Hand-to-hand Facebook ad delivery process details the notes
Here's the body
As the saying goes, interlaced, such as the mountains, the game industry and cross-border e-commerce advertising methods are actually a little bit near. And most of the current market is about e-commerce, so I also put some of the game launch experience to share to everyone.
Because I am better at summing up and like to organize the experience derived from the theory and practice, so every time I will use x-mind to comb the logic, and constantly update and perfect their methodology. The following methodology is my year and a half of experience summed up (mainly for Facebook delivery and optimization), there may be a lot of things to improve, i hope there are imperfections or wrong places we more advice. The following direct lying goods:
1, pre-launch preparation
1.1 From the user's point of view, output the experience documentation, pros and cons and suggestions for improvement of the game product
As the saying goes, sharpening a knife doesn't cut woodwork by mistake. Be prepared to give you a clear direction and a rigorous implementation plan when you start. Because most of the products I'm pitching are hand-travel products, I'll go to the game and include head market competitions like this product before I start advertising new products, and then i'll export a game product experience document. Specific game product experience documents how to write, there are a lot of information on the Internet.
Why play a good game?
As a delivery person, why go for fun games you want to put on? Because only you play the game well, you will know what kind of game style, game play, mechanism, attraction points and levels of difficulty, etc. , you will be to do what kind of material and copy heart has the bottom, also know in which levels dot may be more for conversion and so on.
1.2 Investigation and analysis of head competition on the market
Including but not limited to their download and installation, ranking and revenue for different countries, etc.
In general, this data can be found through app annie, including the release time of the competition, the release of several bits, the frequency of update iterations and package size, and so on, further data also contains their daily ranking and historical ranking, downloads and revenue seems to require higher permissions to see, but if the Android version can learn through google play the installation of this product is probably at what level.
1.3 Competition channels, materials and copy
Learn which channels and regions where the competition was delivered, and in which countries, and what materials and copywritings were used
Here to recommend a good material collection site called social peta, need to pay a subscription, cheap domestic app growing. But through it, you can see the head of the contestants creative material form is more video, pictures or html, and the creative material for the performance of the whole delivery effect is very important; And these for your own product to do what kind of material, targeting the audience and may use the delivery channel probably a reference.
1.4 Learn about user portraits of the same category
Overview of user portraits of the same category of games (gender, age, interest, education, occupation, etc.)
Audience analytics on Facebook's business platform shows how much audience size audiences interested in head competition look like in each country and region. This is actually more important, such as you want to put the game is a card game, and you need to put in Japan, but do you know how many people in Japan are interested in card games?
The chart below shows that facebook users over the age of 18 in Japan are interested in card games in the audience of about half a million people. The small core audience in your region can affect the frequency and conversion performance of your ads.
1.5 Develop a launch plan
According to the characteristics of this product to develop a launch plan (country, budget, channels, etc.)
According to the characteristics of your own products (is a light game or moderate-heavy game, is a puzzle leisure class or role-playing class, etc.), to determine which countries, because different countries and regions of the game users of different categories of game preferences are not the same; and how much budget each country has budgeted for; The proportion of each country's budget allocated across different channels, etc.; I usually do a table similar to the following before I launch, and update the budget changes at different stages.
1.6 Footage and copy
Combine the characteristics of this product with the market head competitive material selling this product material and copy
Look at the market those head advertisers of the material and copy is how to do, after all, they can occupy a category of the top of the head, their delivery material and copy is certainly of reference value. But it is best not to directly "borrow" the elements or persona in their material, because it is likely to violate Facebook's advertising policy, light will receive warnings and modify those infringing or with deceptive lying strong guidance material, heavy is directly your advertising account sealed, so be very careful! Since you run ads on Facebook, you must know the advertising policies and mechanisms associated with this platform very well!
2. Initial testing work
2.1 Objectives and directions of testing and what test methodto determine
Test with a portion of the budget before you go on full play, determine the objectives and direction of the test, and determine which test method
Initial delivery tests are necessary, typically at least 1-2 weeks to achieve what you need to know. Before the initial test, we must first determine KPIs and metrics, first determine the structure of advertising accounts, advertising objectives, creative materialand and copy, audience positioning, placement, bidding, optimization methods and so on. In general, the purpose of our testing is to find out what the cost of getting an installation (CPI) is, how well the feed is performing, what the quality of the game users we bought, what is the return on advertising spending (ROAs), and so on. Meet the goals you need to understand, determine what kind of ad account structure you need to use, test and compare methods for footage, and target audiences, and more.
2.2 Test period data analysis
Organize and analyze the delivery data during the test to find out what the CPI benchmark is in different countries
If you've been testing for a while and end up getting a country's customer-receiving cost CPI higher than you'd expect, consider whether to stop the country's launch or adjust and optimize the audience for the country you're delivering.
2.3 Summary and material improvements
Organize and analyze footage performance data during testing, follow and understand its important metrics, and give recommendations for improvement of footage
The delivery of material for the overall delivery effect is very important, especially to the late delivery, creative creative for optimizing the cost of customers plays a decisive role, although you can optimize the audience positioning, location, bidding strategy and optimization methods to optimize the cost, but the individual feel that the material creative in the overall delivery optimization is the largest proportion.
Which indicators need attention and understanding for the material? (Take video footage, for example)
There are even more specific indicators of video performance, which can be added in the Custom Bar. By analyzing this data, you can see that some video footage is viewed less than 30% of the number of times the video is played in the first 3 seconds, which means that 70% of the users who see your video ad leave without seeing it for 3 seconds. And the first 3 seconds of a video footage is very important, if not in the short term to catch the user's eye, then you can imagine the conversion rate will not be very good. And as a delivery person, you have to understand the significance of these data indicators, so that the material how to improve and optimize the data support, as well as to the creative production staff to rely on the recommendations.
2.4 Post-use Behavior Analysis
Focus on the quality of buying users, including secondary stays, 3 and 7 retentions and average duration per user, and ARPDAU (average active user value) etc.
In general, we use third-party data tracking platforms such as Adjust or AppsFlyer to track how much we buy. With Adjust, we can analyze the retention of buy-to-let users and natural traffic over a period of time, with data such as the average user duration and the value of the user's in-house purchases. If the purchase of the user's retention and natural flow is too much difference, or the retention situation than the market category competition is too far apart, it is necessary to consider whether the product also needs to be optimized and put on.
If retention is even better than expected and recycling is better than expected, consider whether you want to increase your budget and plan for it.
3 Mid-term optimization efforts
At the beginning of the official launch, the specific basic operating process will not be written, because very simple can be learned in a short period of time, if there is a need can contact me, you can send the basic operating documents. Let me talk about how to optimize the delivery in the middle of the launch from which direction.
3.1 Focus on ROAs ROI
Ads run well or badly, the most important thing is to see if the money spent can be collected back in the expected recovery cycle.
So whether we are optimizing other channels, or material, location or audience, is to be able to spend money to buy users, can contribute more than the cost of the value. How do you calculate your ad ROI and what is your return cycle?
Here's a small example.
For example, if you spend 1000 dollars to buy 1000 installation users, then the installation cost per user is CPI of 1 U.S. dollar, this 1000 users in the installation 3 days to contribute 20 dollars of revenue, then the average per user within 3 days of contribution revenue that ISPU is 0.02 U.S. dollars, so you these 3 days OF ROAs , ARPU / CPI , 2%, can also directly use the three days of revenue divided by advertising spending. And if the 1000 users contribute a total of $1000 in revenue in the third month, your return cycle is 3 months. And if all 1,000 users lose in three months and contribute no $1,000 in revenue, that means you're losing money.
3.2 Optimization of key KPI indicators
For example, CPI, user retention, average duration per user and average number of daily active users
To get a higher ROAs, then we definitely want to control as much as possible the cost of getting each user installed or paying, and how to maximize the user value that each game user contributes. The retention of users, the length of time they play, and the number of active daily users affect in-house and advertising revenue (for example, the three-way game modelled on advertising and internal purchase revenue).
3.3 Optimization of budget allocation
Spend your budget on the most valuable users
In a limited budget, it is possible to get a higher ROAs if you get as many users to install and pay for it. How to optimize budget allocation can be analyzed from the following perspectives:
A) Consider the proportion of budgets in different countries in a limited budget
B) Percentage of budgets in different channels in the same country
C) Budget allocation between different campaigns in a country in a single channel (FB channel for example)
D) Budget allocation between different ad groups under the same campaign
The methodology corresponding to the above can also be summarized as the following four points:
a) Which country performed better and allocated more budgets
b) In the same country, which channel performed better and allocated more budgets
c) A single channel, which campaign is better delivered and allocated to more budgets
d) Which ad group performs better under the same campaign and allocates more budgets
That's what you need to do, based on ROAs, to analyze and optimize the distribution of the various channel budgets, campaign budgets, and ad groups based on the performance of the delivery data.
3.4 Optimization of audience targeting (core audience/Custom Audience/Lookalike audience)
In the early stages of the promotion, it is recommended to use core audience users (through behavioral targeting, precision interest words and competing words) to acquire new users. In the middle of the rollout, core audiences target new users, custom audiences promote active/remarketing for older users, and lookalike audience targeting to capture new users of similar high value. Later in the rollout, use Your Custom Audience to promote live or remarketing, Lookalike Audiences continue to look for high-value new users, and use core audience targeting to pull new users in new markets.
As for what kind of lookalike audience to choose, first establish a Custom Audience. Custom Audiences can create Custom Audiences by interacting with Your Page, adding payment information or paying, video viewing, setting up Custom Audiences through levels or achievements, and then using these seed sources to build lookalike audiences to find new, high-quality Lookalike audiences. It's very effective to optimize your performance by constantly optimizing lookalike audiences.
There are several ways I know about how to create the Custom Audience you want:
Custom audiences can be created directly with the audience feature under FB BM background
Create a Custom Audience with the Facebook app analytics analytics analytics analytics feature and then share it in your ad account
In addition, if your product is connected to a third-party platform (e.g. Adjust for example) and has the appropriate functionality, you can create an audience through the Audience Grouping tool in Adjust and share it with your ad account
As for the specific operation here is not a description, if there is time, then may write a detailed operation article
3.5 Location Optimization (mainly divided into automatic and manually adjusted bit)
First of all, what you need to know is that ads on FB will appear on these locations below. Details of each site can be learned directly from this link:https://business.facebook.com/business/ads-guide/video/(Wall needs to be turned)
In the early stages of promotion, personal advice is more suitable for automatic locations, because FB's system learning will more effectively help you find the users you target in different locations. Sometimes you'll find that the user quality of the motivational video location under the audience network will be higher, and you may manually turn off the other locations, but at the same time your ad exposure is smaller and you won't be able to get it from some potential users.
3.6 Ad Optimization (MAIA, AEO, VO)
Advertising is optimized in three main ways: MAIA (installation maximization), AEO (event optimization) and VO (value optimization). These three optimizations correspond to buying a typical download installation user, and those who are more likely to complete an event, such as paying for it, and the group of users who have the highest ability to pay.
Refer to the illustration below ：
Choose a different way of ad optimization, and the cost of each goal will be different. In general, VO and AEO optimization mode of users to buy each cost will be higher than install users, but at the same time the quality of users will be higher, the corresponding recovery will be better. In general, in the initial stage of promotion, most will choose install and AEO optimization method, later can be optimized by VO to find more paid value users.
But regardless of which optimization is chosen, our goal is to buy more users and the ad recycling cycle will meet our expectations. So you can optimize the budget allocation ratio according to different optimization methods to achieve a balance between the size of the amount of users buy and the amount of ad recovery.
3.7 Optimization of the Bid Strategy (Lowest Fee Bid VS Target Charge Bid)
The main role of the bidding strategy is to control the cost of the amount you buy. Here are two main bidding strategies:
1. Minimum fee bid (with no upper limit)
2. Target fee bid.
The biggest difference between the two is that, in practice, the former may rise as your budget increases, your CPI will rise, and the latter is that when you set a target bid, the average installation fee is about the same as the bid you set, but there is a good chance that your bid is too low to run out of volume, the latter of which is more accurate to bid. In general, it is recommended to use the minimum bidding strategy early in the promotion, as Facebook's machine learning allows for as many installations as possible on a limited budget. And, at the moment, Facebook is actively promoting the CBO (campaign budget optimization) feature, and it's likely that it will be gone to adjust the budget in the ad group later.
3.8 Optimization of creative spouts
The entire ad delivery process can be summed up in one sentence, that is,"Use the right channels to deliver the right advertising vector to the right people."As a very important link in the advertising process, advertising material plays a decisive role in optimizing the cost of buying in the later stage. And for the optimization of creative material is also a very long-term process, later time will write a specific optimization of creative material. However, there are some basic production requirements for the production of commercial video footage.
4. Summary and continuous optimization after launch
Regarding the continuous optimization of the late stage of delivery, the process of continuous optimization of those indicators in the medium-term delivery is mainly the process of continuous optimization, and with the different promotion period to adjust the pace of promotion, and later more need to analyze and summarize the data of different promotion strategy period. As far as individual experience is concerned, the later stages of promotion need to pay more attention to the overall promotion strategy, including research and expansion of new markets, budget management and allocation of current marketing channels, and localization of different channels and national production.
Another point is that it is also an important point in terms of the growth of natural quantities. At present, this aspect of research is also in the exploration, later if there are some ideas and harvestwill will also be updated, but also welcome your advice and exchanges.
■ End ■
Go to "Discovery" - "Take a Look" to browse "Friends are watching"
Sent to have a look