This is a purely practical article about DOU+ delivery. The article has more than 6,000 words and is quite brain-burning, so I suggest you save it first.

Have you ever encountered such problems during DOU+ delivery?

1. New account, every video is voted on DOU+, but not a single hit;

2. Carefully produced video content, invested hundreds or thousands of dollars, but did not wait for the expected hit;

3. I heard from the "teachers" that the best time to advertise is when the likes ratio and fans ratio reach a certain value. But it's useless to advertise.

4. After waiting for a hit product, the more you invest, the worse the effect becomes;

5. DOU+ targeted delivery is useless, the system automatically recommends "a cure-all"? ;

6. Live broadcast of DOU+, a lot of investment has been made, but the maximum number of online users is less than 100;

7. The traffic provided by DOU+ is very strange, and none of them are my target users;

…….

The author has been hanging out in various short video exchange groups for many years, and the above DOU+ problems are not isolated cases.

Not only ordinary users, but also many content organizations and agency operation companies often encounter this problem. Especially those institutions that started out with high-quality content have paid too much tuition in this "pit" of DOU+, and finally concluded that "DOU+ is useless."

Is “DOU+” really useless? Of course not.

The biggest reason for not making good use of "DOU+" is that you don’t fully understand "DOU+".

This article will tell you the principle of using DOU+ to heat up short videos, and explain in detail the three major ways to play DOU+:

1. Accelerate the cold start of new accounts through customized delivery ;

2. Things to note when using DOU+ to heat up short video content;

3. How should DOU+ be broadcast live?

What exactly is DOU+?

Let’s first look at an official explanation from TikTok:

▲Image source: Juliangxuetang "DOU+ Mobile Version Product Introduction"

As shown in the above picture, the essence of DOU+ is to purchase traffic for content.

However, unlike information flow advertising, the traffic purchased by DOU+ does not have advertising tags and shares frequency control rules with the natural traffic of the video, so it can be seamlessly integrated into the Douyin recommendation flow. This is the greatest value of DOU+. It can obtain additional traffic from the "natural traffic pool" for video content in a short period of time.

We all know that in the natural traffic pool of Tik Tok, the better the video content performs (playback volume, completion rate, interaction rate, and attention rate), the more exposure opportunities it will get. For example (this is just an example, the data does not have any reference value).

As shown above: Assume that video A is played 10,000 times 20 minutes after it goes online, and the completion rate of the video reaches 20%. Based on the completion rate, the system allocates 4,000 views to this video within the next 20 minutes.

After another 20 minutes, the system will re-examine the user feedback indicators of the video. If the video completion rate drops to 15%, 3,000 playbacks will be allocated for the next 20 minutes; if the video completion rate rises to 25%, 5,000 playbacks will be allocated for the next 20 minutes. That is, more playback volume is allocated to content with better feedback rate.

Every piece of content in Douyin’s natural traffic pool must go through this process (rating - sorting - release) from the time it goes online, and then gradually becomes a hit.

In this process, DOU+ can supplement the playback volume from the natural traffic pool for us.

For this reason, the additional playback volume provided by DOU+ is a double-edged sword.

If the content is excellent, that is, the user feedback indicators are excellent, then DOU+ can help you get more views faster. On the contrary, if the user feedback indicators are not good, the additional playback volume added by DOU+ will accelerate the "death" of the content. This is why, after many friends ended their DOU+ campaign, the number of views decreased, or even reached zero.

At the same time, we mentioned in the previous article that hot spots have a life cycle. In fact, every piece of content on Douyin has a life cycle.

On the one hand, no content can satisfy the tastes of all users. As the number of views increases, more and more users will see it, and user feedback will inevitably show a downward trend. On the other hand, Douyin itself will continue to look for new explosive content and give more traffic to videos with potential.

In other words, as time goes by, the amount of natural playback a video can get will definitely decrease. If video has entered a period of decline, purchasing DOU+ will hardly increase the distribution of natural traffic. This is why chasing DOU+ investments for hot stocks from a few days ago is usually ineffective.

After reading this, I believe everyone should have a clearer understanding of the role that DOU+ can play: DOU+ can obtain additional playback volume for excellent short video works in the early stage of launch (before entering the decline period) to accelerate the growth of the video.

So, how should DOU+ be used in actual combat?

Accelerate account cold start through DOU+

The first function is to define precise users for content through customized delivery and help cold start the account.

Before we begin, let’s talk a little bit about recommendation algorithms.

▲Image source: Zhang Junlin’s Zhihu column “Technical Evolution Trends of Recommendation Systems: From Recall to Sorting to Rearrangement”

Whether it is Tik Tok, Kuaishou, YouTube, or Facebook, in any content platform or even e-commerce platform, the recommendation system has two stages: recall and sorting.

Recall refers to quickly finding a small amount of content that users are potentially interested in from a large amount of content based on some of their characteristics, and then handing it over to the sorting stage. The sorting stage introduces a large number of features and more complex models to provide users with accurate personalized recommendations;

For the platform, the ultimate pursuit is to explore user needs, push accurate content to each user, and thereby improve user retention, user stickiness, content distribution efficiency, etc. Among these, the most basic is to be able to accurately identify users and content.

Imagine that when an account is cold started, the system does not have account characteristics and historical content characteristics, and can only match based on extremely limited information (video subject, video description, etc.). Therefore, it is difficult to accurately locate your target users. If I can help the system locate my target users, will it improve the matching efficiency?

DOU+'s targeted delivery function can play such a role in account cold start.

DOU+ targeted version has two delivery modes:

The first is customized targeted recommendation, where you can choose four feature labels: gender, age, region and interest. Among them, the interest tag is a new option added after the revision, and multiple interests can be selected.

Compared with before the revision, interest tags can greatly improve the accuracy of delivery, but compared with the targeted tags that massive information flow ads can provide, it is still relatively rough overall. Moreover, by using this method, the price of the entire CPM doubles , which means that for the same price, the number of views we get will be half that of other methods.

The second type is recommendations from fans who are similar to experts. As the name suggests, it is recommended to users with similar fan characteristics to the added influencer. The more focused the fan portrait of the target influencer is, the more accurate the delivery will be.

This delivery method is cost-effective, but the process is also more complicated. It requires locking in target influencers through the Douyin Star Map platform or other third-party data platforms, and then returning to DOU+ for delivery.

Taking the Kas Data Platform as an example, you need to go through the following three steps to find the influencer you want who matches your fan profile:

Initial screening: In the expert search function, first select the content track of the account, and then add restrictions to the account in that track from different angles. For example, if you want to create a beauty tutorial account, you can first select the beauty label, and then select an account between the ages of 18-32 and with more than 80% of female fans.

Detailed screening: After the initial screening, we can already see many accounts that meet the requirements. Next, we choose to sort them by the increase in followers in the past 30 days. A large increase in followers in the past 30 days indicates that the influencer is relatively active and has high-quality content. To a certain extent, you can avoid accounts with too many silent fans. Next, we will judge the screening results one by one, focusing on whether the content tone matches yours, especially the content tone of popular videos.

Delivery: After careful screening, you can lock in a group of influencers. Then go back to the DOU+ delivery page and select these accounts for delivery. There are two points to note when placing an ad. First, choose a small amount and a short duration. After the campaign ends, check the settlement report and video surveillance data immediately. If the effect is good, you can continue to invest; if the effect is not good, stop loss in time. Second, try not to choose top or super-top influencers for advertising. The larger the number of fans, the more dispersed the fan characteristics are.

At this point, the entire delivery process is over, and the next step is to look at the data.

We can view video data and user portraits through Douyin’s creator service platform (creator.douyin.com), or we can monitor videos through third-party platforms and view user portraits under a single video.

If the target users are accurately positioned, we can then analyze the specific data and find room for content optimization based on the data.

For vertical accounts, it will be more effective to cooperate with DOU+ for delivery during the cold start process. When the account characteristics are obvious enough, there is no need to use a targeted approach to delivery, and you can directly use the system's intelligent recommendations.

By cold-starting the account through DOU+, our core purpose is not to create a hit video, but to try to ensure that only target users see the video.

How to heat up short video content through DOU+?

We mentioned at the beginning that the essence of DOU+ is to purchase playback volume for video content, and the key to determining whether a video can become a hit lies in user feedback on the video content.

Therefore, from the perspective of playback volume, although the traffic purchased by DOU+ is helpful for the video to a certain extent, the effect is not obvious.

However, from the perspective of feedback rate, if the video content has good feedback, then DOU+ can play a very important role. DOU+ has the ability to allow videos with the potential to become hits to quickly enter the "explosive stage". This is why Douyin officials have always emphasized that DOU+ is a tool for “heating” short video content.

I won’t go into details on how to put DOU+ in videos, as many people have already talked about it. I will just make two additional points.

First, you need to use a "data-driven" mindset to play DOU+.

Data-based means trying to create data reports for each video each time it is released, and finding your own content G-spot.

Don't pin your hopes on any "absolute indicators". Every account has different characteristics, and the content of every track is also different. You cannot replicate other people's data experience. Only by establishing your own "database" can you use data to provide reference for future content creation and DOU+ delivery.

Second, we should use a "refined" operation model to operate DOU+, and avoid being "simple and crude".

Doing DOU+ is a form of advertising, and the purpose of advertising is largely for results. Content is different from commodities. For video DOU+, the likes ratio, comments ratio, and fans ratio are our ultimate assessment goals. Split the budget, do more testing, and only invest when the video has the potential to become a hit. At the same time, we must form a united front with the content team and the operations team. Under appropriate circumstances, it is also necessary for content to give way to results.

In addition , DOU+ is also a good way to check whether the video is compliant. DOU+ review is much stricter than system review. For content that you are very confident in, you might as well give DOU+ a try. At least it can ensure that the content will not have no playback volume due to pitfalls.

Okay, that’s all about video DOU+. Next, let’s focus on live broadcast DOU+.

How to heat up live content through DOU+?

The content formats of live broadcast and video are completely different. Compared with video, live broadcast is small-scale, impromptu, and effect-oriented. It also lasts longer and is more real-time and interactive.

Therefore, whether it is the recommendation algorithm or the delivery logic, there are huge differences compared to short videos.

Let’s first look at the differences between the two types of DOU+ products:

▲ Screenshot of the DOU+ delivery page, screenshot time: May 27

1. The billing methods are different . Video DOU+ is recommended to "a certain number of people" and each person plays it once. It is charged based on CPM (M=Mille, thousand), and is about 20 yuan per thousand plays. Live broadcast DOU+ is based on the expected number of viewers. There is a click behavior between the time a user sees the promotional content and the time they enter the live broadcast room, so it is charged based on CPA (A=action).

Careful friends may find that, with the same amount of investment in live broadcast DOU+, the expected effects of different forms of investment are very different. When I took the screenshot in May, regardless of whether I chose video investment or directly heating up the live broadcast room, the "expected audience for live broadcast" was 25-150 people. In the most recent screenshot, the recommended video heating effect is still 25-150 people, and the direct heating effect of the live broadcast room has become 50-300 people, doubling the effect.

2. The delivery targets are different . The delivery targets of video DOU+ are mainly interaction (likes/comments, visiting account homepage) and attention. The delivery target of DOU+ for live broadcast is mainly based on the behavior of users in the live broadcast room, including: seeding the audience - users click on the shopping cart, view the product details, and place an order; audience interaction - users give rewards, comment and other interactive behaviors; live broadcast room fan increase - users pay attention to the live broadcast room; live broadcast room popularity - the number of viewers and online users in the live broadcast room (user stay time)

3. The exposure time is different . The delivery time of DOU+ targeted version of video can be selected as 2 hours, 6 hours, 12 hours and 24 hours. The online duration of DOU+ live broadcast is 0-24 hours, with an increase unit of 0.5.

4. The delivery materials are different . Video DOU+ can only deliver video content as the main body. In addition to optional video content, live broadcast DOU+ can also choose to directly heat the live broadcast room, that is, deliver synchronous live broadcast content as the main body.

In addition to the above four points, the recommendation logic of the two methods is also different . Video DOU+ will make recommendations based more on the user's interest preferences in the video content, such as: What type of videos are watched more? And there have been likes, comments, reposts, and follows; while DOU+ for live broadcast makes recommendations based on the user's interest preferences in live broadcasts, such as: In which type of live broadcast room do people stay longer? And there have been purchases, gift giving, comments and other behaviors.

Now that we understand these differences, let’s think about how to deliver DOU+ live broadcast.

Usually, people are more concerned about the "viewing rate" when launching DOU+. " Viewing rate", as the name suggests, is the proportion of users who enter the live broadcast room through the delivery of content . This proportion is usually determined by the materials we deliver.

However, in addition to the "viewing rate", the delivery goal you set (conversion goal within the live broadcast room) is also the key to determining the delivery effect.

Taking "planting grass for users" as an example, assuming that in this delivery, you choose "planting grass for users" as the goal, then the system will give priority to displaying your delivery content to users who have been "planted with grass" in the live broadcast room, because they are more likely to achieve the task goals you set.

In actual delivery, whether the users attracted into the live broadcast room are attracted by the product (i.e., the shopping cart opening rate, product viewing rate, etc.) will affect the subsequent exposure times and quality. Because if there are many demands for "seeking for users" at the same time, the system will inevitably give priority to displaying ads with higher conversion rates to the few target users with conversion potential.

Therefore, live broadcast DOU+ has two major content elements:

First, the content delivered must be attractive enough to attract more users who see the content to enter the live broadcast room.

Second, the delivery content, delivery target, and live broadcast content must be highly relevant to ensure that users will convert after entering the live broadcast room.

Now that we know what kind of content to deliver, let’s look at DOU+. DOU+ provides two content formats. The first is short videos, and the second is to directly heat up the live broadcast room.

Short videos are used to heat up the live broadcast room

First of all, we must ensure that the short videos we produce meet the two major content elements mentioned above, especially the second point, which must be highly integrated with the delivery goals and live broadcast content.

For example, if it is a live broadcast with goods, the purpose of the live broadcast is to sell goods, so the content of the short video must be highly relevant to the "goods". If it is a talent show live broadcast, the short video content must be highly relevant to the talent.

Secondly, formulate a video strategy based on your own strengths.

Short videos used to drive traffic to live broadcasts can be divided into two categories: the first is the warm-up category, and the second is the traffic-driving category.

Warm-up short videos are usually released one day before or on the day of the live broadcast. Their purpose is to attract fans to enter the live broadcast room as soon as possible and bring huge traffic to the live broadcast room from the beginning.

It is more suitable for KOLs/celebrity who already have fan effect. Such as Mr. Luo, Tu Zi Ya, Bei Bei Tu, Zhang Ting, Chen He and so on. In addition, it can also be achieved by creating a "gimmick", such as Zhu Guagua. Before the live broadcast, Zhu Guagua will shoot an "official" endorsement video for the main product of the live broadcast. Through the subtext of authenticity and low price, it can not only bring huge traffic to the live broadcast room, but also attract many brand fans to become personal fans.

The purpose of short videos for traffic generation is to drive traffic to the live broadcast room. When users watch the video, they can click on the avatar to enter the live broadcast room. This type of short video is simple and crude. Just take out the most attractive part of the live broadcast room, perform it appropriately, and then shoot it into a short video.

There are more and more such videos on the platform, and they are more suitable for accounts with low numbers of fans, especially those with low numbers of fans that bring goods. When releasing this type of short video, be sure to publish and release it within half an hour to three hours before the live broadcast begins, because short videos usually have peak traffic at this time, and we need to leave some free time to do some interactive guidance for the content. You can also publish a few more short videos to get some traffic from natural recommendations.

However, as more and more videos of this type are produced, the cost of this gameplay will become higher and higher. Because the short video recommendation rules of the default recommendation page are different from the rules of live DOU+short video heating, the former pays more attention to video quality, while the latter pays more attention to budget and conversion effect. For live streaming of products that pursue ROI, FEED stream advertising should be considered in the later stage.

In addition, the short video delivery of live DOU+ can be combined with video DOU+ . On the one hand, video DOU+ is paid per thousand impressions, and the traffic it can obtain is more stable. On the other hand, video DOU+ has a must-see function for fans. Compared with unfamiliar users, fans will have a higher degree of recognition of the anchor, which is conducive to improving various interaction/conversion data in the live broadcast room.

Launch "Direct Heating Live Room"

Directly heating up the live broadcast room means recommending the live broadcast content synchronously on the video recommendation page, which is equivalent to moving the form of "natural live broadcast recommendation" to the default video recommendation page through payment. And the recommendation format is very bold, almost the entire screen can be clicked to enter the live broadcast room.

We have said before that, unlike short video content, live content is longer and more real-time. Therefore, in the recommendation algorithm, the system must not only consider the matching degree between the host characteristics and the user characteristics, but also the real-time characteristics of the live broadcast process. For example: real-time number of viewers, number of gifts/amount, likes, comments, number of new followers, product transaction volume, whether live broadcast/PK/red envelopes are in progress, whether talents are being demonstrated, etc.

For example: Zhang San likes to watch female anchors dancing, but no female anchor can dance for three hours continuously. So for the system, does it need to push beautiful female anchors who are dancing? Moreover, if the female anchor pushed by the system is very good at guiding users to give rewards, then will it not only retain Zhang San but also encourage Zhang San to consume?

Back to DOU+, if we broadcast DOU+ live, we must also choose our own "highlight" moments.

What to do?

First, prepare a detailed live broadcast script and plan at least the content topic for every 0.5 hours of live broadcast.

Then, set core optimization goals for every 0.5 hours of live content (options in Live DOU+) . For example, in the first half hour of broadcasting, the popularity of the live broadcast room can be used as the core optimization goal.

Next, think about the corresponding words or gameplay based on the optimization goals . For example, in order to increase the popularity of the live broadcast room, we can retain users by greeting and interacting with every audience member, and we can also assist through countdown red envelopes and other gameplay.

The last step is to deliver the content. Of course, we choose 0.5 hours for delivery. The amount should be tested based on actual results and ROI . The real-time popularity effect of live streaming is much stronger than that of short videos. Every user who is successfully retained and converted by you will contribute to the weight of your live streaming room.

The direct heating method in the live broadcast room is beneficial to many anchors who started out with show live broadcasts, and it greatly lowers the threshold for such anchors to acquire users. However, these anchors should also be aware that they should never rely too much on this model. Tik Tok is different from traditional live streaming platforms, and making videos is not a bonus but a must.

Finally, I would like to emphasize that when playing Douyin, "content thinking" must be the first element. Whether it is live broadcast or short video, high-quality content is absolutely the core. However, if you want to “play well” on TikTok, in addition to “content thinking”, you must also have “advertising thinking” and “data thinking”.

Advertising thinking: According to the logic of advertising, there is no content that everyone likes. The ultimate goal of all advertisers is to deliver their appeals precisely to the target audience.

Data thinking: Data thinking is a quantitative thinking mode. We should develop the habit of relying on data to discover problems, analyze problems, solve problems, and track problems.

Author: Kas Data

Source: caasdata6