Hello everyone, I am Li Zimu, a liberal arts student who has been a monk halfway through. I have turned from writing work for 8 years to operating work. I would like to share some of my experiences with you, hoping to bring you a different perspective.
At the end of last month, I took on a new job: the promotion and operation of the company's new APP. Although I have been engaged in marketing for many years, I mainly focus on the direction of copywriting and brand activities. The promotion and operation of APP is a relatively unfamiliar field to me, and the data of last month is not very good, and the daily conversion cost is up to 1500/piece .
Fortunately, in the past two weeks, the results are still good: with the same content, the same platform, and the same registration steps, the single conversion cost has been reduced from the original 1500 to the current 17 yuan. How did this happen?
Before you start, please think about a question: what are the variables of ad conversion?
The physics teacher once talked about the "control variable method". After reading more than a dozen APP articles and an operation book, I found that the fundamentals of the delivery strategy are exactly the same as the "control variable method".
What is the control variable method?
In physics, for multi-factor (multi-variable) problems, the method of controlling factors (variables) is often used to turn multi-factor problems into multiple single-factor problems. Only change one of the factors at a time, and control the other factors unchanged, so as to study the influence of the changed factor on things, study them separately, and finally solve them comprehensively . This method is called the control variable method.
- Lu Xiongwen . Dictionary of Management : Shanghai Dictionary Publishing House, 2013
The so-called "AB test" is a kind of variable control method: adjust a certain part of a page (such as the top banner) (such as changing colors, map elements, etc.), in the same delivery channel, the same Time, deliver to 2 groups of similar target groups, collect different experience data (click-through rate, etc.) to select a better version.
In the "AB test", the content of the page (the color of the top banner, the picture elements) is a variable, and the delivery channel, time, and target population are also variables. Keep the latter variables consistent, and only adjust the content variables to find the best Content, and then, you can adjust variables such as channels, time, and crowds, and through continuous testing, find the best matching variables.
2. Find advertising delivery variables through the user experience path
Any variable that affects the results is an experimental variable. In advertising, we can look for variables based on the user experience path.
The user experience paths for different product promotions are different, and can be roughly divided into three parts:
Take the app I operate as an example.
Let's first introduce this APP - face to:
Target users: managers or teachers of training institutions, custody classes, and kindergartens;
Features: You can use the APP to turn your mobile phone and tablet into a face-swiping check-in tool. The school does not need to buy additional hardware facilities. It can also share data with the school housekeeper cloud school system to handle the school's roll call and class elimination;
Market competition: Because its usage scenarios are very subdivided, although there are many face recognition software on the market, there are very few such software for training institutions check-in and class cancellation, and there are no large-scale competing products for the time being.
Compared with the face recognition system telemarketing list that is often completed by thousands, this APP has the characteristics of low cost and high flexibility, and it can also be used by small organizations with several people, so we define his target user as "Little B", If the C-side delivery method is used for delivery, there are the following variables during delivery:
3. Reduce the experimental cost of control variables
In commercial promotion, input and output are telemarketing list the first consideration. There are many variables in advertising. If you do not pay attention to the method during the experiment, the adjustment cost will be uncontrollable.
As a B-side software with a narrow audience, Liandao consumes very fast advertising costs, and the download cost is only about 7 yuan , but the effective registration is extremely low, 900 downloads, only less than 5 valid registrations, and the effective registration cost is as high as 1500 Yuan/piece .
Operation and product often fall in love and kill each other, and it is not easy to move the path in the APP for operation. If only based on the point of "low effective registration", the product is required to change the registration path from the beginning, and debug on the basis of 1500 yuan/piece. Not only is it time-consuming, but the cost of placement m