This is the best user growth analysis model I have ever seen

When conducting data-driven growth analysis, data analysts should not only use AARRR to present growth results, but also quantify the entire process of growth decision-making to discover deeper problems. This article goes deep into the entire analysis model around six modules. Data analysts, come and learn.

Data-driven growth is a requirement for data analysts in many companies, but when it comes to specific operations, everyone starts to get tangled up. Although the five words AARRR are clearly written on the growth hacker, when it comes to analysis, they are always complained about:

“There are 10,000 new customers, so what?”

"The activity rate is 50%, so what?"

"The conversion rate has dropped, what can we do to raise it?"

Simply looking at the five indicators of AARRR, it is easy to come to a mindless conclusion like "the conversion rate has dropped, so we need to increase it." How can we analyze it in a systematic way to output a conclusion? Today, I will share a systematic way.

00 The Essence of User Growth

Forget about models, data, and methodology. Just ask the simplest question: "If you were to start your own business, what would you think about?" You would definitely not listen to the success guru's talk about "mindset", "model", or "underlying logic" (if someone really sells these, run away quickly!), but ask a few simple questions:

1. What do I want to sell?
2. Which customers do you sell to?
3. Where to find customers?
4. How to get customers to pay?
5. How much should I invest?
6. How much can I earn back?

These practical issues are the key to business success. The same is true for enterprises. No matter how many new terms are invented, to increase user growth, you must solve these six core problems:

1. Track selection (online/offline, fast-moving consumer goods, durable goods, snacks, services, etc.)
2. Customer group selection (high, medium, and low-end customers in a specific market)
3. Customer acquisition channels (advertising, stores, fission, traditional sales...)
4. Conversion methods (buy one get one free, group buy, flash sale, discount, purchase limit, etc.)
5. Input costs (product cost + advertising cost + marketing cost + operating cost)
6. Output income (sales profit, financing target)

There is an internal logic between these questions (as shown below):

The entire analysis model is based on these six modules. Not only should AARRR be used to present growth results, but also the entire process of growth decision-making should be quantified to discover deeper problems.

01 First: Track selection

There are two main ways to increase user growth:

Start-ups need to compete in the public market and fish in the ocean.

Incubating new businesses within the group company can draw traffic from within and raise fish in the pond.

The data is viewed differently in two ways:

Sea fishing style: It requires evaluation of market space, market growth rate, competitor situation, and requires a large amount of secondary data.

Pond fish farming style: There is already a data base for internal customers, and the scope to be converted is limited, so you can directly do customer group analysis.

In comparison, the sea fishing method is more troublesome, so let's focus on it. At this time, you need to collect three data (as shown below):

1. Current Market Assessment: Assessing How Much Room for Growth Is There
2. The scale of existing players: discover the competitive landscape and assess the difficulty of competition
3. Growth rate of existing players: Find growth benchmarks and choose benchmarks

It is unlikely to obtain accurate data directly for these three data points, so they need to be obtained through a combination of third-party data, industry reports, news from major competitors, or even some less glorious means.

At this point, you don’t have to worry too much about the accuracy of the data (it is definitely not 100% accurate), but you should evaluate the information from various channels to see if they are consistent. For example, if all channels reflect that the industry is a monopolistic competition and the industry is in a period of rapid expansion, then the qualitative judgment is accurate.

02 Second: Customer Selection

When the track is specific to a specific field (fast-moving consumer goods, durable goods, snacks, services, etc.), the portrait, consumption power, and population of the target user group can be locked. There are two things to focus on here: user consumption power stratification and user repurchase behavior. These two points directly determine the growth strategy.

1. User Consumption Power Stratification

In principle, the stronger the spending power of the top customers, the fewer their number. The more we should adopt a growth strategy of "sifting the sand through the waves". After acquiring a large number of customers, we should screen out the big customers through high thresholds + heavy services and firmly grasp their needs. If the spending power of the top customers is not much different from that of the bottom customers, or if users generally have a rigid demand for large-scale consumption, we should adopt a strategy of "letting water flow to raise fish", do a good job in basic services, and expand the customer base.

2. User growth path

If the natural repurchase rate of customers is high and a small amount of investment can trigger repurchase, then a user growth path can be created to encourage users to spend more and encourage cumulative consumption. If the natural repurchase rate is low, a harvesting strategy should be adopted: acquire a large number of new users and encourage old users to bring in new users, so as to maintain continuous growth. (As shown in the figure below)

Note that there is a typical analysis trap here: treating your existing users as the total number of users in the market. When a company does not have a monopoly in the market, it is very likely that the existing users are only a part of the total users. The overall picture of users in the market is different from the user portrait based on the existing analysis (as shown below).

Therefore, when conducting analysis at the customer selection stage, it is necessary to combine it with research/competitive product analysis to timely understand the customer structure of competitors and avoid groping in the dark, which will only make things more confusing.

03 Third: Acquisition Channel

User acquisition channels and conversion methods are directly related to the positioning of user groups. In theory, there are four common forms to choose from:

1. Online advertising (based on the channels and content preferred by target users)
2. Online user fission (there must be a KOL among the target users/the KOC must have sufficient willingness to share)
3. Offline stores (target users are concentrated in a specific city/specific area)
4. Offline sales (there are enough big customers that it is worth one-on-one follow-up by sales)

These four methods correspond to the needs of specific user groups. Therefore, when evaluating customer acquisition methods, the first thing to look at is whether each method can reach the corresponding users, and then look at the conversion effect. Therefore, it is necessary to distinguish between local and global influencing factors, and give priority to the delivery channels and the number of people reached to see whether the goals have been achieved (as shown in the figure below).

After that, we will analyze the conversion funnel of each type of method. Conversion funnel analysis has been mentioned in many articles, so I will not repeat it here. Traditional delivery analysis/customer acquisition analysis also often does this.

04 Fourth: Conversion method

The analysis of conversion methods has been discussed in many articles, so I will not repeat it here. In fact, traditional delivery analysis/customer acquisition analysis also studies conversion methods, and many AB tests are also conducted around "which conversion method is more effective".

For example, to test the customer acquisition effect of an online course, you can use the method shown in the figure below and gradually implement it through multiple version tests.

It should be noted that testing is not unlimited . Each strategy may have an upper limit on its conversion capacity. Therefore, when designing a plan, you can preset the number of tests, investment costs and expected values. For example, if you test three times a month and are still not satisfied, change the plan decisively to avoid getting too deeply involved in the details and missing the forest for the trees.

05 Fifth: Input-output accounting

Input-output calculation is the most important yardstick for evaluating growth. At this step, conventional delivery analysis/customer acquisition analysis will also be done, but people often get bogged down in details and overly focus on the ROI of each channel, resulting in a situation of "picking a general from a group of lame men" (as shown in the figure below).

To do input-output accounting, you should first package the growth strategy and several specific promotion measures/activities under the same strategy as a whole. Evaluate the overall effect first, then look at the details .

As a strategy package, it:

1. Is the growth rate satisfactory?
2. Whether the growth quantity meets the requirement
3. Is the input-output ratio acceptable?

After the overall assessment, look at the details. This will not only help you set benchmarks internally, but also help you avoid "missing the forest for the trees". If you find that a competitor has launched a new strategy, you can also track and observe its effects, and immediately verify the feasibility of the new method, so as to avoid being limited to past experience and missing new growth opportunities.

06 Summary

This growth model approach is mainly to avoid the short-sighted problem caused by focusing only on the immediate growth analysis. The deep insights that leaders expect, such as the following three questions, can only be obtained from a global perspective and systematic observation:

1. Are there any measures that have not yet been taken but are very useful?
2. Are there more techniques to break the routine and achieve unexpected success?
3. Have you reached the upper limit and need to change the track/customer group?

Of course, there are challenges in doing this, that is, the scope of data analysis breaks through the existing data and requires a large amount of industry data and test data to draw conclusions. This is a great challenge for the work of data analysts, but it is very helpful for growth.

Author: Down-to-earth Teacher Chen

WeChat public account: Down-to-earth Teacher Chen (ID: gh_abf29df6ada8)