Guangdiantong Advertising Creative Optimization Operation Guide

This article will bring you the specific steps for making popular materials.

1. Template material screening

We click on the creative rankings in the background, then select the video category, and we will enter the "Popular Video Rankings" section. Here we can see the popular material cases that have been running on the platform recently. We can also sort them by multiple dimensions such as exposure, conversion rate, and ratings.

According to my experience, I would choose the two items of exposure in descending order and the last 30 days. The materials screened out by these two items are the most representative, and are also the material templates that are easiest to increase in volume after imitation and production.

This time we are looking for a game app, so we select games and role-playing in the industry options and filter the list in the background. We can also filter by placement. The dimensions currently provided include Tencent News, Tencent Kandian, Tencent Video, QQ and Youlianghui.

After we filter out the materials we want according to the above dimensions, we can click to enter the material page, which contains detailed data and label analysis. A popular material often has many labels. We can choose according to our needs and extract the selling points of the products we promote.

The cover shows users the first-time recharge benefits (what gift packages can be obtained for purchases ranging from 1 yuan to 6 yuan, and what props/rewards are included in the gift packages). It is displayed in the form of a discount, which forms a sharp contrast with the original price and reflects the selling point of the first-time recharge.

After we have screened the case materials and analyzed the selling points of the popular materials, we can send the selected reference videos to the directors, planners and editors to discuss the feasibility of borrowing the selling points of the popular materials, the extension of the selling points and the feasibility of material production. Combined with the suggestions given by the designers, it is very likely that the video we produce will surpass the original version and achieve twice the result with half the effort.

2. Subdividing the material direction

After analyzing the highlights and tags of popular materials, we can match the user's population portrait with the highlights emphasized by the materials from the perspective of user tags, so as to refine the account and plan model.

After determining the direction, what we have to do is to implement the material production, from script writing to editing and output. What we should pay special attention to here is that in the early stage of script writing, the more detailed and specific we write, the better. It is best to be specific to each storyboard, each line of text, and provide reference videos and reference images, etc. The advantage of doing this is that it can reduce the communication costs with creative personnel, and can also produce the materials you want with the highest efficiency, reducing the probability of modifications after production. certainly. If you work well together and provide a reference video, your creativity can create a work that is similar in spirit but not in form, meeting the needs of the launch, which can also greatly improve the efficiency of the early preparations.

3. Material Data Analysis

Once the material is produced, it is time to release it. Here we absolutely cannot just put the video online and ignore it, otherwise all the previous analysis and collision will be in vain. We must conduct data analysis based on the effectiveness of the delivery, analyze your material labels, and the user labels that bring conversions from this material, whether they match and meet expectations, if there are more people than expected, what kind of people are they, and what attracted them to convert, repeat this cycle over and over again, summarize a large amount of data, and then continue to use it in making the next video material.

In addition to CTR and CVR, very important indicators in the game industry are payment and retention. Our ultimate goal is to improve these two indicators. Summarize the key points of your materials that can improve payment and retention, and repeat the same cycle in new materials.

Let’s compare two types of materials that we ran recently. Because of privacy issues, we will not release the cases.

The above is the data of two themes A and B in the past week. A is a high-volume material and its consumption is much higher than B, but the ROI of the assessment is still somewhat different from the expectation. The volume of material B is relatively small, but the ROI can meet the customer's requirements, and the retention data is also very good.

Although the consumption of material B is much less, the back-end payment effect and retention rate are very good. By comparing the effects of the above two materials, we can know that as long as good materials have one of the above two points, they have a lot of room for optimization.

By checking the data, we found that elements labeled as "beautiful graphics" and "cool skills" can gain volume faster. And if we want to increase CTR, we can deliberately set some slots in the video (write typos, set rankings) to attract user comments, so that our ads can enter a higher traffic pool.

Compared with the PC game, the backend payment rate and ROI of the materials are better. These features can attract old PC game players and big spenders, stimulate users' desire to recharge, and can be reused in subsequent material production.

If we want to find points with a high probability of running volume, we need to conduct more detailed research, conduct tests, analyze the data and summarize them. Because we run different products, the points that attract users cannot be the same, and there will be more points to attract users to pay and stay.

We can conduct combined testing from the three dimensions of activation, payment and retention, and through a large amount of data feedback, we can derive the best material labels for each of the above three types of data. Similarly, we can repeat the cycle over and over again.

4. Material label iteration

"Every time I write a script, I look at cases on the Internet, and I will write the good and attractive points into the script. However, most of the materials either cannot be purchased at a high price, or they cost a lot of money but the backend is poor. In the end, I gave up on all the materials." In the early information flow industry, many optimizers did make materials based on their personal feelings, because at that time everyone was actually crossing the river by feeling the stones, and the industry did not have unified standards and mature strategies. There is also a certain traffic bonus. Testing a small amount of material may hit the point of volume, but the data has not been tested a lot, which leads to wrong conclusions. Finally, the wrong video is made with the wrong method, and the result is doomed to failure.

Now, with the competition for purchasing volume becoming increasingly fierce, the traffic pool has become saturated, and everyone is competing on the quality of the materials and the degree of creative innovation. If we don’t have scientific analysis at this time, it will easily fail and waste time and energy.

As the industry matures, many companies are gradually shifting from extensive to refined delivery. Large companies will establish material label libraries specifically for KA customers. In other words, by analyzing all historical materials produced, and analyzing them based on their performance and various data from the front-end and back-end, it will be broken down into separate labels and categories, building an exclusive database, and then output into industry gameplay. When customers with similar products come online, the waste caused by the direction of material output will be greatly reduced.

Through data analysis of large quantities of materials, we break down the key points of each explosive material and conduct variable control testing. This way, we can know the real point of a piece of material that attracts users to convert, and reuse it.

5. Summary

In fact, placing an order is not just a matter of luck, and the characteristics of explosive material can indeed be traced. Today we explained in detail the production process of popular materials. In the entire material production, product analysis, user and material label analysis, material data analysis, material label iteration, etc. are all very important. I hope everyone can follow my suggestions and launch tests together.