More than 100 articles in two months on the daily news show, I understand the three secrets of the traffic playing method recommended by the public account

1、 Understand the source structure of public account article traffic

In recent years, with the rise of short video platforms, most of the traffic of the original graphic platform has been stolen; Therefore, many public account owners are complaining about the poor reading rate of the article delivery, especially the majority of 10w+public account owners once lamented that the image and text content is becoming more and more difficult to do.

In this regard, WeChat officials are constantly adjusting the public domain recommendation traffic; Let new public accounts publish articles that can also be recommended by the platform's public domain traffic, thus giving content creators more traffic sources.

To put it simply, from the source of public account article traffic, the recommended traffic is shown in the following figure:

🔺 Article traffic structure: subscription+sharing+search+recommendation traffic

Remember! Once the content platform is involved in the recommendation algorithm mechanism (i.e. crowd+content tagging matching)

The new principle of content creation is to do what users want to see on the premise of pleasing the algorithm.

If the above two points are met, the probability of article reading exceeding 10w+will greatly increase!!

The figure below shows more than 100 articles in more than two months of Qishu's daily trading accounts, and the traffic data of a public account.

Through the early content positioning and strategy, 76.7% of the traffic from the starting number depends on the article to see the recommended traffic (now the system background has been collectively referred to as the recommended traffic), and 12.7% of the traffic comes from the search.

But today, Uncle Seven mainly shares how to get the recommended traffic from the public account, and will share with you from the following three aspects:

1. Recommended traffic algorithm
2. 5 recommended flow inlets ➕ Data optimization direction of each flow inlet
3. How to use WeChat public account articles to recommend traffic on the public account

2、 In depth analysis of public account recommendation traffic 1 The traffic recommendation algorithm understands WeChat recommends the content most suitable for users' interests by analyzing various data such as user interaction behavior (following, reading, liking, watching, searching, appreciating, commenting, forwarding), hobbies, social relations, LBS location, etc.

🔺 Main social recommendation+tag recommendation

Mainly, the public account recommendation flow mechanism includes the following three aspects:

(1) Recommended based on hobbies. Through the identification and classification of user interest tags, the content related to interest is recommended.

(2) Recommend based on reading interest and interactive behavior. According to the user's reading interest and interaction behavior, recommend the content related to his or her previous reading.

(3) Recommendations based on social relationships. Through the analysis of users' social relations, users can recommend the public accounts they follow or the content shared by friends.

Through the above multi-dimensional Data analysis WeChat recommends the content most suitable for users' interests, so as to improve users' reading experience and make the public account get more traffic.

2. Recommended traffic inlet

"Recommended traffic" not only increases WeChat public domain exposure channels for articles, but also extends the continuous exposure and communication cycle of articles, and increases the flow of public accounts and advertising traffic revenue.

However, as the content creator of the public account, you don't know anything about the recommended traffic portal. It can really be said that you will earn n weeks less

Because how much you know about the traffic portal directly determines the direction of your future optimization of content recommendation traffic!!

Entrance 1: Like at the bottom of the article, recommend after reading

The entrance location is more based on your reading interest, interaction behavior, social chain, etc., to recommend relevant content.

When you have done many interactions at the bottom of certain types of articles, the system has marked you with a user tag, and then you will recommend similar articles under the following interactions. In addition, the user tag changes with your subsequent behaviors.

As you can see from this traffic portal, the three core factors that mainly affect the click through rate are: number of readings, title, and home map.

If the number of subscribers to the public account is not large, it is suggested that the optimization focus on the title (accounting for 90% of the layout), followed by the first picture (accounting for 10% of the layout)

Entrance 2: Subscribe to messages

The entrance location is more based on your reading interest and social relationship chain to recommend relevant content.

If you swipe a few more screens of this traffic portal, you will find that there will be a reminder in the lower left corner of the system (the number I follow, shared by friends, read by friends) in two forms: large picture+title, small picture+title.

But the two core factors that mainly affect the click through rate are: title, home image, and the other two potential factors: left subscript hints such as friends' sharing, friends' reading, and the number I follow

However, the seventh uncle found that the articles with the left bottom corner mark are easier to display the large picture+title display, which can greatly improve the click rate of articles.

Entry 3: Discovery page - take a look (watching, hotspot, video tab page)

① Reading the tab page (recommended by friends)

The entrance location is more based on the social relationship chain and reading interest as recommended reading, and the system prompts according to the lower left corner of friends' attention, friends' reading, etc.

As can be seen from the screenshot, "Friends are Watching" has been folded and weakened, so it is not very important to guide users to click "Watching" at the end of the article.

So it is necessary to leverage the social relationship chain traffic of customers who have WeChat in this traffic pool. In addition to the article title and the first picture, there are other potential factors that affect the click rate. The bottom left corner prompts: friends' attention, friends' reading, friends' reading, etc

② Hotspot Tab Page

The entrance location is more based on reading interest, interactive behavior, LBS location, etc. to recommend relevant content.

As can be seen from the screenshot, the two core factors that mainly affect the click through rate of this portal are: title and home image.

In addition, hot traffic portals, original and non original articles can be displayed, but the order of arrangement has no obvious correlation with article related data. (There is no obvious data regularity after the research of Uncle Qi, and even articles with few hundred readers who don't like them can get on this traffic)

Entrance 4: Picture message (more content)

The entrance location is more based on reading interests, social chain relationships, etc. to recommend relevant content.

As can be seen from the screenshot, the two core factors that mainly affect the click through rate of this portal are: title and home image.

Other potential factors affecting click through rate: tips on the lower left corner: friends' attention, my attention number, etc.

At present, picture messages are still supported by official traffic, so sometimes you can send pictures and text messages alternately.

Entrance 5: Collections Tab

This entry location is more recommended based on the same collection tag name.

Research more about the collection tags used by the benchmarking account or the peer size, and you can still use this tag.

With such simple actions, you can get such vertical and accurate traffic. If you don't build a collection label, you can blame the platform for not giving you traffic.

So to sum up, we need to get good public domain recommended traffic on public accounts at each traffic entrance; In addition to content quality, more attention should be paid to title optimization+image optimization+communication of existing users' social chains

The details of these three cores are optimized.

Pay special attention to the picture optimization, which refers to the optimization of the picture on the cover and in the article, including clarity, no messy watermark, relevance, etc.

3、 How to use WeChat public account articles to recommend traffic

In the practical operation of more public accounts and the exchange of experience with peers on the second month of Qishu, I summarized the secrets of public domain recommendation traffic of public accounts in the following articles. It is mainly shared from the account dimension and content dimension:

1. Account dimension

The account dimension is mainly viewed from the following four perspectives:

① Public ID type

There is no obvious correlation. The subscription number or service number may be listed.

② Registration duration

There is no obvious correlation. Both new and old numbers are possible, with the focus on stable update frequency.

③ Certified or not

There is no strong correlation. It is OK to authenticate, and it can also add some account weight. However, the new number operated by Qishu was not certified, and the recommended traffic system was also very good.

④ Violation record

No serious violation record of the account

At present, it is not sure whether there is any account entering the recommendation pool (if you have worked as an e-commerce platform, you can roughly understand what it means to enter the pool). In Uncle Qi's opinion, the account entering the pool cannot be fully believed.

Since Qishu counts the articles published by different accounts, the published article data is not bad from time to time, but he does not recommend the public domain traffic (the content is guaranteed to be free of violations).

However, after all, the algorithm black box of WeChat platform is unknown; In addition to the official announcement, others are not very credible even if they say it out of the mouth of Big V.

As long as you adhere to the output value content of a stable frequency, it is really a matter of time before articles get the public domain recommended traffic of the system.

2. Content dimension

Content dimension involves more important details! Uncle Qi believes that every detail of the public account content operation should be paid attention to, because how much attention you put into it, it can give you the corresponding output!

① Content verticality

Why do we say content verticality?

From the mutual selection of advertising account tags and public account recommendation traffic algorithm, you should realize that now the public account also needs to do the tagging matching interest distribution of accounts, people and content.

(However, according to the current practical experience of Qishu, the current WeChat traffic distribution of tags and interest recommendations is not as accurate as expected, and it is still a long way from the headline system algorithm. After all, the algorithm needs a lot of data modeling to be gradually improved, which is very understandable)

Therefore, when an article is published, it can be accurately recommended to other people with similar tags in the public domain by tagging the account, content and readers through the system.

But note: try to avoid some political articles.

② Update frequency

The stable update frequency of articles is very important!!

If the new account can be better every day, the old one is recommended to write 2-3 articles a week, which can give the system an indicator of the stability of the activity of the account.

If an account that has been connected to the recommended traffic is disconnected for a longer time, the recommended traffic will also drop precipitously, or it may return to 0.

③ Article originality

Although non original articles can also get public domain recommendation traffic, the recommendation proportion is certainly smaller than that of original articles.

Needless to say, it is better to publish articles with original logo!

④ Content keywords+sensitive words

If you understand seo , then you must know that the system grabs the core content tag of an article.

It is mainly based on the information density of the title, subtitle, and article keywords. So it is very important to layout keywords in these areas!!!

Some industry sensitive words are also worth noting, most of which rely on experience.

⑤ Content data indicators

According to the comprehensive analysis of the article data on the public account background made by Uncle Qi and the articles that can be used to recommend traffic, there is no obvious strong correlation with the number of readers, likes, forwards, comments, and appreciations.

However, the system will probably assess the data after the recommended traffic is distributed, because Uncle Qi found that some articles recommend traffic for a long and large duration, while some articles recommend traffic for a short and small duration; Of course, there will be differences in different fields.

So don't spend money to brush meaningless data indicators. If you can brush the data with recommendations, there will be more hot articles, which will affect the user's reading experience. After all, the government has a long head!

After in-depth analysis, Qishu summarized the articles that can recommend traffic on public accounts, and found that the four core indicators are the most critical: opening rate, completion rate, sharing rate, and length of stay. How to optimize these four indicators?

There is room for your own thinking.

Never publish for the purpose of publishing content. Use tactical diligence to cover up strategic laziness.

⑥ Article content quality

When it comes to content quality, Uncle Qi believes that a good piece of content can be described as meaningful, interesting, useful and tasteful in four words. After reading it, users must have a sense of gain;

According to this standard, each time you finish writing, you should review your article again, and feel whether the article can meet one of the standards from the user's perspective.

In addition, the number of words in the article should be kept in a reasonable and comfortable range. It is suggested that the number of words should be more than 500 but not too many, which may affect the reading rate!

⑦ Article publishing setting method

Select the corresponding field for the article category. If you are not clear, select "Other".

The first article and the second article in the group can get the recommended traffic in the public domain under the test of Qishu.

How to determine whether your account has received recommended traffic?

View the total data of recommended traffic in the selected time period (PC side)

Operation steps:

Select the sidebar "Data" - content analysis - data trend - recommendation

View the detailed traffic of recommended articles in the selected time period (PC side)

Operation steps:

Select the sidebar "Data" - content analysis - content channel analysis - recommendation

As long as the recommendation module displays data, it means that your public account article has received system recommendation traffic.

If you ask your seventh uncle if there is any way for new or old people to get the recommended traffic quickly, there must be!

4、 Conclusion

Therefore, if you want to get more content recommendation traffic from public account articles, it is very important to control the recommendation traffic algorithm, recommendation traffic entry, content dimension, and account dimension. For we media practitioners, there are three main sources of traffic on the platform: content recommendation traffic+search traffic+(delivery traffic).

Maybe most of us media people know the content recommendation traffic and search traffic, but the delivery traffic is limited to such things as Tiaojia, French fries, and video number heating.

About how to play with high-level paid delivery traffic, Uncle Seven will share it later.

Understand the traffic mechanism of the three modules of the content platform, and then do the corresponding traffic strategy. In this way, it is only a matter of time before we stick to doing the right thing in the right direction and bear fruit!

We media content in the past of Uncle Qi Marketing I have summarized the following two sentences from my experience to encourage you:

The essence of getting platform traffic is algorithm+content, and the essence of content is understanding human nature;

And content operation is not only about quantity, but also about quality. Both can give consideration to the best.

You are not alone on the way of we media entrepreneurship, when you are typing on the keyboard late at night; Maybe someone is grabbing his head in the lighted study not far from you, thinking about how to write the next sentence.