K12 education promotion cases & crowd targeting strategies!

How to improve the quality of Internet users’ access and increase clicks while stabilizing costs?

How to obtain more effective conversions and reduce ineffective consumption among high-conversion populations?

How to optimize the delivery strategy through short-term sub-plan testing?

These troubles in the delivery process always make people feel bald...

So in order to solve everyone’s doubts, today the editor will take the K12 education brand as an example to provide marketing solutions and analysis methods, hoping to be helpful to everyone.

Business Introduction

Market Environment

• Increased demand for tutoring and mentoring for children

• Fierce competition in the market

• National policy support to support the development of the education industry

Industry

• Education and training-K12, basic education

Main Business

• Elementary, middle and high school tutoring

• 1-on-1 intensive coaching

• College entrance examination retake tutoring

Promotion services

• Tutoring: Primary school Chinese tutoring, primary school mathematics tutoring…

• Tutoring classes - subdivided into: College Entrance Examination retake data tutoring, retake English tutoring...

Account Information

Promotion budget: 160,000 yuan/day

Promotion area: Nationwide

Delivery time: 6 am to 12 pm

Promotional business: Primary school Chinese, mathematics and English tutoring, middle school, high school... 1 to 1 tutoring, pre-exam tutoring and other training and tutoring services are currently

Promotion needs: Willing to try new products, eager to reduce conversion costs because conversion costs are higher than those of peers.

Background:

Promotion objectives:

1. The current cost is in a stable stage, but we will continue to reduce costs while maintaining stability, and expand the volume again after stabilization. I hope the CPA will be reduced from the current 440 to 410, reducing the CPA by 30 yuan

2. Increase account quality while stabilizing costs, improve the quality of Internet users' visits, and increase clicks. We hope to increase the number of clicks from an average of 6,000 conversations per day to 8,000.

3. Through the above goal setting, the reduction coefficient for people under 30 years old can be adjusted in the population attributes to reduce clicks from people without decision-making experience and reduce invalid clicks. Improve click quality, reduce invalid clicks, and lower CPA to 410

Market Analysis

1. Industry traffic trend analysis

2. Crowd-Analysis of Internet User Attributes

3. Competitive product analysis

Delivery strategy

Because crowd targeting involves premiums, customers are more sensitive to premiums. Their perception of premiums is that they will simply increase bids without any other effects.

1. As the customer decision-making group tends to be over 30 years old, we can target them based on their attributes.

Through population attribute settings, you can lower the premium for people under 30 years old, set a premium for the core population between 30-44 years old, increase consumption among decision-makers, and reduce the number of non-decision-makers who only view without conversions. Ultimately, you can obtain more effective conversions among the high-conversion population and reduce ineffective consumption.

2. After the premium, the ACP will increase, the number of clicks will decrease, and there is a worry about reducing the CPA

When setting the population, the system will filter by the population attribute labels, and the acp in the account may increase, but under the final calculated result data, the acp may not increase. Communication phase tests, such as AB comparison and setting plans at the same time, plan A sets the population while plan B does not, and the test results will draw conclusions.

Project Description:

⚫ The crowd is differentiated into people under 35 years old. Because the customer business involves K12 tutoring business, the people who need tutoring may be children. They have demand but no money and cannot complete the payment conversion. Reduce the premium of this part, reduce traffic and set corresponding creatives. If there is display, it will be displayed to the netizens in need.

⚫ The group is set to be over 35 years old. This part has absolute decision-making power, and the final decision and payment are made by parents, so most of the decision-making power is over 35 years old. Increase the premium coefficient, attract clicks from this group of people, and create creative ideas that meet the decision-makers' needs to influence clicks individually.

⚫ Conduct small-scale tests on the recommended population through the system to discover the uniqueness of the population attributes and facilitate the selection of subsequent populations.

Data analysis ideas

1. Crowd comparison:

People under 35 years old should adapt to first-tier cities, such as Beijing, Shanghai, Guangzhou, etc., because in second-tier cities, they will basically have tutoring needs and corresponding decision-making power at the age of 27-35. We should change our thinking and divide this part of the traffic by region.

2. Data comparison:

After setting the population, the data has obvious changes. Consumption, ACP, CTR, display, click, consumption keywords and other data dimensions have been significantly improved, driving the overall market to rise. Due to the increase in ACP, the final CPA was checked with the customer. The customer replied that the result was a decrease of 10 yuan. It can be understood that the cost has decreased after the overall test.

Through product data analysis, it is found that consumption plans will fluctuate. The data shows that while the conversion cost is decreasing, consumption in some regions has increased significantly. Through this, we can understand that the demand in some regions has increased again, and we can increase the plan and create a memory list of the population attributes in this part of the region to conduct short-term sub-plan testing.

3. Data and delivery summary

After one week of launch, we pulled out some indicator data and found that the expected data in the account data were relatively consistent. Customers' ACP concerns were gradually dispelled. They would not give up a product because of premium issues, but would understand a product from the final conclusion.

Initial stage of optimization: do it carefully and make reasonable layout

Analyze the attributes of the crowd in the industry, clearly understand the similarities of the crowd attributes, analyze the attributes of the industry, and adjust the distribution of account plan attributes one by one.

Mid-term optimization: when problems arise, conduct a comprehensive investigation

The investigation of problems in the line needs to be comprehensive. For example, if the same attributes are set, the low click-through rate in some regions or plans should be analyzed and investigated to determine whether it is due to the incompatibility of creative materials and population attributes.

Late stage optimization: Strengthen confidence and improve conversion

Multi-layer materials can be segmented by different plans, population attributes, and regions to obtain high CTR materials for optimization analysis.

Can today’s case help you who are still in distress?

Author: Baidu Beijing Branch

Source: Baidu Beijing Branch