App Delivery Growth: Marketing API

Marketing API (hereinafter referred to as MAPI) is an open API launched by major mainstream advertising platforms based on their own marketing systems to help advertisers better place advertisements. Based on MAPI, advertisers can build a Trading Desk to assist in advertising delivery.

So what capabilities does MAPI provide? How can advertisers and developers use these capabilities to help improve advertising efficiency? This article will take you to explore the details. I hope readers will be inspired...

1. Introduction to MAPI

Marketing API is an open API launched by major mainstream advertising platforms based on their own marketing systems. It can provide comprehensive technical and data support for advertising developers, and realize all-round automated marketing and efficient management.

Let's review first. If an advertiser wants to place an advertisement on a certain media (platform), there are mainly the following common methods:

1. Open an account and place advertisements on the advertising platform: The advertising platform is mature and stable, with complete functions. Advertisers only need to open an account and recharge to place advertisements on the platform.
2. Connect with ADX: Advertisers can build their own DSP and deliver advertisements by connecting to the ADX service of the advertising platform. But generally speaking, since connecting to ADX will distribute the platform traffic to the advertiser side, advertisers can freely choose the amount of placement, so the ADX of mainstream platforms is generally not open to the public. It is necessary to evaluate the consumption of advertisers, and only when the consumption threshold is reached will they be allowed to enter.

So readers may have doubts in their minds: since all major advertising platforms have their own marketing systems and ADX, and advertisers can place advertisements through the above two channels, why is it necessary to launch MAPI?

I personally think that there are mainly the following reasons:

1. ADX has a consumption threshold limit. Unless you are a KA customer, the advertising platform usually does not provide docking.
2. The marketing system of the advertising platform is aimed at all customers. Both the operation process and the functional points are highly unified, which cannot meet the personalized customization needs of customers.
3. When clients conduct marketing activities, in order to maximize their reach to users, they generally choose to place advertisements across the entire network (i.e., place advertisements through different advertising platforms). At this time, it is impossible to aggregate the data from each platform for analysis, making operations inconvenient.

Based on the real needs of the above customers, major platforms have launched MAPI for advertisers to access, providing comprehensive technical and data support for advertising developers to achieve all-round automated marketing and efficient management.

So which customers are suitable for conducting advertising campaigns by connecting to MAPI? How can advertisers improve the efficiency of advertising through MAPI?

1. Technical knowledge: Since MAPI connection requires certain technical development costs, the access party must have good technical accumulation;
2. Understand the business: After all, the interface is separated from the system operation page, so you need to understand the capabilities that each interface can provide and accurately determine whether it can meet your own business needs;
3. Understand effect analysis: Since major platforms can only provide basic data reports such as booth consumption, and cannot display on-site conversion effect data (unless advertisers send conversion data back to the platform), you can connect to MAPI to recycle booth consumption data and perform conversion attribution, so as to analyze the traffic quality of each channel;
4. Need to place ads on multiple channels: If advertisers want to place ads on multiple platforms, they must use MAPI to collect data from each platform for aggregate analysis;
5. Improved advertising efficiency: Since the functions provided by the advertising platform are unified, they cannot meet the personalized advertising needs of advertisers. For example, if all advertisements are placed with different targets, N strategies must be manually created on the platform side. However, MAPI can be used to place advertisements in batches according to customized rules.
6. …

In short, MAPI can help advertisers better deliver advertisements.

Seeing this, you must be thinking that since MAPI is so easy to use, you should quickly connect to it, and it will be easy to place advertisements in the future. Don't worry, let's first look at the two shortcomings of MAPI:

1. Real-time nature of MAPI functions: Currently on major platforms, for the same function, MAPI support programs must be later than the platform side, that is, there is a lag, but the possibility of API taking precedence in the future is not ruled out.
2. Functional completeness of MAPI: Whether from a security perspective or a business layout perspective, MAPI cannot cover all platform-side functions. Advertisers need to evaluate whether the existing functions can meet their own needs.

The following figure summarizes the advantages and disadvantages of the three delivery methods for readers' reference~

2. MAPI Function

Currently, mainstream advertising platforms in the market generally support marketing APIs, such as ByteDance, Guangdiantong, Baidu, Kuaishou, Alihuichuan, etc. Although the degree of support varies from platform to platform, the main functions are basically supported, including advertising, data reporting, account management, etc. Better platforms can provide more auxiliary tools. The general functions are as shown below:

3. Make use of MAPI

The advertising platform's marketing API provides so many functions. Based on this, how should advertisers build a Trading Desk to meet their own advertising needs?

Building a Trading Desk is a long and painful process. As the saying goes, “It’s painful but fun”. I have compiled several different docking stages. Advertisers and developers can refer to these stages and carry out their work step by step.

1. Market exploration stage

Advertisers at this stage are generally just starting their business and are exploring the market by placing a moderate amount of advertising. At this time, advertisers can directly open an account on the platform and place ads without having to invest R&D resources to connect to MAPI.

2. Initial stage of launch

After a period of time, advertisers need to conduct more analysis on the data to verify the effectiveness of the campaign. Since the platform cannot provide advertisers with on-site data, it is necessary to recover relevant data from the platform and combine it with on-site conversion data for data analysis. However, advertiser developers need to prepare the following technical support.

1) Self-built monitoring system

Currently, all major mainstream platforms support the addition of exposure and click monitoring. Advertiser developers can build their own monitoring systems and collect relevant data on advertising spots. The monitoring link must carry relevant identification parameters to ensure the uniqueness of the link. The identification parameters can be used to analyze the effectiveness of different advertisements on different channels.

2) Conversion attribution

Based on the recovered exhibition point data, the core fields are device number and timestamp, which can be attributed in combination with the advertiser's on-site conversion data.

3) Data monitoring - traffic quality

With exposure and click data, you can monitor traffic quality and determine whether there is abnormal traffic. You can refer to the relevant knowledge about abnormal traffic, or add the author on WeChat to discuss some common anti-cheating judgment criteria in detail.

3. Initial launch phase

After verifying the effectiveness of the delivery, advertisers enter the delivery volume phase. During this stage, advertisers invest a large amount of advertising in order to quickly capture the market, which requires higher operating costs. In order to improve operational efficiency, advertiser developers need to provide the following technical support:

1. Data recovery: connect to APIs related to advertising delivery and recover advertising-related data from the platform, including data on plans, advertisements, and creative exhibition points;
2. Advertising rankings: Analyze the conversion data of each creative display point through creative data, and output advertising rankings for operations staff to analyze the delivery effect;
3. Advertisement module: connects to the advertisement add, delete, modify and query operation API, and can operate advertisement delivery;
4. Automation rules: Use advertising ranking data to establish standards for good and bad ads; use automation rules to create batches of ads for high-quality ads to grab market share; and automatically shut down low-quality ads to avoid wasting budget. PS: The automation rules here are a process of continuous exploration, which requires joint discussion and formulation by the operations, finance, and algorithm teams, and full use of AB experiments to obtain the best rule configuration;
5. Data Monitoring - Account: Connect to the account-level interface to obtain the account's consumption and balance. The account status needs to be monitored in real time to prevent insufficient balance or no consumption;
6. Data monitoring - review: connect to the advertising delivery level interface to monitor the advertising review status to avoid the situation where the advertising review fails but the operation is not aware of it, resulting in the inability to increase the volume of advertising delivery;
7. Make good use of the platform's automation rules: For some simple monitoring, you can make full use of the automation rules provided by the platform for advertising/account monitoring;
8. Auxiliary tools: Auxiliary tools for connecting with media, such as copywriting assistants and video tools, etc., to obtain high-quality copywriting, produce dynamic videos, and provide guidance for operational delivery;
9. More tools are waiting for readers to discover further...

4. The stage of selecting the quantity of the product

After a large amount of marketing investment in the early stage, the cost of acquiring customers through advertising has become increasingly high. This is the time to enter the stage of refined traffic operation, that is, to select the amount of advertising investment. There are generally three ways to select the amount of advertising.

1) Media Tags

Advertising can be targeted through media background tags, but this depends on the richness of the media tags. Currently, they are basically some conventional tags. It is still difficult to combine them with your own business.

2) DMP (crowd)

By connecting to the crowd management API provided by the platform, advertisers can calculate the crowd packages that need to be targeted/excluded based on their own algorithms, and synchronize them to the advertising platform through the crowd management API for crowd targeting. However, this involves the exposure of advertiser data, and it is necessary to discuss with the company's information security to decide the form and granularity of data output. The data must be desensitized, otherwise there will be a risk of leaking user privacy data.

3) RTA

In order to avoid synchronizing data to the advertising platform while at the same time wanting to operate traffic in a more refined manner, major platforms provided the RTA interface in 2020, which means asking advertisers whether to place ads for this traffic before recalling the ads, giving the advertisers the choice. For more information about RTA, readers can search on Baidu. There is a lot of information on the Internet, so this article will not go into too much detail.

5. Intelligent delivery stage

In fact, some algorithm-related content has been involved in the above stages, but why do we need to pull out a separate intelligent stage to introduce here? I would like to look at the usage scenarios of MAPI from another perspective.

Most companies, especially those providing SaaS services, usually package and customize the capabilities of a single advertising platform when using MAPI.

But have you ever considered a scenario where you package the ad space based on its attributes? For example, you could package the information flows of Bytedance, Guangdiantong, and Baidu into a unified information flow. In this way, the ads can be divided into splash screen, information flow, search ads, etc. This involves powerful algorithm capabilities and product disassembly and packaging capabilities:

1. Product disassembly and packaging capabilities: Products must be familiar with the MAPI capabilities provided by various platforms, be able to strip out common modules, and integrate and map them to the same level;
2. Algorithm capabilities: Due to the packaging of each company's product capabilities, various algorithms are involved here, such as budget allocation, automatic bidding, targeted strategies, etc. Since the daily budget rules of each company are different, budget allocation is a difficult problem and needs to be paid special attention.

IV. Conclusion

Although Marketing API has some issues with functionality and timeliness, it can help advertisers to better deliver ads to a large extent. This article analyzes the capabilities supported by various MAPIs, and proposes some solutions and path planning for the use of MAPI for readers' reference, hoping to bring inspiration to everyone.

Building a Trading Desk by connecting to the MAPI of the advertising platform is a long and painful process. Let's work together to solve the problem.

Author: jdyx728

Official account: Daily notes on commercial products. Commercial product managers who understand products, operations, and code are welcome to discuss daily commercial ideas and product strategies~