What secrets are hidden in the background data of public accounts that you don’t know about?

Is it true that every time you look at the backend data of a public account , your eyes and mind are in a mess? WeChat officials are providing more and more data, but all you can see are the number of new users, number of people who unfollowed, open rate and forwarding rate.

In fact, there are many secrets hidden in the background data.

User Analysis

User analysis mainly includes two parts: user attributes and user growth .

1. User attributes

The user attributes directly provided by the WeChat backend include gender, language, province, city, terminal, and model.

1) Gender

The gender ratio generally varies by industry. Public accounts related to science and technology and economics have a larger proportion of male users, while public accounts related to emotions and psychology have a larger proportion of female users. Operators can appropriately adjust the layout design and article style based on this data. Some advertisers also need to decide whether to place advertisements based on this data.

2) Province, city

The distribution of provinces and cities is of reference value when initiating offline activities or holding offline training. For example, 24.4% of Sister Chen’s fans are located in Guangdong, of which Guangzhou accounts for 8.8% and Shenzhen accounts for 7.5%. So if Sister Chen considers doing offline activities in the future, they will most likely be in Guangzhou or Shenzhen, because this is the only way to ensure the maximum participation rate.

3) Terminal and model

There are two main aspects of the impact of terminals and models on operations: one is the reward situation, and the other is the display effect. In April 2017, after Apple’s new regulations came out, WeChat removed the reward function for Apple phones. So, if rewards are an important source of income for you, and Apple phones happen to account for a large proportion of user terminals, you can consider using the reward function in a different way.

The most important information that the model can convey is: what size screen most users use to read your content. Each model has a fixed size, such as the iPhone 6 is 4.7 inches, and the iPhone 6 Plus is 5.5 inches. Based on this, we can further analyze the raw data in the background (you can see that 4.7 inches and 5.5 inches are the main models):

Based on the conclusions, when you evaluate the typesetting effect, you can give priority to using models with a large proportion for preview to ensure the best reading experience for most users. For example, the commonly used "indent at both ends" setting, the appropriate indentation is directly related to the size of the mobile phone screen. If most users' mobile phone screens are small (such as the elderly), it is best not to set the indent at both ends.

In addition, terminal models can also convey users' consumption behavior. Because there are large differences in price and launch time between different models, in-depth mining of these data can help understand users' consumption habits to a certain extent, which is valuable for operators and advertisers.

4) User Tags

In addition to the above 6 user attributes directly provided, we can also use the label function in the background of the official account to add special labels to all users who leave messages, reward, and repost, and carry out layered operations. Teacher Li Shaojia’s public account “Shaojiadianban” conducted a one-year statistical analysis: the fan loss rate of fans who gave rewards and positive reviews was about 1% of the overall fan loss rate.

Not only will the churn rate be much lower, but they will also be much more motivated to participate in activities and experience new products. Therefore, we can gather these users together and focus on operations. Some activity-related tweets can be sent specifically to such users (just select the tag type in the group message object).

2. User growth

User growth is undoubtedly the data that public account operators are most concerned about. The analysis of basic data such as the number of new followers and the number of people who unfollowed varies from account to account and will not be the focus here. Sister Chen mainly wants to talk about the easily overlooked data "all sources" in the "new number of people" column.

There are 6 types of sources of new fans that can be counted in the background of the official account, and all other sources are summarized in "other totals". By analyzing the sources of new fans, we can determine the drainage effect of each drainage channel and strengthen the operation of key channels in future operations.

1) Public account search

Most of the fans who come through "public account search" are directed by other platforms, because most of other self-media platforms such as Zhihu, Toutiao, and Jianshu do not allow the direct placement of QR codes. Around September 2016, Zhihu automatically converted QR code images to links:

Toutiao is directly banned. The only way Sister Chen has found to divert traffic from Toutiao is through the watermark on the picture. Although Jianshu did not directly prohibit it, it gave humane suggestions:

"This type of QR code has a very low conversion rate, but it greatly affects the user's reading experience. If users are reading articles on the Jianshu app, the QR code is useless to them. But when users read your articles on the web, few people will take out their phones to scan them. In contrast, we recommend that you upload the QR code to your personal settings, or write your WeChat ID in your profile, so that the people you need will naturally find it."

This suggestion applies to all platforms except WeChat, so after reading it, Sister Chen obediently uploaded the QR code to her personal settings on Jianshu and wrote the ID in her personal profile.

2) Scan the QR code

All places where the official account QR code appears may contribute new fans to this channel, so you need to first figure out where your official account QR code appears. For example, for Sister Chen, the public account QR code mainly appears in two places: one is the website that reprints Sister Chen’s articles (of course, the premise is that the reprint is standardized), and the other is at the bottom of Sister Chen’s articles (when the article is opened after being shared and is scrolled to the end, it may be scanned and followed).

3) Menu in the upper right corner of the image and text page, and the name of the public account in the image and text page

Both of these ways of following are possible only after the article is shared, so this set of data is directly related to the number of articles shared and the quality of the articles. The conversion process of the menu in the upper right corner of the picture and text page is complicated (it takes 2 steps to jump to the public account details page, as shown in red 1→2 in the figure below), and is generally rarely used. It is more common to convert faster through the public account name within the picture and text page (only one step is required to jump to the public account details page, as shown in yellow 1 in the figure below).

If the number of shares is large, but the number of conversions of the public account name in the picture and text page is small, you may need to consider whether the "guide attention" function is not set up or the "guide attention" function is not clear enough.

4) Business card sharing

Increasing followers by sharing business cards means that your official account has been shared and recommended, which is a manifestation of fans' high recognition of the overall content and style of the official account.

Image source: Provided by fans of Strawberry Academy

5) Pay attention after payment

This channel is suitable for certified enterprise accounts and does not require development. When a user successfully pays using WeChat payment methods such as card payment, official account payment, and QR code payment, he/she will be automatically followed by the relevant official account. APP payment requires an additional application. Friends who want to know more can click "Read original text" to view the relevant answers on Zhihu.

6) Other total

According to Sister Chen's observation, this data mainly includes the traffic brought by the business card displayed at the bottom after the article is reprinted by a single hook. Combined with the number of readings at the bottom, it is possible to roughly estimate the diversion effect of the single hook reprint.

In order to make the data more accurate, you can consciously set different traffic-generating logos (QR codes, business cards, public account names, etc.) in different channels. For example, do not put QR codes on Zhihu and Jianshu, but only put the public account name.

Graphical analysis

Graphic analysis mainly includes two parts: single graphic and text and all graphics and text.

1. Single picture and text

1) Conversion rate

Conversion rate is a very important operational indicator for all operations. The open rate and sharing rate directly reflect the operation quality of the official account. However, when judging such data, it is best to compare them within the industry, because the gap between industries is very large and the reference value of data from different industries is relatively small. For example, the average opening rate of the e-commerce industry is only 1.18%, while the average opening rate of the agriculture industry is 8.78%.

Image source: " WeChat Public Accounts' Picture and Text Mass Message Data Report for the First Quarter of 2017" released by the third-party data platform Host

If you really can’t find industry data, you can compare it with the overall performance of the official account:

Image source: "WeChat Public Accounts' Picture and Text Mass Message Data Report for the First Quarter of 2017" released by the third-party data platform Host

2) Read the sources

The reading sources counted in the public account statistics include public account conversations, friend forwarding, Moments , historical messages and others.

Public account conversations refer to the number of times the original text is opened, which reflects the number of your active fans and the attractiveness of the title ; forwarding by friends includes directly sending the original text link to friends, and also includes the number of times it is opened after sharing to the group; historical messages refer to the number of times historical articles are viewed before following the public account, which is related to the number of new fans. If the content is tool -based, existing fans will also continue to read historical articles through historical messages; other items mainly include data on jumping to the historical article page for reading through the menu bar, as well as data on jumping to read through links in the text.

2. All pictures and texts

There is a very important but often overlooked data in all the pictures and texts, which is the "hourly report". This data can help us find the time period with the highest opening rate, so as to decide in what time period to tweet.

When looking at this data, you must consider the interference of the current posting time on the data. For example, Sister Chen usually posts around 10 o'clock in the evening. Because of the system prompts, there will definitely be a reading peak during this period. The peak will continue the next morning, but this data does not mean that 10 o'clock in the evening is Sister Chen's best posting time.

Reading curve of most of Chen Jie’s tweets on the second day

In order to eliminate interference, Sister Chen chose to observe the data two days after the posting. The following is the data of Strawberry Academy two days after a posting in November. It can be seen that the reading peak is from 9 to 11 in the morning. Of course, in order to reduce errors, we can collect as much relevant data as possible and take the average value for estimation.

summary

There is no summary here, every piece of data is important!

The author of this article @琛姐草莓学堂 is compiled and published by (青瓜传媒). Please indicate the author information and source when reprinting!

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