APP advertising growth: dynamic product advertising

Dynamic product ads can be precisely delivered based on user preferences, thereby pushing content that users may be interested in, encouraging users to place orders, and driving growth. So, how does Dynamic Product Ads (DPA) work? What changes has the emergence of DPA brought to advertisers and advertising platforms? The author of this article summarizes DPA delivery, let’s take a look.

The commercialization of advertising is perhaps the most important evolution in the history of Internet advertising. It has evolved from the previous "users looking for products" to the current "products looking for users", providing users with high-quality advertising experience while improving the effectiveness of advertising. All this is thanks to today's topic: dynamic product advertising.

1. Introduction

Before formally introducing dynamic product ads, let’s take a look at a story that everyone often encounters in daily life:

The story above is an example of precision marketing during the advertising process: Douyin APP recommends refrigerator product information to Xiaobao based on his record of browsing refrigerators on Taobao APP. The precise and high-quality repeated information push evokes Xiaobao’s memory, triggers his desire to buy, and ultimately successfully promotes conversion.

In this conversion process, Xiaobao's browsing behavior + the product information of the refrigerator played a huge role. By pushing the refrigerator advertisement to Xiaobao who has potential demand, the crowd is accurately targeted and order conversion is promoted. This is product advertising.

2. Dynamic Product Advertising

Dynamic Product Ads (DPA) dynamically displays the most suitable product content to users based on the user's interest behavior data on the advertiser side and the product library information provided by the advertiser, thereby improving conversion effects.

DPA's vision: to push more relevant product content to users, not just advertisements.

Since it is a dynamic product advertisement, what does “dynamic” mean?

1) Dynamic content

The so-called dynamic content refers to pushing suitable content to users based on their different behavioral data. For example, for different users, if two users have the same intention to buy a refrigerator, Xiao Bao likes double-door refrigerators and Xiao Xu likes single-door refrigerators, then at the same time, the product advertisements they see are different; even for the same person, if his recent online shopping changes, the content of the advertised products he sees will also change accordingly.

2) Dynamic landing page

The so-called dynamic landing page means that the landing page that users see is no longer the same. It can be a product details page, an aggregation page of similar products, or even a bundled sales page of different products. These are all dynamic landing pages generated programmatically based on user interests and product libraries.

3) Dynamic users

The so-called dynamic users mean that the audience of advertisements is no longer defined in advance. Daily product visitors are dynamic audiences that change continuously.

For dynamic product ads, the data comes from users. It does not feel like a marketing ad to users, but more like an off-site personalized recommendation that understands user needs. Therefore, it will get more attention and clicks, and have a better conversion effect.

Since dynamic product ads are so good, advertisers are already eager to try them out and must be wondering whether their ads can be placed on DPA? It should be noted here that although DPA has good effects, its threshold for placement is relatively high and is not suitable for all advertisers. The following conditions must be met:

• Product library: This refers to commodities in a broad sense. First, there are many categories. Second, it has obvious structural characteristics and can describe commodity information through structured language. Therefore, it is more suitable for retail, tourism, hotels, etc.
• User behavior data: Advertisers need to collect user behavior data on the client and calculate the user's intended products. At the same time, they need to use technical docking to allow the advertising platform to perceive the user's intended data.
• Number of users: The number of users of the advertiser's APP must reach a certain level, so that when matching with the media of the advertising platform, it can cover more users and maximize the effectiveness of DPA.

3. DPA delivery process

I believe that advertisers who meet the above conditions are eager to learn about the DPA delivery process. In this section, we will take a look at how a complete DPA advertising delivery process works and the supporting facilities that advertisers need to build.

DPA deployment requires a whitelist, so before the formal connection between production and research, the business side will first perform the whitelist operation. After the whitelist is activated, the real DPA delivery docking will begin.

1. Step 1 Product Library

Advertisers need to synchronize their product data to the product library of the advertising platform and maintain product information on the platform. There are three main ways to synchronize:

1. Manual entry: Applicable to situations where the number of products is small and the product information is rarely updated. There are relatively few usage scenarios for this situation.
2. File upload: Applicable to situations where the quantity of goods is average and the product information is not updated frequently. If advertisers need to update product information, they can do so by uploading an XML file on the platform.
3. Scheduled pull: Applicable to situations where there are a large number of products and product information is updated frequently. By configuring the data source address and account password, the platform can synchronize the latest product information regularly.

Advertisers can dynamically choose the appropriate method based on the updates of their products. It is recommended to use the scheduled pull method to synchronize the product library information as soon as possible.

Updates to the product library are generally full updates. As far as I know, the current mainstream platforms do not support incremental updates (if there is any mistake, please correct me). When the product information is updated, there will be an update protection, which is mainly to prevent the product from being accidentally deleted. It is an automatic protection mechanism provided by the system. After enabling update protection, if the difference between the product before and after the product file is updated exceeds a certain ratio, the system will automatically stop the update and retain the product file from the last successful update.

2. Step 2 Behavior data synchronization

Users will generate various behavioral data on the advertiser side, such as searching, clicking, collecting, placing orders, etc. Advertisers can calculate the user's intended product based on these behavioral data. If the user does not complete the final conversion on the end, DPA can be used to deliver intended products to these users on the media of the advertising platform to recall the users and promote conversion.

For user behavior data synchronization, advertising platforms generally provide the following docking channels:

1) SDK

By accessing the SDK of the advertising platform, the SDK can automatically collect user behavior data, and the user's intended product will also be calculated by the advertising platform. Advertisers only need to access the SDK interface, and the development workload is relatively small. However, the problem with this solution is that advertisers are fully dependent on the advertising platform to participate in the calculation logic of the user's intended product.

2) MAPI

After the advertiser's BI team calculates the user's intended products, it transmits the user behavior data back to the advertising platform through the advertising platform's MAPI. This requires the advertiser to set up a data team to analyze the user's behavior data, and at the same time connect to the advertising platform's behavior data MAPI (not DMP MAPI), which requires a large amount of development.

3) RTA

For the first two solutions, advertisers need to return user behavior data to the advertising platform, which involves user privacy and security issues. Therefore, the advertising platform launched the RTA interface to solve this problem. For more information about RTA, you can refer to the relevant content of "APP Delivery Growth: RTA".

3. Step 3 Advertising

The product library and user behavior data are ready. The next thing to do is to associate the product library data with user behavior by creating an ad.

The biggest difference between DPA advertising and traditional advertising is convenience: for the delivery of massive amounts of goods, traditional advertising may require the creation of a large number of plans, while DPA advertising only requires the creation of one plan, which greatly improves the efficiency of advertising.

When placing advertisements, there is another factor that will affect the effectiveness of the advertisements, and that is the creative template. Since the product library contains main product images of different sizes, it is necessary to apply a creative template to unify the creative size and integrate other product information into the creative, such as product location, price and other information. By disclosing more information in the creative, the conversion rate of the product can be improved.

Advertising platforms will provide creative template production and editing tools. Advertisers can upload PSD files themselves or use the creative templates provided by the platform.

Regarding template splicing, Alimama has proposed a technology for dynamically describing advertising creativity. Those who are interested can refer to "Farewell to Splicing Templates - Alimama Dynamically Describes Advertising Creativity". It is hoped that all major advertising platforms will be able to provide similar technologies in the future to truly serve advertisers.

4. Step 4 Advertisement display

The display of advertisements is entirely the capability of the platform. The platform will combine the user behavior data provided by advertisers and its own understanding of users to display the most suitable product advertisements to users.

SDPA

There is another derivative form of dynamic product advertising, SDPA (Single Dynamic Product Ad), also known as single product advertising.

If DPA is used by advertisers to attract customers and achieve personalized delivery for each individual (finding suitable products for each person), then SDPA is mainly used by advertisers to deliver popular products (finding people who understand high-quality products).

Compared with DPA, SDPA not only has different starting points, but also uses different data models.

• DAP: Combines user behavior data uploaded by advertisers with the advertising platform’s understanding of users to push the most suitable product advertisements to users.
• SDPA: Combining the advertiser's historical delivery data with the advertising platform's understanding of users, popular products are pushed to the most suitable users.

V. Conclusion

This article explains DPA through a case that everyone will encounter. Now let’s go back and see what kind of help DPA provides to advertisers and advertising platforms.

1. Advertisers

• Reduce the cost of creating advertisements. By associating the advertisement plan with the product library, you can launch a large number of products, reach more users, and achieve large-scale launch.
• Through template production, it is possible to dynamically display product prices, names and other related information in advertisements, thus reducing the cost of creative production;
• By connecting with user behavior data and combining it with the model technology of the advertising platform, users with product intent can be redirected to improve ROI.

2. Advertising Platform

• By accumulating user behavior data and product data, we can better understand users, industries, and products, continuously optimize platform models, feed back to advertisers, and promote positive development of the industry;
• Advertising commercialization: From pushing advertisements to users to pushing commodities, it promotes the transformation of the industry.

Author: Baozi

Source: Daily Diary of Commercialized Products