How to find the rules for short video ads to become popular, 3-step rule!

Is there a scientific method that can quickly discover the patterns of popular video materials, accurately locate the target audience, and place short video advertisements efficiently and at low cost?

This article will share with you how to use data models to drive advertising, and interpret the experience of scientific advertising with practical cases.

Contents of this article:

1. How to find the rules of popular materials

2. How to select the most accurate targeted groups

3. Human-machine integration and automated delivery

1. How to find the rules of popular materials

How to find the patterns of popular materials can be summarized into three steps: disassembling the materials to find the patterns, choosing the right traffic pool, and conducting combination testing .

Step 1: Deconstruct the running volume material and find out the material pattern

First, we will disassemble the entire video material. The idea of ​​disassembly is to break down the video into 5 seconds before and after, and the story summary (5W) in it according to the 5W2T method.

T1, the content of the first 5 seconds of the video, common examples include background images, BGM, popular jokes, etc.

5W, the first is when, which roughly includes three parts: product entry time, download entry time, and video length . This is based on the analysis of past historical delivery materials, and the median data is obtained based on the online timeline (product entry time: 2 seconds; download entry time: 35 seconds; video length: 40-60 seconds).

The second W, where, refers to the scenes that appear in the entire video material, including what elements; the third W, is the relationship between the characters in the video; the fourth W, what, is the scenarios that users may use; the fifth W, why, is how to introduce the product to users when they don’t know that there is such a product that can meet their needs.

The last T, T2, is the last 5 seconds of the video. It is about how to guide users, including some arrow instructions and small interactions and design elements such as the search box at the end.

At this point, we have completed the framework structure of the entire video. T1 determines the CTR of the entire video, and the 5W and T2 in the middle determine the CVR of the video.

Next, I will introduce a case to show you how to use the 5W2T method and give you a specific example.

Let me first briefly introduce the background of this case. This is a calligraphy product, and it is also based on the 5W2T method we just used to disassemble the excellent materials in its release history and obtain the high-quality elements shared by these excellent materials. Based on the data results that the CTR deviated significantly at the time, I optimized the first 5 seconds of the video and replaced them with red envelope material. Everyone is familiar with red envelopes, which can naturally lead to the relevant functions of the product and greatly improve CTR.

In addition, in addition to finding the commonalities of good materials, this model for disassembling materials can also find out what elements are included in bad materials. For templates with high-quality materials, you can change the templates at will to get more materials.

For example, like who, or the example just mentioned, the calligraphy product changed the character relationships in it and adjusted it to use the "bridesmaids and groomsmen picking up the bride" relationship and introduced it into this product.

Step 2: Choose the right traffic pool

In advertising, it is not enough to have the right target audience or the right creative material, but to match the target audience with the creative material in order to achieve better advertising results.

Let me first briefly introduce the rules for traffic pool screening. Taking Toutiao as an example, the creative classification labels and creative keywords have the highest weights in targeting, because the creative classification represents the labeling of users' advertising browsing behavior.

For example, when some people are watching shoe advertisements on Douyin, they interact with this type of shoe advertisements or stay on the site for a long time, then Douyin will continue to push shoe-related advertisements to them. For example, if there is a person who likes clothes more and interacts with clothing advertisements, Douyin will continue to push clothing advertisements to him. For this tag, there is a filtering item in the AD background of Toutiao, which is the creative tag.

Here I will also explain a case to you so that you can better understand it. Let me first talk about the background. I counted the creative classifications and creative keyword settings of all the more than 100 accounts that were investing in a novel product at the time, and rated the accounts based on the secondary retention dimension. S-level accounts had the highest secondary retention, and D-level accounts had the worst secondary retention. Compare the accounts in the two categories and observe the differences in their creative tags.

Let’s first look at the first material, two different levels of creative labels for a book A. The selection of creative keywords for S-level books A is more particular, and keywords with a strong relevance to the book are selected. For example, if my book is "Grave Robbers' Chronicles", then the creative ideas I choose may include keywords such as supernatural, ghosts and gods, and cultivation, which are keywords that have a strong relevance to the book. For D-level, words are related to novels but are relatively general.