Traffic analysis in fission growth mode

Fission growth has always played an important role in marketing promotion. As we all know, traffic is difficult to obtain and costly nowadays, so more platforms choose to use fission to attract traffic. The author of this article shares traffic analysis under the fission growth model. Let’s take a look at it together.

In the past two years, growth-related topics such as growth hacking, traffic pools, fission marketing , and private domain traffic have always attracted much attention; in an environment where the cost of paid traffic continues to rise, "free" traffic acquisition methods represented by fission have been favored by more and more platforms; no type of traffic can be obtained completely free of charge, and brands, content, and communication all require costs. The "free" here is relative to paid channels.

Regarding fission marketing, there are more and more successful cases in the market, and I have also experienced several projects at work. Today, I will take fission activities as an example to summarize the traffic data analysis content and analysis ideas required in each link of fission activities from planning to completion.

01 Project Process

The complete fission activity process and the corresponding data analysis framework for each link are as follows:

It can be seen that each link of the fission activity is heavily dependent on data analysis, and the correctness of data decision-making directly determines the success or failure of the activity; the various links in the framework are connected in series, involving aspects such as traffic structure, traffic quality, and user behavior. Therefore, fully participating in a fission activity analysis is equivalent to conducting a global review and analysis of product traffic.

02 Target Assessment

1. Evaluate whether the quantitative goal setting is reasonable

At work, in some scenarios where there is no experience or data to draw on, the method of goal setting is just "shooting off the cuff". However, user growth is goal-oriented, and the goals of fission activities must be set reasonably. The standard for judging whether they are reasonable is actually very simple, which is to see whether the fission degree (K) and the proportion of users participating in the activity are reasonable.

Fission degree (K) = number of new users invited / number of old users who initiated invitations

The percentage of users participating in the activity = the number of users participating in the activity / the number of active users

Suppose that the DAU of an APP is 200,000, is it reasonable to set the total number of new users of the fission activity to 100,000?

K=1 means that every user participating in the activity has invited new users, which is difficult to achieve in actual business. According to the estimated qualified standards for event planning, K=0.6, and the number of users participating in the event is 167,000, accounting for 83% of active users. Obviously, the target is too high.

So what is the recommended target value?

Conservative estimate: K=0.5, 30% active users participate, then the target should be set at 30,000; aggressive estimate: K=0.7, 50% active users participate, then the target should be set at 70,000; this gives an estimated range, and ultimately the target can be set based on the actual business situation.

2. Select target seed users for fission activity

‍‍‍‍Target seed users are users who can participate in activities and have high conversion capabilities. In order to make K as high as possible, while globally covering the activity entrance, we should also make focused activity recommendations; when selecting seed users, we need to analyze the behavior of active users.

Taking information flow content APP as an example, when users make choices, they need to pay attention to the proportion of user browsing, commenting, forwarding, favorites, question feedback and other related behaviors; for example, user A browses and favorites more; user B browses and comments more; user C browses and forwards more; then the seed user selection priority is C>B>A.

In actual work, you can set scores for user behaviors, calculate user scores based on the weighted average of the behavior proportions, and select users with higher scores as seed users.

Seed user selection is also an inventory of user behavior and user (traffic) quality assessment of the current product. The behavior scores of selected seed users can be reused for traffic quality assessment. ‍‍‍‍

03 Numerical Planning

‍‍‍The numerical planning part needs to be carried out in sequence.

• Step 1: After the product designs the sharing page, it is necessary to estimate the sharing page conversion funnel based on the designed page sharing path and conversion experience data;
• Step 2: Calculate the spread value based on the conversion funnel, the number of seed users, and the number of target new users;
• Step 3: Calculate the cost value based on the activity budget and the degree of spread

1. Estimation of conversion behavior data of users sharing the page at each stage

The key points of estimating the page conversion funnel are:

• Clarify the product path that users access;
• Don’t forget to consider the loss of each action. The activity designs of different products vary greatly. Just follow the key points of estimation in the actual estimation.

The calculation formula roughly summarized based on common product paths is as follows:

Activity entrance conversion = DAU * activity entrance page visit ratio * activity icon click rate

Sharing path conversion = activity page exposure * share button click rate * shared page successful exposure rate * shared page reach user button click rate

2. Numerical calculation of fission model propagation degree and cost

The numerical calculation of the degree of dissemination, in layman's terms, is routine design, which calculates how much dissemination old users need to generate in order to obtain equity rewards.

Taking the activity where users initiate a lucky bag opening activity to win cash red envelopes as an example, after an old user initiates an invitation to help open the lucky bag, how should the amount that each helping user can get be set? Do new users demolish more than old users? Old users definitely can’t open it? A series of problems followed.

Key points for numerical setting:

• Don’t raise expectations too much;
• Attracting new customers is the main goal of communication, and exposure is the secondary goal.

The numerical setting method can be as follows:

Assume that the amount of the lucky bag is fixed at 10 yuan:

• The amount of money that the first user helps to dismantle is set to a random number of around 5 yuan, to increase the user's expectation of receiving a reward;
• Before inviting new users to help with the opening, the opening amount for old users starting from the second one is set to a random number of around 0.5 yuan, with a 1 yuan protection. When the remaining amount in the lucky bag is less than or equal to 1 yuan, the random number is adjusted to around 0.01 yuan until all the lucky bags are opened;
• If the second user who helps to dismantle the account is a new user, the amount of help to dismantle the account will be set to a random number of around 1 yuan. If a new user has been invited, the 1 yuan protection will be cancelled.

So how can we ensure the cost?

By combining the activity budget and the expected target of new users, the acceptable cost of acquiring new users can be calculated; the range of costs for acquiring new users under the fission model can be inferred based on the maximum and minimum scenarios set by numerical values; finally, by adjusting the parameters that affect the difficulty of opening the lucky bags, the cost of acquiring new users under the model can be controlled to be within an acceptable range, and the model can take effect. ‍‍‍

04 Activity Monitoring

1. Activity monitoring data template creation & data monitoring

Taking the activity of opening lucky bags and getting red envelopes as an example, data needs to be collected from two perspectives: the initiating user and the user who helps to open the lucky bags.

The key points included are:

• Monitor shared links, shared pages, browsing of intermediate pages, and click indicators;
• Monitor the number of users reached and the number of rights issued.

From the user's perspective:

Help dismantle the user's perspective:

2. Timely evaluate the results based on data performance

The core result values ​​of the fission activity are summarized as follows:

• Number of old users reached by the activity → Number of old users participating in the activity → Activity participation rate
• Number of old users who have obtained benefits → Number of new users brought by the activity → Fission degree K
• Activity expenditure equity amount → Activity budget consumption progress → Activity acquisition cost

The core result value can indicate the achievement of activity goals, budget expenditure, and whether the activity is successful or not. During the activity, it is necessary to pay attention to changes in core indicators at any time, and adjust the activity difficulty and budget investment in a timely manner according to different data performance to achieve the best results of the activity.

05 Results acceptance

1. Activity effect data inventory

The core result indicators of the activity effect data inventory are consistent with process monitoring. The content that needs to be inventoryed is: how much money was spent on this activity? How many new users were acquired? Is the customer acquisition cost more or less than that of paid channels? How many old users participated in the event? What is the degree of fission? Which link has a lower conversion rate? How to optimize?

2. Introducing traffic quality analysis

The traffic introduced through activities also needs to monitor user behavior and evaluate traffic quality. The monitored data indicators and analysis methods are consistent with the payment channels.

Author: Dapeng

Source: A data person’s private land