3 ways and 1 case study for new operators to quickly develop data thinking

There are many communication problems within the company, including communication between operations and products, and communication between products and technology, but there is no common "language" to solve these problems. However, with the advent of the big data era, we have found that communication between departments through data can ensure that everyone looks at things from the same dimension, abandoning positions and identities, and only using objective data to illustrate the problem. Today, Zhuge Jun shared 3 commonly used data analysis methods and 1 growth plan to help novice operations students quickly build data thinking and improve communication efficiency in daily work.

1. Data Trend Analysis

Generally speaking, trend analysis is suitable for long-term tracking of core product indicators, such as click-through rate , GMV, number of active users, etc. Making a simple data trend chart is not considered trend analysis. Trend analysis requires more clarity on data changes and analysis of the reasons for the changes.

For trend analysis, the best output is the ratio. When analyzing trends, we need to clarify several concepts: month-on-month, year-on-year, and fixed base ratio. Everyone is familiar with the month-on-month and year-on-year comparisons. The fixed base ratio is a comparison with a certain base point. For example, if January 2016 is taken as the base point, the fixed base ratio is a comparison between February 2017 and January 2016. Another core purpose of trend analysis is to explain the trend. For the obvious turning points in the trend line, a reasonable explanation should be given for what happened, whether it is external or internal reasons.

In the process of data analysis, there are many factors that affect the indicators, so we can examine them one by one from different dimensions, such as: channels , product versions, sources, keywords , networks, regions, IP, system browsers and versions, etc.

2. Data Comparison and Analysis

Looking at the trend changes in data independently, in fact, in many cases it cannot explain the problem. For example, if a company's profit grows by 10%, we cannot judge whether this company is good or bad. If other companies in the same industry as this company generally have negative growth, then 5% is a lot. If the average growth of other companies in the industry is 50%, then this is a very bad data. Comparative analysis is to provide a reasonable reference system for isolated data, because isolated data is meaningless .

Taking A/B testing as an example, the most important thing is that the two groups A/B only keep a single variable and keep other conditions consistent. For example: to test the effect of the homepage redesign, it is necessary to keep the quality of the two groups of users A/B the same, keep the online time the same, and the source channels the same, etc. Only in this way can we obtain more convincing data.

3. Data Segmentation Analysis

When some preliminary conclusions are obtained, they need to be further broken down, because some key data details will be obliterated in the use of some comprehensive indicators, and changes in the indicators themselves also require analysis of the reasons for the changes. The segmentation here must be carried out in multiple dimensions. Common splitting methods include:

Time-sharing: whether the data changes in different time periods.

By channel: whether there are changes in traffic or products from different sources.

By user: Are there any differences between newly registered users and old users? Are there any differences between loyal users and novice users?

By region: whether there are changes in data in different regions.

Composition splitting: For example, if the search is composed of search terms, different search terms can be split

Segment analysis is a very important means. Asking more whys is the key to reaching a conclusion, and splitting it step by step is the process of constantly asking why.

4. Business scenario examples

One scenario is to speed up the user decision cycle around the user's key behavior process. Taking e-commerce products as an example, this key behavior process includes: receiving coupons to promote activity, potential customers to promote registration, registration to promote orders, and insights into habits to promote repeat purchases. Speeding up the user decision cycle means helping users recognize value, dispel concerns, and pay for purchases.

1. Holistic approach: Accelerate decision-making in core processes

When we are attracted by the selling point of a product, our subconscious mind will quickly make a decision for us "based on feeling". The reason we don't execute the decision immediately is because we are still looking for reasons to convince ourselves and make the decision seem reasonable.

Human desires are endless. Many times, human desires exceed consistent self-cognition and behavioral norms. At the same time, humans are extremely self-consistent animals. When a decision intention of ours is obviously inconsistent with self-cognition and behavioral norms, we will subconsciously look for more reasons to make our decision "look" reasonable.

For example, you see a lipstick that you like and you feel very excited. Then you have the urge to buy it. However, you already have many lipsticks, and stocking up another one does not conform to your self-awareness of “not wasting too much”, so you start looking for the product details page. When you see information such as “The color of You Who Came From the Stars is no longer popular”, “The same exquisite and silky style as Andy”, “Affordable version of Tom Ford”, etc., your purchase decision can be reasonably explained by these reasons, and you finally choose to buy it.

We need to give users enough reasons to complete the rationalization process, which can be product features, price advantages, or even just a reasonable explanation on a psychological level. These reasons may not need to be particularly prominent, because they are not intended to attract users to generate purchasing impulses. Users with desires will naturally find information to "rationalize their orders" in order to convince and comfort themselves.

Taking a certain e-commerce product as an example, any user who does not create an order within 24 hours after successful registration will receive a text message. The content of the text message is roughly: You have a novice red envelope worth 128 yuan to receive, download the APP to enjoy the discount. A small red envelope will stop some users from "hesitating" and successfully place an order. PS: The data in the above figure comes from our customers. Based on Zhuge io’s precise reach, 68 people successfully completed the payment, that is, this text message brought about the conversion (monetization) of the core business.

2. Business Preferences: Targeted Marketing for Users with Different Needs

After we acquire users, we will analyze their relevant data to understand their behavior habits for use in product operations . By analyzing the data obtained from user behavior monitoring, we can understand the user's behavioral habits in more detail and clearly, so as to find out the problems in the operation, which will help the APP to discover high- conversion rate pages, make marketing more accurate and effective, and improve the user's conversion rate.

For example, during what time period are users most active? If you push activities during this time period, participation will be more positive and users will be more receptive. For example, if analysis shows that women between the ages of 20 and 30 are the most active between 22:00 and 24:00, then at this time you can choose activities that are more valuable to female users for targeted push, set the push activities, time points and user groups, so as to enhance user vitality and extend the user life cycle .

3. Silent activation: intimate reminders around active features

Most of the silent users are low-involvement users. Is it possible to further subdivide their respective needs, and then conduct user reach and user operations to meet their needs at a certain time? If the user is only using basic functions, don't harass him. For example, if you reach out to him for a month and he doesn't respond, and this happens several times in a row, don't harass him again, otherwise he won't use the product. Such long-term silent users should be handled with caution.

In short, trends, comparisons, and segmentation are the most basic thinking in data analysis. Whether it is data verification or data analysis, you need to constantly look for trends, make comparisons, and make segmentations in order to obtain the final effective conclusion.

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

Product promotion services: APP promotion services, information flow advertising, advertising platform