How to operate the remaining traffic of the content platform?

Content platforms have surplus traffic every day. How can we make good use of this surplus traffic? The author of this article gives corresponding solutions based on his own experience.

Many content platforms have surplus traffic every day. The total amount of this traffic varies greatly. How to manage this surplus traffic? Now I will share with you “What to do with the remaining traffic” based on my work experience.

What is ad remnant traffic?

The so-called residual advertising traffic refers to the remaining advertising traffic after a content platform removes advertising sales and membership benefits. These advertising traffics vary greatly, but can provide advertising support for the platform's content.

Remaining advertising traffic = main site traffic - advertising sales - member benefits

Its changes mainly depend on three factors:

1. As for advertising sales, general content platforms have better sales during holidays, such as Christmas and New Year's Day. This part of traffic can be known in advance in the advertising booking background;
2. Membership benefits. Generally, content platforms have a large number of members when they have membership activities. There are also some trial membership benefits distributed outside the platform, such as three-day trial privileges, membership renewal discounts, etc. The traffic in this part varies greatly, so we can only estimate the impact of membership benefits on the remaining traffic through action effects.
3. The main site traffic depends on the supply of sticky content on the main site, including the quantity and quality of sticky content . For example, for long video platforms, TV series are highly sticky content. If the supply of TV series content weakens, the main site traffic is likely to decline.

How to manage the remaining traffic?

Estimates and strategy adjustments should be made before placing ads for the remaining traffic.

1. Master Station Guide

It means to operate the content within the site strategically and meticulously with the needs and goals of the entire site as the core in order to achieve the goals of the main site.

Estimated remaining traffic : This part can be estimated based on the historical data of the main site, member rights actions and the advertising reservation background to obtain a reasonable predicted value. The unit of this predicted value can be days or weeks. This can be adjusted according to the specific operation strategy. For example, after estimation, a content platform has 10 million remaining traffic on January 6th.

Clarify the needs and purpose of the main site: For the main site, there is a wealth of content of different types and levels. The remaining traffic should be scheduled together with the content slots and PUSH, and make it more in line with the unified goals of the main site: retention, new users and recall.

1. Rating resources for content to be delivered

Generally, when resources are released, the traffic capacity of each content will be re-evaluated, and the content's functions in the three strategies will be scored and rated. The data of each strategic function can be obtained through the data background:

Finally, by integrating various data, we can derive the capability value of a content unit in three dimensions, thus serving as a reference for resource allocation.

Generally speaking, content scheduling has the following characteristics:

• Retention: Focus on the most important sticky content, such as TV series
• Attract new users: Focus on stimulating content that has a greater impact on peak users, such as variety shows
• Recall: Focus on strategically targeting lost users, such as targeting lost ACG users with animation-related content

Generally speaking, the remaining traffic contributes to all three goals of the main site, but has different emphases in different strategic periods, which is mainly reflected in the traffic share of strategic actions.

1. Develop a reasonable traffic distribution strategy

For the remaining traffic, it is not enough to get an estimated value. The most important thing is to distribute the traffic. At this time, it is necessary to carry out combined delivery according to the staged goals of the main site and the rating of the content. For example, the estimated 10 million remaining traffic on January 6 should be distributed as follows under the retention strategy of the main site:

• Since it is Sunday, users have more leisure time and a greater chance of generating traffic peaks. We should focus on content with better new user acquisition capabilities, and give priority to large-category content with stronger new user acquisition capabilities. According to past data, when the investment in such content accounts for x%, the input/output of next-day retention is optimal;
• The next day is Monday, so the retention and digestion of the site traffic the next day is very important. According to previous data, when the investment ratio of this type of content is y%, the input/output of the next-day retention is the best;
• The rest is traffic that needs to be introduced through activities to achieve strong retention within a week.

1. Advertisement scheduling based on traffic strategy and content combination

The content is filtered according to the ranking of its ability value in each dimension. Generally, the top content with good playback volume and newly launched content are filtered out from the TOP5. If the abilities of a certain content are relatively average, it can also be filtered out. In addition, too many content items will lead to more dispersed delivery and inability to collect effective data feedback. Generally, an average release benchmark value for a single content item should be set to control the number of effective deliveries. If it is close to or lower than the benchmark value, the system should automatically stop advertising for subsequent ranked content items.

There are many dimensions of content, and the content evaluation system is also relatively complex. This article only tries to extract proven and effective operation methods from personal experience. I would like to ask colleagues and predecessors for advice on how to improve the empirical methods.

Note: The pictures in this article are not for commercial use and are only used as case references.

Source: Nalan Feifei