Analysis of channel operation: 3 major elements + 1 major core point

In the Internet environment, it is said that traffic is king, so seizing the traffic entrance is a battleground. It can be said that whoever can gain users will have the initiative. The traffic entrance is the theme of this sharing, analyzing the art of channel operation .

The topics of this sharing mainly include 3 contents:

1. Brief analysis of channel operation

TalkingData proposed the AARRR model based on the user life cycle and with the return on investment as the goal. These five contents correspond to the five important processes in a product life cycle, namely, from acquiring users, to increasing activity, improving retention rate , and obtaining revenue, until finally forming viral spread.

From this diagram, we can see that the first step of operation is undoubtedly to acquire users (Acquisition). The so-called channel operation means introducing more users from different places. The more superficial operations usually only focus on the number of users. In fact, in addition to paying attention to the number of users, the quality of users is actually more critical to operators.

To improve the quality of promotion and operate the channel well, the principle that needs to be followed is essentially: recommend the right content to the right people at the right time and place.

2. Three Elements of Channel Operation

Based on this principle, I have summarized three core factors that affect channel operations. Channel + user + material means that we need to choose the right channel, conduct user insight analysis, and customize the materials.

Then we will explain them one by one

Channel Analysis

First, let us understand the classification of channels.

It mainly includes: search engines , information flow advertising , technology websites, media promotion, social advertising, third-party electronic markets, resource replacement, ranking brushing, manufacturer pre-installation, etc.

Are all channels suitable for us? Of course not, because each channel has its own advantages and disadvantages. We need to understand the characteristics of each channel and then make a choice based on our own situation.

For example, when we launched a certain product in the past, we found that the advertisements of competitors were everywhere. Then we found that they were actually launched through Baidu Alliance and Google Alliance ads . Let’s analyze the characteristics of these two channels.

Disadvantages: Both the click-through rate and the registration conversion rate are not high, and there are a large number of malicious clicks. Malicious clicks mean that many affiliate websites use plug-ins to simulate clicks to earn traffic fees in order to increase their revenue.

Advantages: large traffic, low CPC cost, and a large amount of exposure can enhance the brand. If the company has an investment budget, we also choose to try these two channels. In addition to increasing brand awareness, the more important thing is to seize the market with competitors.

Let’s briefly go over the characteristics of each channel. (The following part is excerpted from the summary of channel characteristics in game data analysis )

1. Search engine, information flow advertising

• Billing method: CPA, CPC.
• Advantages: Large traffic and controllable costs.
• Disadvantages: Risk of cheating and poor quality.

Overall comment: Monitoring costs are high and user quality is complex, but it is gradually moving towards precision. Advertising networks such as Tencent Social Advertising and Alibaba , Baidu, etc. have strong DSP capabilities. (DSP: Demand-side Platform)

2. Media promotion

• Billing method: CPC, CPT.
• Advantages: Good quality, can drive a large number of downloads in a short period of time, can be exposed in a short period of time, and is topical and popular.
• Disadvantages: The number of app users is limited, and the display and exposure of ads is restricted, which reduces the effect.

Overall review: Poor experience, not sustainable. Media promotion includes media soft articles, reviews, press releases and other means, which are all ways of brand promotion . Appropriate media promotion can effectively connect products with users. In addition, traditional media, TV commercials, micro -movies , videos and other terminal game promotion models are now gradually gaining attention.

3. Social Promotion

• Method: Use Weibo, WeChat , forums, communities and other means to promote, including Weibo big account marketing , grassroots hype and other means.
• Advantages: The promotion cost is extremely low, but it can be spread efficiently.
• Disadvantages: The publicity effect is good, but it is not the best way to acquire users.

Overall comment: Social promotion provides great help in product publicity.

4. Resource replacement

• Advantages: Traffic replacement, better effect quality
• Disadvantages: Unstable traffic, high maintenance and monitoring costs

Overall comment: Accurate replacement and resource optimization are the keys to achieving the best results in this type of promotion. However, we are now facing the problem of user quality, and the overall effect is not as good as before, but there are still a small part of the results that are good, so you can choose to use this promotion method as appropriate.

5. Third-party electronic markets

• Type: Mobile assistant (360), mobile phone manufacturer store (Xiaomi), operator store (China Unicom), third-party market (Anzhi), other channels (NetEase)
• Billing: CPC, CPA, CPT. Now most market channels adopt the joint operation model, namely CPS model.
• Advantages: Good user quality, stable traffic, and maintaining partial operations, namely optimization work.
• Disadvantages: There are hundreds of such market channels, and the maintenance cost is high.

Overall review: The main way for users to download games and the main channel for acquiring game users, directness, authenticity, good positioning, recommendations, etc. will all affect the download and user scale of the product, but the core is still the quality of the product.

7. Pre-installed by the manufacturer

• Billing method: CPA, pre-installed quantity.
• Advantages: Equipment shipment volume control, low cost.
• Disadvantages: User quality is complex, accuracy is very low, and conversion effect is poor.

Overall comment: The capital requirement is large and ordinary developers cannot afford it, so it is not the preferred method.

8. Parallel imports

• Billing method: CPA.
• Advantages: Large number of users, fast and obvious effects.
• Disadvantages: Quality cannot be guaranteed, the phone may be reflashed or cleaned up.

Overall review: Similar to pre-installed versions, quality cannot be guaranteed.

9. iOS jailbreak

• Type: Assistant (PP), Community (Weiphone).
• Billing method: CPT, CPA.
• Advantages: high quality and fast version update rate.
• Disadvantages: higher cost.

Overall evaluation: It has rich user resources and high quality, and is a key choice for testing and promotion.

10. Ranking

• Billing method: ranking position.
• Advantages: Higher ranking, increased exposure, driving natural downloads, and better user quality.
• Disadvantage: Risk of being taken off the shelves. If you stop brushing the ranking, the ranking will drop rapidly.

Overall review: This method is basically no longer used. Instead, points walls are used to improve game rankings and acquire more users. This method mainly focuses on the iOS platform.

User insights and analysis

After understanding the characteristics of the channel above, we will focus on the second factor: users. There are three main methods for user insight and analysis.

1. Third-party data platform
2. Industry related reports
3. DMP data management system

Third-party data platform : Taking 51 Credit Card Manager as an example, let’s take a look at its usage scenarios: intelligent bill management, repayment reminders, credit card repayments, online card application, strategies for getting bargains, financial management, lending, installments, etc.

User positioning: Business-related people: credit cards/loans/installments/financial management, etc., industry-related people: insurance/ securities /stocks, etc.

Through the iResearch data analysis platform, we can see that its users are mainly distributed in first-tier cities, accounting for 43%; secondly, male users account for as high as 83%, mainly those born in the 1980s and 1990s, accounting for 71%, and consumers with medium abilities account for the majority, followed by middle and high-level users.

There are many third-party data insight and analysis platforms, such as Baidu Statistics, Tencent Analytics, Google Analytics, etc.

Industry-related reports : Let’s take a look at how to gain user insights through industry data analysis, taking credit card users as an example.

DMP data management system : Let's take a look at the data insight analysis of industry-related reports. Taking 51 Credit Card Manager as an example, according to the credit card analysis report related to the "China Banking Industry Development Blue Book (2016)", credit card users are mainly born in the 1980s, accounting for 42%, aged between 27 and 37, males account for as high as 86%, with a monthly income of 5,000-10,000 yuan, and have mortgages and car loans. Their current consumption scenarios are mainly supermarket shopping, online shopping, food industry, etc., which are mainly daily payment behaviors. Overall, 60% of users are not satisfied with their credit limits.

Based on the characteristics of credit card users, our marketing advice is: we recommend mainly recommending products that men prefer in credit card installments, points redemption, credit card discounts and other activities. Since their income level is not high and they have great pressure to consume, we can also try to recommend loan-related services to them.

We can also recommend suitable credit card categories to users based on preferential policies and limits of usage scenarios, such as travel cards, high-end cards, foodie cards, etc.

Crowd insight and analysis process based on DMP: Let us first understand the user portrait analysis method based on DMP. DMP actually divides the huge number of users from different dimensions, then enumerates the attributes of each dimension, and finally performs a detailed split on each attribute. Based on the DMP model, we can mine high-quality demographic attribute portraits, user interest portraits, social graph portraits and portraits of related business fields to support advertising marketing, recommendations, product operations , DMP data management and other business scenarios.

So let’s take a look at the crowd insight and analysis process based on DMP.

First, we conduct a portrait analysis of our core users to better understand ourselves. Second, we gain insight into the group characteristics of core users so that we can clearly understand the characteristics of potential users. Based on these characteristics, we can mine more potential users through the Lookalike extension model and the relationship chain extension model. Finally, we test and analyze each orientation and continuously optimize it.

Personalized customization of materials

We have completed the insights into the crowd and understood the characteristics of the channels. During the entire channel delivery process, we also need to do a good job in personalized material customization and channel selection. Personalized materials require comprehensive consideration of factors such as brand, product, activity, channel, activity format, customer base, etc. Different materials should be designed for different groups of people, different activities, and different channels, so that our advertisements can appear in front of the right people at the right time and in the right format.

The choice of channels is based on activities/products/targeted groups and KPIs, and the conversion cost is evaluated before selection is made. The delivery process follows the principle of precise users + massive exposure.

Precision delivery focuses on evaluating conversion effects, channel selection focuses on assessing CPA, ROI, AR PU values, etc., and massive exposure mainly focuses on evaluating CPC and CPM costs.

No matter what kind of activity it is, the delivery strategy is to continuously attract traffic and improve user conversion, while at the same time finding ways to seize competitor traffic and weaken the other party's competitiveness.

3. Data-driven channel optimization

After we understand the three major elements of channel operation, let’s finally look at how to do a good job in channel operation. The channel characteristics, user insights, personalized materials and customization we mentioned earlier are the basis for a good operation. To do a good job in channel operation, you also need to combine data analysis for continuous optimization.

Whether it is prediction or analysis, data is the basis. Without data, there is no operation. In the process of channel promotion , we need to clarify the data goals, such as activity time, investment amount, and expected income. We need to submit data to the development team before the activity goes online, and then analyze the channel data during the activity, from the effect of the entire activity and each channel to the conversion of each connection, the data situation of each time period, etc. Based on the specific details of the data analysis, we can finally find the actual channel position, find the problem, and then optimize the promotion.

In the actual project process, our targeted delivery process is as follows, which is mainly divided into 4 steps: testing, volume expansion, sedimentation and expansion.

• Testing phase: We need to conduct small-batch, multi-material, and multi-channel AB tests for each targeting. Filter out high-quality targeted users, materials and channel resources .
• Release + Sedimentation: According to the testing period, we screen out high-quality targeted users, materials and channels, accurately release them, and continuously optimize the activities. During the activity, we need to sediment relevant methodologies, user groups and labels, such as: what materials are suitable for what users, what channels are more suitable for attracting new users , etc.
• Expansion: Finally, we will accumulate more users during the event. Based on these users, we can conduct crowd insight analysis again to tap into more potential users and more channel resources.

The effectiveness of marketing is mainly related to three factors: targeting, materials and resources. Therefore, in addition to better positioning user groups, user portrait analysis can also help us better understand our users, and then combine them with brand products to customize personalized materials and channel recommendations.

The above is Big V Sister’s summary of channel operations this time. Everyone is welcome to learn and communicate with each other.

The author of this article @Big V Sister is compiled and published by (Qinggua Media). Please indicate the author information and source when reprinting!

Product promotion services: APP promotion services, information flow advertising, advertising platform