微信視頻号推薦算法介紹及運營方式

來源：https://www.chinanovo.net   發(fā)布時間：2024-01-28 17:53:51      

平台算法及規則

Platform algorithms and rules

了解微信視頻号平台的算法和規則，對(duì)創作者十分重要，知己知彼方能(néng)百戰百勝。

Understanding the algorithms and rules of the WeChat video account platform is crucial for creators. Only by knowing oneself and the other can one win a hundred battles.

1. 微信視頻号推薦算法

1. WeChat video account recommendation algorithm

從陌生人社交，到熟人社交。

From socializing with strangers to socializing with acquaintances.

雖說視頻号的目的不全是爲了對(duì)抗“抖快”，但卻免不了與在同一賽道(dào)的他們對(duì)比，而視頻号與抖音、快手大的差别，便是分發(fā)推薦模式。

Although the purpose of the video number is not all to fight against "Shuaikuai", it is inevitable to compare with them on the same track. The big difference between the video number and Tiktok, Kwai is the distribution recommendation mode.

抖音在國(guó)内的發(fā)展勢如破竹，在國(guó)外也沒(méi)有幾個科技巨頭能(néng)幹得過(guò)TikTok的算法。字節打的，是陌生人社交，數據爲王。快手老鐵則向(xiàng)來是社交和算法的平衡好(hǎo)手，“土味文化” 形成(chéng)了快手獨特的半熟人社區。

Tiktok's development in China is overwhelming, and there are few technology giants abroad that can do TikTok's algorithm. ByteDance is socializing with strangers, and data is king. Kwai Laotieze has always been a good balance between social interaction and algorithm, and the "unfashionable culture" has formed a unique semi acquaintance community of Kwai.

而身上流着微信 “社交血脈” 的視頻号，在出生便擁有 “抖快” 都(dōu)不能(néng)企及的牢固熟人社交關系鏈，這(zhè)恰恰是視頻号崛起(qǐ)的大優勢。

The video account, which carries the social bloodline of WeChat, has a strong social network of acquaintances that even Doukuai cannot reach from birth, which is precisely the major advantage of the rise of video accounts.

目前視頻号的推薦，基本上是基于社交關系來進(jìn)行推薦分發(fā)。例如：

At present, the recommendation of video accounts is mainly based on social relationships for recommendation distribution. For example:

發(fā)現頁中的小紅點提醒，是 “好(hǎo)友點贊”、“好(hǎo)友在看的直播” 等；

The small red dot reminders on the discovery page are "like by friends", "live broadcasts watched by friends", etc;

進(jìn)入視頻号，默認打開(kāi)的是“朋友”版塊，而“朋友”也是處于頂部四大信息流的中間位置；

When entering the video account, the "Friends" section is opened by default, and "Friends" is also located in the middle of the top four information streams;

從朋友圈、聊天等途徑看到的視頻号動态，更是純粹的社交推薦。

The video account updates seen through social media platforms such as Moments and Chat are purely social recommendations.

在近曝光的視頻号算法機制中顯示，朋友點贊的“社交推薦”在算法中占比55%，而優質内容的“熱門推薦”隻占15%。

In the algorithm mechanism of recently exposed video accounts, "social recommendations" from friends who like account for 55% of the algorithm, while "popular recommendations" from high-quality content only account for 15%.

這(zhè)就證明：充滿微信特色的社交推薦分發(fā)，依然是視頻号的優先級。

This proves that social recommendation distribution full of WeChat characteristics is still the priority of video accounts.

對(duì)創作者來說，社交推薦也更有利于賬号漲粉引流。用戶對(duì)某個視頻号作品點贊後(hòu)，系統會把這(zhè)一視頻作品推薦到他的社交圈，如果他的微信好(hǎo)友看到後(hòu)也點了贊，那麼(me)這(zhè)個視頻作品會再推薦給此好(hǎo)友的社交圈……

For creators, social recommendations are also more conducive to increasing account followers and attracting traffic. After a user likes a video work on a certain video account, the system will recommend the video work to their social circle. If their WeChat friend also likes it after seeing it, the video work will be recommended to this friend's social circle

這(zhè)種(zhǒng)社交裂變，某種(zhǒng)意義上複制了公衆号在朋友圈 “刷屏” 的傳播路徑，可以無限 “出圈”。

This social fission, in a sense, replicates the official account's "screen brushing" communication path in the circle of friends, which can be "out of the circle" indefinitely.

理論上來說，一條視頻号不停地在各個圈層間傳播，可以潛在觸達的人群邊際是無限的。

In theory, a video account constantly spreads between different circles, and the potential reach of the audience is infinite.

對(duì)運營者來說，首先要意識到，這(zhè)是巨大的流量潛能(néng)。

For operators, the first thing to realize is the enormous potential for traffic.

站在運營實操角度，可以充分利用社交裂變進(jìn)行冷啓動。發(fā)布視頻後(hòu)，點贊、轉發(fā)、評論、收藏是關鍵指标，這(zhè)決定了你的作品能(néng)夠獲得多少曝光量。

From the perspective of operational practice, social fission can be fully utilized for cold start. After publishing a video, liking, forwarding, commenting, and bookmarking are key indicators, which determine how much exposure your work can gain.

很多創作者爲了提升曝光量也想盡了各種(zhǒng)辦法，比如找朋友互贊，加入“互贊群”，付費推廣等等，折騰了半天發(fā)現效果并不明顯，還(hái)有被系統降低權重的風險。

Many creators have tried various methods to increase exposure, such as finding friends to like, joining "like groups", paying for promotion, and so on. After struggling for a while, they found that the effect was not significant, and there was also a risk of being downgraded by the system.

視頻号能(néng)不能(néng)做起(qǐ)來，作爲創作者，還(hái)是要回歸其重要的流量和推薦算法。

Whether a video account can be developed or not, as a creator, we still need to return to its important traffic and recommendation algorithms.

先談談抖音的算法機制：

First, let's talk about the algorithm mechanism of Tiktok:

熟悉抖音的創作者應該都(dōu)知道(dào)，抖音視頻的作品推薦，主要依靠算法邏輯，根據作品完播率、複播率、點贊數、留言數、轉發(fā)數等數據，來綜合判定作品情況，人工審核主要是針對(duì)熱門和爆款類視頻的。

The creators who are familiar with Tiktok should know that the recommendation of Tiktok video works mainly depends on the algorithm logic, according to the completion rate, rebroadcast rate, likes, messages, forwarding number and other data to comprehensively determine the work situation. Manual review is mainly for popular and popular videos.

具體來講，抖音的算法，其實是一個漏鬥機制，跟今日頭條的去中心化的推薦算法原理基本一緻。

Specifically, the Tiktok algorithm is actually a funnel mechanism, which is basically consistent with the principle of the decentralized recommendation algorithm in Headlines today.

那麼(me)，視頻号會不會也采取這(zhè)樣的模式呢？

So, will video accounts also adopt this pattern?

微信不像抖音，視頻号和朋友圈、微信群是打通的，未來也肯定會打通公衆号和小程序。

Unlike Tiktok, WeChat is connected with the video account, the circle of friends and WeChat groups, and will definitely connect with the official account and the applet in the future.

所以，用戶可通過(guò)轉發(fā)來獲得基礎播放量，不需要平台給量。

So, users can obtain basic playback volume through forwarding, without the need for platform feedback.

根據大多數創作者這(zhè)段時間的測試來看，新作品視頻号是不給量的，更新的視頻作品隻有達到一定熱度後(hòu)，才會獲得平台推薦，而這(zhè)樣的基礎播放，需要你通過(guò)轉發(fā)和分享才能(néng)獲得。

According to the testing conducted by most creators during this period, new video accounts are not given in quantity, and updated video works will only receive platform recommendations after reaching a certain level of popularity. Such basic playback requires you to obtain it through forwarding and sharing.

這(zhè)就意味着，在剛開(kāi)始做視頻号時，基礎播放量，要通過(guò)你的轉發(fā)和分享來獲得。所以，視頻号的算法和抖音恰恰相反，采用相反的算法機制，也即是將(jiāng)抖音的 “算法+人工” 變爲 “人工+算法”。

This means that when starting a video account, the basic playback volume needs to be obtained through your forwarding and sharing. Therefore, the algorithm of video number is the opposite of Tiktok, and the opposite algorithm mechanism is used, that is, the "algorithm+manual" of Tiktok is changed to "manual+algorithm".

因此，運營視頻号的創作者不必過(guò)分擔心算法，如公衆号閱讀量一樣，優質的轉發(fā)數據和點贊數，才是評價作品好(hǎo)壞的根本。

Therefore, the creators of operating video accounts need not worry too much about algorithms. Just like the reading amount of official account, high-quality forwarding data and likes are the basis for evaluating works.

2. 微信視頻号審核機制

2. WeChat video account review mechanism

視頻号每天都(dōu)有大量的創作者發(fā)布視頻，肯定是要設定視頻審核過(guò)濾機制的。目前都(dōu)有有雙重審核，分别是機器審核和人工審核。

Video accounts have a large number of creators posting videos every day, so it is necessary to set up a video review and filtering mechanism. At present, there are dual audits, namely machine audits and manual audits.

微信視頻号的審核标準：

Review standards for WeChat video accounts:

機器審核：

Machine audit:

提前設置好(hǎo)人工智能(néng)模型來識别你的視頻畫面(miàn)和關鍵詞，用來審核你的作品和文案是否存在違規行爲。

Set up an artificial intelligence model in advance to recognize your video visuals and keywords, and use it to review your work and copy for any violations.

如果視頻規則可能(néng)會導緻作品被：降低推薦、删除作品、自我可見等。

If video rules are used, it may lead to works being: reduced recommendations, deleted works, self visibility, etc.

嚴重違規則直接删除或者設置自我可見。如果無法直接判定違規則會進(jìn)入人工審核。

Serious violations can be directly deleted or self visible. If the violation cannot be directly determined, it will be subject to manual review.

人工審核：

Manual review:

對(duì)機器篩選出來疑似違規的作品，由人工進(jìn)行審核。人工審核主要有圍繞三塊：視頻标題、視頻封面(miàn)、視頻關鍵幀。

Suspected works of violation will be screened by the machine and manually reviewed. Manual review mainly revolves around three aspects: video title, video cover, and video keyframes.

如果發(fā)确定違規，將(jiāng)會進(jìn)行：删除視頻、降權通告、封禁賬号等處罰。

If it is determined that the violation has occurred, penalties such as video deletion, downgrade notice, and account suspension will be imposed.

上一篇： 5種(zhǒng)網站設計的布局方法
下一篇： 做一些軟件開(kāi)發(fā)原型圖到底起(qǐ)到多大作用？