APP operation: Can the revenue from application push be increased by 400%?

Push is the most cost-effective and direct marketing operation method for App applications. It has many details and is a great test of the skills of operators. This article will analyze how to improve the benefits of marketing push from the perspective of ROI. (Non-IM, system and other functional service push)

Taking an application with 1 million daily active users as an example, its daily push cost ≈ cloud cost + push platform/self-built platform cost + operation/R&D manpower cost, roughly calculated as 2,000 yuan. (If you don’t push it every day, the cost may still exist, so it can be said that it is almost free)

Daily push revenue = number of pushes × number of people pushed each time × reach rate × click-through rate × average conversion value. The "average conversion value" is calculated differently depending on the profit model. For example, for news applications that use information flow advertising as the main profit model, the average conversion value = average number of ad impressions × eCPM ÷ 1000; for e-commerce applications that use product purchases as the main profit model, the average conversion value = purchase rate × average order amount × average profit margin. Based on the industry average, the daily push revenue is ≈ 3×1 million×30%×1%×0.5=4,500 yuan. In addition, the improvements in the number of active users and health of the application are also very significant (often the KPI of operators), which are not discussed here. Below we analyze them one by one.

Push times

Generally, an app pushes 2 to 5 marketing messages per day (to a single user). Some apps push more than 10 messages per day, but this often causes user disgust. In some cases, the push notifications are turned off, and in some cases, the app is uninstalled directly. Therefore, for operations, there should be a certain degree of restraint in the frequency of push notifications. It is best not to push more than 3 marketing messages per day to a single user.

Number of people pushed

It is generally not recommended to push all messages. For specific user tags, the intersection and difference method is used to select the target and then push it. This can increase the user's click-through rate without disturbing the user too much.

Reach

Arrival rate = number of arrivals/number of sends. There are two main reasons for low arrival rate. One is that the user has turned off the push, and the other is due to technical channel reasons.

Regarding the user push switch, on the one hand, we need to examine the push frequency and push copy to see if there are too many places that harass users. On the other hand, we can also use some techniques to guide users to turn on the push switch, such as price reduction reminders, point gifts, etc., and remind users in some scenarios that are valuable to users, rather than popping up a window to remind users to turn on the push switch right away. If your push content is useful to users, most users will not turn off the push switch.

There are three types of push channels: third-party push platform, manufacturer channel, and long connection. Most applications use third-party push platform, which provides a one-stop push service. It does have advantages in convenient integration, customization, and data statistics. However, its arrival rate is often not 100% guaranteed. Therefore, many applications will also integrate the push SDK of mobile phone manufacturers, such as Huawei push service, Xiaomi message push, etc., and send push messages through the manufacturer's resident notification center process to avoid the message from reaching the user after the application process is killed. Its online message arrival rate can generally reach more than 99%.

Long connections require more R&D resources and there is a possibility that the process will be killed by the system, so it is not recommended.

Click-through rate

The click-through rate is related to many factors, such as copy quality, push style, push time, portrait matching, etc.

As shown in the figure below, the push copy is highly homogenized. How to stand out from the crowd and stimulate users' desire to click is a test of the copywriter's skills. There are many specific methods, such as @user, riding on hot spots, directly displaying discount information, personalized ringtones, adding >> symbols, etc. It is recommended to choose a method that matches the tone of the application, or you can conduct experiments through A/B Testing. In addition, a single push copywriting style can easily make users numb. Changing the style appropriately may bring users a refreshing experience.

The push style is generally in the text format shown in the picture above. Some of them will also be equipped with small picture previews. The addition of Emoji will also make the push style more lively. The industry is currently also trying the push format of large pictures, which attract attention through the display of larger pictures including animated pictures, thereby greatly increasing the click-through rate of push.

As for the time period for push notifications, generally speaking, the push time should be when people are relatively free and use their phones frequently, such as on the way to work and breakfast time (9-10 am), lunch break (12-2 pm), on the way home from get off work (6-7 am), before bedtime (21-22 pm), etc. However, this cannot be generalized. The timeliness of some hot content must definitely be considered. If you are confident in your app, you can also adjust it according to the user habits of your product. There is no need to push push notifications at the above-mentioned time points. You should analyze the usage time periods of your users and fix the time for regular push notifications so that users have psychological expectations for this time and form a certain tacit understanding with them.

Pushing information based on user portraits can not only increase user click-through rates, but also reduce users' aversion to push information. The main user portraits of the current industry push platforms include: age, gender, interests, recently active applications, places visited, etc.

It is also worth noting that in order to improve the experience of mobile phone users, major domestic Android manufacturers have intelligent classification rules for notification messages, which automatically classify and rate notifications pushed by apps and grant different display permissions. Taking Huawei as an example, the corresponding relationship between different levels of notification display permissions on the front end of the mobile phone is as follows:

The display authority of marketing push notifications is relatively low, so many operators find ways to adjust the push copy, adding words such as "Welcome to subscribe", "Message reminder", "You have an unread message", or even "Shipping", "Import reminder" and other obviously misleading words, in order to achieve the effect of being recognized as an important push by the system.

This method is not advisable. Although it can increase the click-through rate temporarily, the gap in the content it jumps to will greatly damage the subsequent conversion of the push and cause user disgust. The manufacturer's algorithm logic will also be adjusted to gradually resolve this type of recognition bias, and even punish such eye-catching push applications, which is not worth the loss. Here you can consider using Huawei's "Quick Notification" service. "Quick Notification" is a value-added service of Huawei's push service, which is mainly for important notifications of applications. It provides an exclusive channel of 1 million per second, reaching users efficiently and in real time, and because of its important notification attributes, it can obtain display permissions such as lock screen display and ringtones.

Conversion

When a user clicks on a push notification and enters the content page, the content must first be consistent with the push copy. Otherwise, the user’s feelings will be deceived, and not only will the brand image be greatly damaged, but the user may even uninstall the app directly. To improve conversion, it mainly depends on whether the content of the application is attractive, and whether the product copy and pictures meet the needs of users. It is the deep-level ability of the entire application. The best strategy is to improve the internal strength. This part no longer falls into the category of push.

Finally, let's go back to our revenue formula. Daily push revenue = number of pushes × number of people pushed each time × reach rate × click-through rate × average conversion value. Assuming the current revenue is 3 × 1 million × 30% × 1% × 0.5 = 4,500 yuan, if the reach rate increases to 50% and the click-through rate increases to 3%, the daily revenue can reach 22,500 yuan, an increase of 400%. The annual revenue can reach 8.2 million yuan, which is still considerable.

Note: All data are reference values. The calculation formula is relatively idealized. The actual data needs to refer to the conditions of each application.

Author: Huawei Mobile Services

Source: Huawei Mobile Services