages of entrepreneurship conducts data monitoring?

The following content is reproduced from PM Faner, the original author is: The Sun and the Moon under Your Feet, and the original text has been slightly edited when reproduced.

ask:

For a startup company's O2O project, a platform product that brings together merchants, which data are more important and must be monitored and analyzed from the first version? Which data is less important and can be put off to later versions? Although there are third parties (such as Umeng) that monitor very comprehensive data, I am not very clear about the priorities. I hope you can give me some advice.

answer:

This case is a good opportunity to summarize some of my experiences in data analysis.
Let me first put forward a point of view: "Data indicators serve the goals"

Several conclusions can be drawn from this perspective:

Corollary 1: Clarifying the goal is the most critical step. How to clarify and set the goal is the first problem that needs to be solved.

Corollary 2: By deconstructing the goal, you can get the corresponding data indicators.

Starting from Corollary 1, the first problem we need to solve is to find the highest priority goal and set accurate goals. What influences the goals of a product?

1. Product life cycle; 2. Product type; 3. Product business model (promotion, profit, etc.); 4. Target user groups; 5. Company goals; 6. Market competition environment, etc.

From the description of the problem, we can get the current characteristics of this product.

1. Product life cycle: Initial stage - exploration;

2. Product type: O2O e-commerce platform;

3. Product business model: Promotion is mainly online and offline; the main profit model is to charge commissions from merchants;

4. Product positioning and user groups: Assume that the product is a platform that provides high-end catering services for white-collar workers, with a toB merchant side and a toC user side. For the sake of analysis, assume that the questioner is responsible for the toc side product.

Assume the following two points:

1. Market competition environment: The market competition is fierce. There are similar competing products, but no giants have emerged. However, leaders have emerged in the same industry.

2. The company’s goal: not to consider profitability within 2 years.

Based on the above factors, we can get the current highest priority goal: rapidly increase users and cover the market. Around this goal, we now begin to deconstruct the goal through multiple factors to obtain key data indicators.

1. Products in the early stages:

Key points: user feedback, meeting needs, and improving user satisfaction. Focus on data indicators: user retention, activity, feedback, user portraits (loyal users, lost users, general users), characteristics and demand analysis;

2. O2O e-commerce platform:

Key point: Product type determines the key service objects of the product. O2O e-commerce is a product with heavy offline business. In the early stages of the product, we must first serve merchants well and accumulate merchant resources. Merchants will only be willing to continue to provide more services if they have orders. Then we need to attract C-end traffic and increase the number of C-end users.

Pay attention to data indicators: merchant users’ order volume, order acceptance rate, average order value, which merchants have the highest order volume, and which types of merchants are the most popular.

3. Product positioning and user groups

Key point: From the product positioning, we can know that it is mainly aimed at white-collar users and provides high-end catering services. This determines that the product structure and design are all high-end services for white-collar workers, and the user type is particularly focused on white-collar users.

Focus on data indicators: Behavioral characteristics of white-collar users: favorite product types, average order value, and characteristics of users with different levels of loyalty.

4. Market competition environment

Key point: Since there are leaders in the market, there are already models to learn from, which means that there are data indicators to refer to. As a result, the cost of trial and error is relatively reduced.
Focus on data indicators: learn from competitors' data standards. (daily activity, next-day retention, average order value, monthly order volume, number of covered merchants), etc.

Data indicators serve the goals of the product. Without goals, the data indicators of the products set will not be reasonable. The most important goal of this product at present is to serve users well and increase the user base rapidly. Therefore, data analysis of users is the most important.

Based on the idea of ​​goal, starting point, and path, let's shift our perspective and look at the funnel model of the e-commerce purchase process, which is often used as a case in data analysis. The user purchase logic of conventional e-commerce is as follows:

“ Browse the product list → View product details → Confirm order → Pay → End”

Assume that in the early stages of the product, there are not many users. We observed key conversion rate data throughout the entire process: list to detail page (80%), details to order (70%), and order to payment (20%). But at this time, our key goal is not the conversion rate of the purchase process, but to add a large number of merchants and have more products to choose from, so we will not waste time on the conversion of the above pages. Because the number of users in the early stage of the product is very small, and there are still many links that need to be done, but the ordering process has at least been run through.

Assume that there is a certain amount of users in the mid-stage of the product. We observed key conversion rate data throughout the entire process: list to detail page (80%), details to order (70%), and order to payment (20%). The key issue at this time is not the conversion rate in the previous stage, but the low conversion rate of the order-to-payment link. Our key goal at this time will be to increase the conversion rate of order-to-payment first. At this time, we will specifically study the relevant data indicators of the order page (page browsing time, user operation behavior), etc.

Assuming that the product is in the mature stage, we observed the key conversion data in the entire process: list to detail page (80%), details to order (70%), order to payment (20%). These data have been adjusted to a more mature state by several former product managers. At this time, we will pay more attention to the detailed data of each stage, constantly make small changes, and perform testing and optimization. At this time, if a certain process can be improved by 1%, it will bring huge benefits to the company.

After establishing the above basic ideas, think about how to improve the ability and sensitivity of data analysis. The first thing you should do every day is to look at the data, deal with the data, and pay attention to anomalies so that you can understand the sensitivity of the data. We often talk about digging deep into data, and behind this is a person's ability to think deeply. I found that this is both simple and complicated. The process is: encounter abnormal data, ask why, make assumptions, deduce, and segment the data. As for segmented data, it means observing data around different dimensions, such as the analysis of users with different loyalty levels, and dimensional analysis based on the purchase process. Different data results can be obtained by exhausting various fine-grained dimensions.