Ten thousand words of practical information | How to achieve commercial monetization of mobile application advertising?

Having been engaged in mobile advertising commercial monetization for many years, I always feel that colleagues in product and business are a pair of "love-hate relationship". From a developer's perspective, it is understandable to focus on optimizing user experience , but the issue of "making a living" is related to our survival. In fact, the two are not so fundamentally contradictory, they just lack a unified goal and relatively coordinated methods.

The outline of this article is as follows:

01 Overview

02 Advertising transaction model

03 Participants

04 Advertising style design

05 Monetization Experiment Process

06 Technical docking process

07Promotional material design

08 Related data indicators

09. Remove advertising

10 Conclusion

01 Overview

During the Dragon Boat Festival , many relatives came to ask me the question they ask every year: "What do you do in your job?" I used to answer: I am in business development, business operations or business strategy, and then attached a lot of explanations of professional terms. This year, I gave an answer that I thought was very down-to-earth: "I help the apps installed on your phones make money," and they seemed to understand it immediately.

If it were a few years ago, when you said the word "app", only people under 40 years old would immediately understand that you were talking about mobile applications. At that time, app developers still focused all their energy on attracting new users and activating users. But in an environment where users are constantly being educated by the market, my 90-year-old grandmother also sent out her first red envelope in her life using WeChat this year. This shows that the era of mobile apps has completely transformed from seizing the entrance to sustainable development. As the entry threshold for late-launched apps becomes higher and higher, it is very likely that investors ’ funds will be burned out simply from the stage of attracting new users to activating users. Therefore, this fast pace requires developers to carry out user activation and user monetization at the same stage.

Six modes of mobile app monetization

To be called a model, it must be able to conform to the principle of "sustainable development". For example, in 2012-2013 when mobile games began to explode, many people believed that financing and listing was the most convenient monetization model. However, in recent years, as VCs have become more cautious about their investments, very few mobile applications have been able to go public . And even if they get a good start, it does not mean that they can support the long-term development of the app. Therefore, if we carefully select the common models that can meet the principles of sustainability, there are six:

1. Value-added services

It is undeniable that the monetization trend of free apps + IAP (in-app purchases) that conforms to "Chinese characteristics" is still the most popular. Most users are willing to spend more money on value-added services of their favorite free apps, and are not willing to spend even a dollar to buy an app they have never used. The IAP of gaming apps is more about encouraging users to pay through "acceleration", which is simply using money to buy time. Social apps have virtual goods such as emoji stores to promote consumption. Tool applications will provide you with more advanced features and privileges, such as manual services or larger storage space.

The monetization of value-added services is undoubtedly the most traditional and relatively ideal monetization model, but the premise for prompting users to pay is that users have very good stickiness and the value-added functions are highly irreplaceable. Many developers will try to include some basic functions in value-added services to encourage users to pay, which will inevitably lead to a large number of user losses.

2. Advertising services

When it comes to advertising services, it will inevitably touch the sensitive nerves of developers. The reason is that it always goes against the user experience. But the emergence of native advertising has undoubtedly minimized the damage to user experience. Native advertising has no restrictions on form and can be integrated into the product without interfering with the user’s normal usage experience. We will discuss this in detail later.

I believe that advertising services are a more sustainable monetization model than IAP because it does not require users to pay in order to use your app features. For users who are interested in clicking on the ads and ultimately make a purchase, the ads do not create a negative experience for them, but rather satisfy their needs.

3. E-commerce shopping guide

Strictly speaking, e-commerce shopping guides are also a type of advertising service, but their potential value is much higher than other advertising service models. The most obvious example is that after most of Sina Weibo's traffic was directed to Tmall and Taobao , its advertising revenue has exceeded half of Sina's portal revenue. And compared to other advertising services, e-commerce shopping guides seem to cause the least disruption to user experience. The reason is that shopping has become the most basic need of users and seems to have become a basic functional section of mobile apps.

But not all mobile apps can generate substantial revenue using the e-commerce shopping guide model. It requires developers to have good technology to make personalized product recommendations to users. Moreover, without large traffic, being an e-commerce shopping guide often leads to silence.

4. Application Distribution

Similarly, application distribution is actually a branch of the advertising service monetization model. But it is more inclined to be used as a tool-type super app. As long as users can get the apps they want more conveniently and quickly, the monetization potential of app distribution is undoubtedly huge.

However, the competition in the mobile distribution market is extremely fierce, with Baidu, Wandoujia , Tencent App Store , and 360 Mobile Assistant showing signs of dividing up the market. As the saying goes, “whoever controls the entrance controls the world.” If it is not a super app that is “a must-have for every computer,” monetization through app distribution seems to be stretched in terms of scale and sustainability.

5. Hardware sales

A small number of apps, such as YiXin 's "Qingguo Camera" and Moji Weather 's "Kongkongguo Smart Air Monitor", have already embarked on the very emotional path of monetization through the "combination of software and hardware". They rely on the extremely high user stickiness for a certain function of the app to provide hardware support. Or they may have a certain level of user authority in a certain field and then release a series of peripheral products. In terms of response, this is undoubtedly the monetization model that has attracted the most attention.

However, mobile apps used to sell hardware are most likely to be characterized by “much ado about nothing”. In terms of the degree of monetization itself, it is very little. There is no denying that there are some outstanding ones, but most of them just win applause.

6. Offline services

By transforming from hardware sales to providing offline services, the path immediately becomes much wider. This is also the reason why many O2O apps can continue to monetize. Photography apps have launched ID photo functions, health apps provide home medical consultations, food community apps provide food delivery services, and travel navigation apps help you call taxis or rent cars. In recent years, O2O has become omnipotent in the country. I believe that many people also think that these apps are making a lot of money.

But it is not difficult to see that after a wave of "burning money", the O2O mobile applications that can survive all have some characteristics. Either they are able to focus heavily on a specific industry (manicure, medical care), or they have touched the high-frequency and urgent needs of the majority of users (taking a taxi, eating out), or they are some retail giants that have formed an online and offline marketing closed loop on the Internet+ ( Suning ). Therefore, it has to be said that the monetization model of offline services is indeed beautiful, but its scope of application is limited.

The inevitability of advertising monetization

After reading these "many" monetization models, if you think that I have many ways to go, then I must make clear this cruel reality. In fact, the six major monetization models are nothing more than two categories, advertising monetization (advertising services, e-commerce shopping guides, and application distribution) and non-advertising monetization (IAP, hardware sales, and offline services). I believe that developers who are willing to spend some time reading this article must have an urgent and pressing desire for money. Then you are almost left with advertising monetization as the only option.

As we mentioned at the beginning of the article, in today's mobile application market , user activation and user monetization must be carried out at the same stage. To put it bluntly, this is actually a "fight to support war" idea. The three non-advertising monetization models either have a very narrow scope of application or require us to spend a lot of time to cultivate user stickiness and accumulate product advantages. Unfortunately, what we lack most is time. "Before the army moves, the food and fodder must be sent ahead." Accumulating financial advantages faster than competitors is often the greatest guarantee for product development.

The Developer Dilemma

I admit that it is not an easy decision to commercialize the mobile app in my hand through advertising, because there will be a lot of problems to face. Here, I intend to list the doubts that developers may face in the form of questions. If you are confused by them, it is worth reading on:

• What placements should I try to advertise in?
• Are advertising and user experience necessarily incompatible? Is there a way to have the best of both worlds?
• Would it be less harmful to the product if we only sold brand advertising?
• How should I price my ad space ?
• Is it very troublesome and costly to connect to a third-party advertising platform?
• How to grasp the qualification review of advertisers?
• The design levels of advertisers’ promotional materials vary widely. How can they be more standardized?
• Why do many advertisers give up on advertising after running for a period of time due to performance issues?
• How do I determine whether the revenue generated by my ad placement is excellent or insufficient?
• Now I want to remove the ad space, does this mean that my work is in vain?

02 Advertising transaction model

Before answering the above questions, we first need to understand some basic knowledge about advertising transactions. From the perspective of advertisers, there are at least five advertising transaction models in the market: programmatic direct guaranteed transactions, preferred transactions, header bidding, RTB real-time bidding open transactions, private transactions, etc. Various transaction models are divided according to whether there is bidding, whether the quantity is guaranteed, the display priority level and the billing method. Please refer to the figure below for details.

From the perspective of us developers, we only need to understand two major categories: one is programmatic direct transactions, which is what we commonly call brand advertising; the other is RTB real-time bidding transactions, which some people simply call performance advertising. In the process of our mobile application advertising monetization, these two models are basically indispensable.

Programmatic Direct (Brand Advertising)

Programmatic Direct is simply a non-bidding model. Once the media and advertisers have negotiated the price and basic traffic estimates, they can directly conduct one-to-one transactions. Generally speaking, the settlement method used is CPD (pay per day) or CPM (pay per impression). This packaged traffic purchasing model generally meets the volume needs of brand advertisers, so it is also called brand advertising.

RTB real-time bidding transaction (performance advertising)

RTB real-time bidding (Real-Time Bidding) means that the media connects advertising resources to the RTB trading market, allowing advertisers to participate in real-time bidding to purchase traffic. Compared with programmatic direct transactions, the traffic inventory and price of RTB real-time bidding transactions are not fixed, and the bidding rule is "the highest bidder wins, and the second highest bid is settled." Since advertisers often set bids based on their own conversion costs, most bidding ads are also called performance ads. The purchasable inventory traffic of performance advertising is generally the part left after the brand advertising is sold, also known as residual traffic. In other words, for advertisers, the priority of traffic acquisition is brand advertising over performance advertising. In the exploration of advertising monetization, the general sign of mobile apps entering a mature stage is access to the RTB trading market, which will make the overall advertising monetization efficiency higher and the sales rate more stable and sustainable.

Brand Advertising VS Performance Advertising

For some mobile application developers with relatively high product quality, selling brand advertisements is almost the first step in advertising commercialization. When we gradually receive payments from brand advertising, we can’t help but ask this question: Do we still need to access the RTB bidding market to sell performance advertising?

Brand advertising does have some natural advantages that make it more likely to be favored by the media. First of all, advertisers who are willing to block off traffic to purchase brand advertising are generally brand advertisers with complete qualifications and a certain degree of popularity. Users will be more receptive to such advertisements. On the other hand, brand advertisers’ material design expertise is generally better. Finally, due to its package-based nature, the price of brand advertising is generally higher, which is very impressive in terms of the monetization level of a single traffic.

Despite this, I still insist that developers must connect advertising resources to the RTB trading market as much as possible to allow more advertisers to bid for the following reasons:

1. The sales rate of brand advertising cannot be guaranteed to be above 60% for a long time, let alone 100%.

Especially for some mobile applications with more vertical attributes, the industry market of advertisers they originally target is very narrow. If it happens to be the off-season for the industry, the sales rate will be even worse. Bidding-performance advertising can almost guarantee that there will be advertisers bidding every day. If operated well, it can maintain a 100% sales rate all year round, and the advertising monetization will be more substantial and sustainable. Therefore, brand advertising is sold first, and the remaining traffic from performance advertising bidding is the most stable advertising monetization model.

2. In this era of brand-performance integration, the boundaries between brand advertising and performance advertising have gradually blurred, and brand advertisers are not fools.

So for the publication rates that are adjusted every quarter, should we charge a premium? What is the basis for our premium? The answer lies in the cost per click of performance advertising. If performance advertisers generally recognize the quality of the media, the cost per click will gradually increase in the process of competing for traffic. From the increase in the cost per click, we can determine whether the selling price of the brand advertising needs to be adjusted. Generally speaking, it is more appropriate to set the daily selling price of brand advertising at three times the revenue from performance advertising. Therefore, performance advertising is the cornerstone for promoting the improvement of our media’s overall traffic monetization level.

3. User Experience

No matter how good the quality of the advertising materials is, everyone would be annoyed to see ads from the same advertiser all day long. Moreover, in this native era where advertising is content, we also need to show users something different to stimulate their desire to click and buy. It is obviously inefficient to negotiate with brand advertisers one by one, so access to the RTB bidding market is the minimum guarantee for a sufficient number of advertisers.

4. Not all ad space can be sold as brand advertising

Today's brand advertisers prefer to purchase advertising space that is large in size and has a certain degree of interactivity. For example, splash screen ads, online activities, skin customization, etc. Even a flow of graphic and text information may not be popular. But on the contrary, performance advertisers like this position very much because more native advertising design can often bring better results. Mobile applications only have one opening screen, but there may be multiple tabs and multiple functional areas. Therefore, only performance advertising can be universally applicable to all forms of advertising.

03 Participants

In the process of advertising monetization, developers will deal with many people who play different roles in the entire advertising delivery node. If we make a very "rough" division, it is nothing more than the owners of traffic [media], the [supply-side platform] that serves the media, the consumers of traffic [advertisers], and the [demand-side platform] that serves advertisers or agencies. But if you look at it more carefully, developers may come into contact with people with more specific functions during the entire process of promoting advertising monetization, such as product PM, reviewers, material designers, etc.

media

From the perspective of advertising carriers, media can be divided into three types: PC web pages, mobile web pages, and mobile apps. In this article, we only discuss mobile apps as a medium. The media acts as a "supplier" in the entire advertising process. Strictly speaking, all decisions, such as what kind of advertising space to provide, how much traffic to allocate, what industries advertisers are allowed to place ads, what selling price to set, etc., are ultimately decided by the media. As the source of the advertising monetization process, media decisions will often determine whether the start is correct. However, developers often have the least understanding of the details of advertising. This gap will cause many media to encounter difficulties in the process of advertising commercialization, and more roles will be needed to participate.

Supply-Side Platform

The supply-side platform is what we often call SSP (Sell-Side Platform). However, at present, the functions of SSP and advertising trading platform Adx (Ad Exchange) have basically been integrated in the country. The media connects advertising resources to Adx for sale. There are generally two types of Adx: one is a third-party Adx that aggregates traffic from various media; the other is private Adx from some large media that have their own traffic. Generally speaking, in the process of media resource docking, we will first deal with three types of people on the advertising trading platform:

• Business Development (BD): Determine the specific advertising resources, price, profit sharing method and ratio.
• Product/Operation (PM/PO): Determine the specific access method (JS/API/SDK) and follow up the entire technical docking process.
• Reviewers: Determine the industries and categories of advertisers that are allowed to place ads, as well as the basic qualifications they need to have. Determine the ad formats that can be delivered ( landing page forms, one-click downloads, etc.).

Demand-side platform

The demand-side platform is also known as DSP (Demand-Side Platform). Simply put, it is a platform that provides real-time bidding for advertising to advertisers and agencies. It can be roughly assumed that the advertising trading platform mentioned above is connected to the media, while the demand-side platform is connected to advertisers. So will our media developers have any intersection with demand-side platforms? There is still an answer.

There are many hybrid DSP companies in the market now, which means that in addition to connecting to Adx, they will also build their own advertising trading platforms to purchase traffic. In other words, the advertising exchange platform and the demand-side platform are combined into one. Youdao Smart Selection and 360MAX are representatives in this regard. When faced with such a hybrid DSP company, as media developers, we will also have the opportunity to contact three types of people:

1. Business operations: People who have a macro-level understanding of the entire advertising resource delivery situation. They will regularly prepare data reports to present the advertiser structure of advertising locations, the pros and cons of monetization, and even the effectiveness of advertisers to media developers. And provide reasonable suggestions for the media's advertising commercialization process (such as how much price limit should be set, the rhythm of traffic release, how to formulate brand sales policies, etc.)
2. Material designer: helps advertisers design advertising materials. While ensuring the advertiser's effectiveness, it is necessary to give priority to the media's user experience, so there is an overlap with media developers.
3. First-line sales and advertising consultants: front-line business personnel who are most familiar with advertisers’ situations and are responsible for introducing advertisers and managing advertising campaigns and their delivery strategies.

Advertiser/Agency

The final demand side that purchases traffic for advertising. Advertisers have various demands for purchasing traffic. Some are for simple exposure, some focus on evaluating users' front-end behaviors (registration, downloading, etc.), and some even analyze users' back-end behaviors in detail (showing intention, placing orders, etc.). Advertisers use a variety of assessment methods to judge the quality of media and ultimately decide whether to place an ad and how much advertising budget to allocate for placement. Therefore, for media that need to monetize advertising, advertisers are the ultimate “breadwinners”.

04 Advertising style design

If you have read this far, you are probably wondering how to execute it. But before that, we should first focus on what our advertisements look like? This requires us to mention the native advertising we mentioned earlier.

Native Advertising

Forgive me for being impolite, but now every advertising trading platform claims that they sell native advertising. Since this term was proposed at the end of 2012, its definition has been expanded and gradually blurred. Here, I think many developers who pay attention to user experience are "deceived". I would like to return native advertising to its three most essential characteristics. You can take a look at the ads in your own apps. Are they really native?

• Native advertising is a profit model based on user experience. This simple sentence tells developers two most fundamental issues—you need to focus on user experience and you need to make a profit. Therefore, when judging whether your ad space can bring in a lot of money, you have to pay attention to the trend of your daily active users, changes in NPS values, and whether users have gone crazy in the comment section.
• Ads have valuable content. What distinguishes native advertising from traditional advertising is that native advertising cannot be simple advertising information, but must be advertising content that can create value for users. This is the transition from "I want you to read" to "I want to read". Of course, everyone has different judgments on what is “valuable”. However, purely transactional advertising information obviously goes against the original intention of native advertising.
• The presentation of advertisements complies with the principle of visual integration. Native advertising emphasizes that the advertisement is the product itself, and its implantation does not destroy the harmony of the page itself, and its construction does not affect the user experience. Therefore, if you insist on telling me that floating banners are also native ads, I will not agree. If the advertisement blocks the original content of the product, it actually does not follow the visual integration and does not start from the user experience.

Specific ad styles

Here are some common ad styles:

Information feed ads first appeared on Facebook in 2006, and were subsequently applied to user friend updates on various social media platforms . Today, information flow ads are no longer limited to social media. As long as there is certain content (regardless of whether the content is produced by users or the media itself), information flow ads can be placed. From left to right, the following pictures show normal text and picture information flow ads, large graphic information flow ads, and icon functional small information flow ads.

Button ads are also called icon ads. Early button ads evolved from banner ads. In the era of native app advertising, button ads are often placed on the same page as the product's function buttons, making them more visually integrated. But the disadvantage is that the content information that can be conveyed becomes very small due to size limitations.

Splash screen ads are the largest ads in an app and are also the most popular ad space among brand advertisers. In fact, splash screen ads cannot be called native in the strict sense, and they also cause interference to users due to startup delays. However, since users have been fully educated on the impression that "normally there is a startup page when launching an app", everyone can accept this setting more obediently. However, in order to further reduce interference to users, many developers will add skip buttons or reduce the size of ad creatives, etc.

The style of ad that is second only to the splash screen ad in display area is generally the focus image ad. But the focus image ad will appear more native. The reason is that the focus image appears in the content position, and users will subconsciously think that it has readable content. And the graphic style + automatic carousel mechanism will more easily attract users' attention.

Incentive video ads have recently become the new favorite in the domestic mobile app advertising market. In fact, we have come across this kind of advertising in many foreign games a long time ago. These ads ask you to watch a game video and will give you hundreds of gold coins or some special props. For advertisers (taking CPC settlement as an example), the effectiveness of rewarded video ads depends more on where the billing node is set. If users are charged when they click on a video, the effect is generally poor (users are just going for the rewards). However, there are also many conscientious advertising platforms on the market that set the billing node on the landing page click after the video ends. Doing so can filter out a large number of invalid users, thus ensuring a fairly good advertising effect for advertisers, while also providing real benefits to users.

The reason for not listing them all here is what I said at the beginning. For developers, these types of ad styles are strictly speaking "fake native" (of course there are exceptions). They will all interfere with users' normal use of app functions to varying degrees. The standard for judging this is still the three essential characteristics I mentioned earlier.

05 Monetization Experiment Process

Now that we have figured out the advertising style, let’s get straight to the point, which may be the most concerned part of the entire article. Starting from scratch, let’s simulate how the initial experimental stage of mobile app monetization would proceed.

Ad location selection

Where should we choose to place the first ad in the app? We need to discuss this issue from two aspects: user experience and monetization potential. First of all, in terms of user experience, we need to provide users with an adaptation period to gradually move from an ad-free environment to an ad-acceptable environment. Here we need to pay attention to the following issues:

(1) Avoid placing ads on the homepage immediately

Do you want the ad you’re still testing to be seen by all users at once? The answer is of course no. Of course, we can also control the number of users who see the ads by splitting the traffic. But don’t forget that in the long run, the location with the greatest monetization potential for a mobile app is most likely on the homepage. If we do not select the right advertiser and do not know the test data of the advertisement, it is easy for us to make wrong judgment if we immediately conduct advertisement testing on the homepage. This seems to be too hasty both in terms of user experience and monetization process.

(2) Avoid placing large advertisements

Prioritize content, prioritize content, prioritize content, I can’t repeat it three times. This is also one of the native advertising principles I mentioned above. On a very limited mobile phone screen, no user wants invalid information to take up too much space. Nowadays, many users hate the web homepage of some portal websites because the double enlarged rectangular ads + couplet ads + floating window ads at the top make them only see one headline of the news headline...

(3) Place it at the bottom and on the right

No user opens the app with the intention of viewing ads, so we need to allow users to achieve their primary goal immediately. The reading habit from top to bottom and from left to right determines that we need to place the advertisement at the bottom or on the right.

(4) Don’t put it at the bottom

After reading the third point and then the fourth point, you will definitely think I am a psychopath. What I need to explain is that ads are not part of the user's reading target, so if the ad is at the bottom of the readable content on the entire mobile phone screen, it will often be ignored immediately. At this time, advertisements will become a stopper for users' reading. Because when the user reads the content up to the advertisement, he/she thinks that the entire reading task is completed.

Therefore, we need to change the role of advertisement from a task terminator to a role that adds the next task starter. This means that, below the ad, you have to reveal a small portion of readable content again. This is especially important in information flow advertising. The specific operation method is to study the size of mainstream mobile phone screens on the market, and then place the advertisement near the bottom of the first screen while revealing some content prompts on the second screen. Then when users read from top to bottom, they will inevitably pass by your advertising position, which will not cause too much interference to them, and can also serve as a "sir, please take a break for a moment" effect.

(5) Don’t try floating occlusion

In the first test of ad monetization, I think floating ads should not be considered. All ads that directly block users from reading information will only bring you user complaints and accidental clicks on ads.

After talking about the archway, let’s start losing some moral integrity. In addition to user experience, monetization potential is also a factor we need to consider. In this dimension, we also need to pay attention to some aspects when selecting the test ad location:

(1) A certain amount of traffic is required

Although we have said not to conduct ad testing immediately on the homepage, it does not mean that we should choose some very corner locations or jump to the second or third level pages. Therefore, too little traffic has no testing value, or it is difficult to draw reliable test conclusions. We can select some functional areas that users will frequently enter after one click, such as the "My" location.

(2) Select a location with a high likelihood of a high click-through rate

Without conducting any ad testing, how can we determine which positions are likely to have a high click-through rate? At this time, we need to look at the click-through rate of the content. The style of native ads will generally remain basically consistent with the style of the readable content on the current page, so the click-through rate of the content can be used to make a certain discount to infer the possible click-through rate of the ad space. However, it should be noted that some locations that guide or encourage users to click cannot be used as a reference for click-through rates (such as check-in locations) and are not suitable for selection as advertising locations. Because locations containing incentives will lead to non-real clicks by users, causing advertisers to lose effectiveness. (Please refer to the description of rewarded video ads above)

(3) You need to consider whether to use the open screen for the first test

Many mobile app developers are keen to use the splash screen to conduct the first test of advertising monetization. But in fact, I think that the opening screen position has many special features, so it cannot actually represent the advertising monetization potential of this app. Moreover, the splash screen ads represent the “most valuable” part of the entire app. We are unlikely to put them on the market at a very low test price, and they must be sold using branded pricing. Generally speaking, only brand advertisers who are looking for exposure will place advertisements, and these brand advertisers often do not provide any feedback on the subsequent effects of the advertisements. In other words, using splash screen ads to represent the advertising monetization potential of a mobile app is an insufficient and unconvincing test.

Having said that, let me give you an example of an ad placement that is suitable for your first monetization test. This is a work efficiency app. We choose the small information flow of the [My] page for testing. You can find that this ad spot complies with many of the principles we mentioned: it is not on the homepage, its average daily traffic is second only to the homepage, its style is native (does not block content), its area is small, and it is placed in the lower middle part of the page (the user's vision must pass through when reading the entire functional area).

Finally, it should be emphasized that the principles mentioned above are all for the first advertising monetization test of a single mobile app. In the initial stage, we need to be more conservative and cautious, but it does not mean that home page ads, splash screen ads and other styles are not feasible, but they should be done later.

Experimental Advertiser Selection

What types of advertisers do we use for our first ad test? Here we need to re-focus on the purpose of our test: to evaluate the monetization potential of a certain ad position without disrupting the user experience to a large extent. It is not difficult to see that advertisements that are beneficial to users can not only satisfy user experience but also bring advertising effects. Therefore, we need to first understand what type of users our users are.

A clear and specific user portrait is a prerequisite for almost all advertising monetization tests. It directly determines the advertiser industry, specific products, average order value range and material production style that we should choose. Without further ado, let’s look at a few examples.

Previously, I conducted information flow advertising testing for a work efficiency mobile app. Since our initial test price was relatively favorable, advertisers from four industries, namely gaming, e-commerce, education, and finance, expressed interest in testing, but we could only choose one. The products promoted by the four advertisers are mobile games, cosmetics discount season activities, children's early education products, and free black card processing services. We looked at the user portrait of the app in detail: male users accounted for 73%, which made us first exclude the test of e-commerce advertisers; the age distribution was 21 to 26 years old, and children's early education products did not seem to have many conversions because there were few parents in this age group; corporate managers and employees accounted for more than 60%, so we felt that the performance of the game would not be too good. Finally, we chose the financial credit card black card product for testing because the attributes of the user portrait are very consistent: more than 50% of users have a pre-tax income of more than 5K and have a certain income ability; more than 60% are in first- and second-tier cities; and 58% have a bachelor's degree or above. In general, we can define the users of this app as male business people with a certain level of income.

A good user portrait can help us avoid many detours when selecting advertisers. The same is true even for material design styles. We once put a pink thermos on the promotional materials of an e-commerce company during its discount season. When they later learned that their users were mostly male, they changed the color of the thermos to black, and the click-through rate doubled.

Qualification and material review

Generally speaking, many advertising trading platforms have their own audit departments, which have standardized audit processes, standardized audit systems, and efficient automatic audit mechanisms. However, when we conduct our first ad monetization test, we often cannot immediately apply these standardized processes. The core reason is that we need to be very cautious about the first advertising test and make sure to select advertisers with excellent qualifications and high-quality materials for testing. From this perspective, the blacklist review mechanism of "which advertisers are not allowed to advertise" is not applicable. The whitelisting mechanism of "only which advertisers are allowed to vote" is more secure. Moreover, in the testing stage of manual operation, the number of advertisers will not be large. We should first abandon the automated system and use a more detailed case by case to select advertisers and serve materials.

Manual operation experimental process

With the advertising location and ready to test advertisers and creatives, you can officially "start". Before accessing the system, the entire advertising delivery experimental process was almost completely manually operated:

(1) Prepare before the advertisement goes online

Test thoroughly before going online to ensure that the newly launched advertising products are functioning normally and that the advertising display/click and other data are reported correctly. In addition, testing of the delivery links provided by the test advertiser is also required. This includes whether the link can be opened normally and whether the response speed of the opening will be too slow. Whether the effect can be successfully monitored when registering, downloading, placing orders, etc. on the page? Speaking of conversion tracking and monitoring, generally speaking, advertisers and media will not fully trust the data provided by each other, so in most non-standard testing environments without access to the system, third parties are used for performance monitoring.

(2) Advertising online

Whether all users should see the ad immediately is a question worth considering. Generally speaking, a safer approach is to first divide some of the traffic to launch advertisements (such as 20%). On the one hand, you can further test whether there are bugs in the advertising function and process online. On the other hand, we can see if the user rebounds strongly.

In addition, we need to cherish every test opportunity. It is best to be able to monitor data as closely as possible. For example, view the traffic and clicks of ads in a time-division manner. This is used to judge the front-end data distribution of advertising locations. For example, we will find that the work efficiency app has a larger traffic and clicks at 9 am, and the reading app has a numerical peak after 6 pm, etc. This will provide a lot of important information for our future operations.

At the same time, we should collect as much data feedback from advertisers as possible during the advertising delivery period. Including the arrival PV of the landing page, real-time conversion data, etc. If the ad bounce rate is abnormally high or the conversion data is abnormally poor, we need to further confirm whether there are problems with the online process of the ad. In the first advertising placement test, the first priority is to ensure that the advertisements are delivered normally.

(3) Feedback after the ad is launched

This may be the most important part of the entire advertising monetization test. What conclusions do we need to draw through this test? Overall, we need to evaluate at least from three aspects: user experience, monetization level and operating costs.

Speaking of user experience, most developers may think of it first by judging by NPS values, but I have to say that this may be quite difficult. Because generally speaking, we don’t send a package separately to launch an advertising function, because the user experience will be very bad. So it is more reasonable to bundle some user functions to post a new version. The NPS measurement is for this entire version, and you cannot judge whether the advertising function affects this value. Therefore, a more direct method should be to go directly to the user comment area to see whether a large number of users complain or criticize the newly launched advertising function. Although this seems less objective, it is the most effective and fast.

As long as the method of assessment of monetization level is appropriate, it will be more "reasonable and well-founded" than user experience evaluation. Many times, we charge advertising fees at a certain test price, but this test fee does not actually represent our possible estimated future returns. Here are two examples that will be more direct:

1. We conducted the first test of information flow advertising with an advertiser, and conducted CPA settlement at a successful registration price of 40 yuan. We divided 20% of the traffic for this test. After three days of testing, we finally brought 300 registrations to advertisers. Therefore, we charged 40*300 to a total of 12,000 yuan for advertising testing, with an average daily income of 12,000/3=4,000 yuan. If the full amount is put into use in the future, the average daily income will be 4,000 yuan/20%=20,000 yuan
2. In fact, the total traffic of this ad location is 300W, and the click rate of this test monitor is 1%. In other words, if it is fully launched in the future, 300W*1%=3W clicks will be generated. On the premise of dividing 20% ​​of traffic, we completed 100 registrations for advertisers per day, so the conversion rate is 100/20%/3W=1.67%. Compared with the same type of media, the average bid level of media with a conversion rate of 67% is about 2.5 yuan. We plan to achieve full bidding investment in CPC in the future, which is expected to be 3W*2.5=75,000 yuan.

Did you find it? In fact, the test costs we see are not equal to our possible monetization capabilities in the future. The reason for this is that data monitoring will become particularly difficult before accessing the system, and advertisers will tend to use CPA for settlement. And without other advertisers competing, neither click unit price nor click rate will show a peak state. So when we make estimates, we need to make an assessment that is more likely to be closer to the true value as B's idea.

After evaluating the user experience and monetization level, we will basically conclude whether to fully commercialize this advertising space. Of course, sometimes a test cannot explain the problem. We need to conduct several more rounds of tests to gradually increase the traffic share of the test, and replace some advertisers from different industries and categories to try. But it should be noted that our testing is based on fully manual operations, which will inevitably cost us a lot of labor costs. There is no system support from material review, online operations to data monitoring, and fee settlement. Therefore, we need to evaluate whether the entire operating cost is equal to our estimated real estate revenue. However, in the final analysis, the most efficient way is to make the decision to fully commercialize this advertising position under the premise of stable test conclusions and connect to the advertising trading platform as soon as possible to achieve standardized system delivery.

06 Technical docking process

It is actually not a very "heavy" thing for the media to implement standardized advertising, but it can bring a huge improvement in operational efficiency and monetization revenue. Currently, the market basically uses API docking or SDK docking modes. Next, we will briefly introduce it separately.

API docking

The server-side docking is just a good choice. The advantage is that it can be quickly connected and launched, and the media's distribution and control of traffic is more flexible and convenient, so the risk of bugs on the client will be smaller. However, the disadvantage is that it is difficult to realize all functions of the advertising SDK, such as user context information capture, ad preloading, rendering acceleration, performance monitoring, etc. The lack of these functions may affect the automatic optimization of the advertising system, which in turn will have an impact on advertising click-through rate, fill rate, and advertising revenue.

SDK docking

SDK docking requires changes to the client, and the development volume will be slightly larger than that of API docking. We also need to ensure that the advertising SDK is stable enough to not cause bugs to the product. Because after all, it is a client integration, if a bug occurs, the client needs to issue a new version to fix it. However, its advantage is that it can utilize all the advertising optimization functions of the SDK to improve the monetization efficiency of media traffic by collecting detailed user information, accelerating ad loading and rendering, and improving advertising data monitoring.

In general, in the early stage of screening advertising partners, using API docking is more flexible and controllable. After basically determining the advertising platform partners for in-depth cooperation, SDK docking is a better choice.

07 Promote material design

Many developers believe that they attach great importance to advertising materials. Because we have fully considered whether the materials will damage the user experience, affect the product tone, destroy the visual interface, violate legal provisions, etc. Therefore, we will perform a very "strict" review of advertiser materials, from size, size, format to tone, text content, and typesetting style.

There is no problem in doing this, but what I am talking about here is considering more than just "legal" standards. Just like the problem of changing pink thermos to black mentioned above, good design materials can directly improve our monetization efficiency.

But the reality is that most of the promotional materials on the market are provided by advertisers themselves. If the material produced by the advertiser has met the "legal" standard, then as a media, we have no reason to require the advertiser to make frequent modifications. Besides, many times we don’t know what the criteria for modification are, but just put forward some suggestions that look more "beautiful" to the naked eye. This is equivalent to our attitude towards brand advertisers' materials. We often think that their materials are of high quality because they look "high-end and high-end". But we have ignored that brand advertisers often do not pay attention to the subsequent performance data of the material, or they are not paying attention to it, but they are just unable to measure it (in the traditional luxury jewelry and fast-moving consumer goods industries, it is difficult to associate their final sales with a certain advertising effect). But the effect of advertisers is different. The real effect of each material will affect their final bid, thereby affecting the real monetization of the media.

How to solve this problem? In addition to our developers, there is also an advertising trading platform that needs to be responsible for the monetization of media advertising. Yes, we must have stricter requirements for advertising and trading platforms in promoting material design. On the one hand, we are responsible for the promotional effectiveness of advertisers, and on the other hand, we are responsible for the monetization efficiency of the media. This makes advertising trading platforms really spend their energy on researching excellent promotional materials.

Next is the question of choice. What advertising and trading platforms are there on the market that do better in this regard? In this regard, I can tell you how they did it first. They will use a large amount of material to put it into the media and continuously collect click-through rate, conversion rate and other data. Then split a piece of material according to layout, tone, elements, copywriting and other units. Track the data performance of these units in different materials and media, and then recombine them. Through A/B test and quantitative analysis, several material templates are finally formed suitable for this media delivery, and then recommended to advertisers or access to programmatic creative backends to form general delivery. That's right, this is definitely not something advertisers can do, nor can they do it through media review. This requires continuous research and operation of a team. On the market, the ones that can do this are Guangdiantong , which has a highly intelligent programmatic creative backend; it has a huge team of designers to continuously provide advertisers with Youdao Intelligent Selection for creative production.

08 Related data indicators

In the entire process of advertising commercial monetization, whether it is the manual operation stage or the standardized operation stage after accessing the system. The data that provides us with all the basis for decision-making is data. So what data do we generally need to pay attention to?

Daily live users contribute income

How much revenue does a mobile application advertising generate is considered high? How to judge the value of the advertising position? Just look at this indicator. Directly divide the revenue from this ad location by daily active users. This metric can be viewed for a single product page or for the entire mobile application. The value of applications that are good for advertising monetization is generally above 3 cents.

Daily live users accept display

Number of ads/number of daily active users. This indicator is more considered from the perspective of user experience. Simply put, it means how many ads will be seen in a day when users use the app. There is no standard for the quality of this indicator, it all depends on our acceptance level and the changes in the NPS value that matches it. But if this value is greater than 10, we have to consider whether the advertisements are being put too much.

ECPM value

The revenue you can earn per thousand impressions on your ad. This indicator can more directly reflect the monetization ability of a specific position. ECPM = Revenue/Advertising impressions × 1000, while in the field of performance advertising, revenue = Advertising impressions × Click rate × Click price, so ECPM = Click rate × Click price × 1000. Most of the time, click-through rate will be affected by the attractiveness of material materials, and the unit price of clicks will often be affected by the advertiser's back-end ROI. Therefore, ECPM can more comprehensively reflect the value of a single advertising space.

09 Dead-advertising

When I was a child, I always had a question. Primitive society was distributed on demand, and in communist society it was distributed on demand. What did you experience in this? The teacher told us that the difference between them is that productivity is at different levels. This problem seems to be similar to our advertising business mutation. When we don’t have ads, we must put them on. And when advertising monetization reaches a certain development bottleneck, we will remove some of it.

I'm more straightforward. The so-called de-advertising method we often use is the membership system. For all app membership services on the market, the first benefit is almost all about removing advertisements. This group of users who choose to become members and remove advertisements often think that the advertising experience is poor and will not click on any advertisements. In other words, the membership system is actually to filter out some traffic that contributes ineffectively to advertising revenue. Moreover, through the membership system, we can also recreate a fortune from this group of users. At this stage, the combination of product and efficiency integration, system standardized operation, and membership system will basically form a stable and sustainable development model.

10 Conclusion

After years of working in mobile advertising commercial monetization, I always feel that my colleagues in products and business are a pair of "happy enemies". From the perspective of developers, it is understandable to pay attention to the optimal user experience, but the issue of "supporting a family" is related to our survival. In fact, the two are not so fundamentally opposed, they just lack a unified goal and a relatively coordinated method.

According to the meticulous historical records, physical advertisements appeared in China as early as 3000 BC. When a tribe needs to exchange cloth for the lambs next door, they start to use every means to "promote" their clothes. At that time, people had more of a primitive need for advertising. Is it true that today, our cute users don’t need advertising?

Write this article to all mobile application developers and to those "tribal people" who are already on or are about to embark on the road of advertising commercial monetization. I hope you will succeed and share your experience with us. In addition, you might as well bring your "little lamb" to us. We look forward to having deeper cooperation with you.

Mobile application product promotion service: APP promotion service Qinggua Media advertising

The author of this article is @Youdao Zhixuan and it is compiled and published by (Qinggua Media). Please indicate the author information and source when reprinting! Site Map