A brief analysis of the AARRR model user operation strategy!

There are two purposes for writing this article: one is to help organize my own knowledge system , and the other is to use it to discuss user operation related work with everyone. So I will only write what I think is right. If you have friends who disagree, please leave a message or send me a private message to discuss it. That would be great~

1. Start by reconstructing the AARRR model

After introducing my personal work structure framework as a whole, I will now start to organize my thoughts on each section of the work framework one by one. Today I would like to talk to you about my understanding and reconstruction of the AARRR model.

First of all, based on my personal work experience, I reorganized and processed the traditional A AR RR, focusing more on what I think is useful. I renamed it "Tuwei Work Thinking", hereinafter referred to as "Tuwei", which is mine.

In “Tuwei”, the original AARRR five-step funnel was transformed by me into an “A” “A” “RRR” three-step funnel. Personally, I think the working part of “RRR” is relatively independent and should be viewed in parallel.

The entire user activity work is integrated into Activation, and Retention focuses more on the essence of user retention - the product should continue to meet the ever-expanding needs of users; Refer is brought forward because I believe that the best time for fission is when user needs are just met, and it should not and cannot follow Revenue; Revenue should also be struck while the iron is hot, and its role is not only to make profits, but also to help retain users.

II. A: User Registration Operation

Throughout my work experience, the work of attracting new customers has always been done by a dedicated person. Since I have no intention of taking the job of the person in charge of attracting new customers, I focus on how to increase the registration conversion rate of potential users who download the app, and provide suggestions on optimizing the channels for attracting new customers through data analysis .

1. Registration conversion operation

(1) Registration process selection

Method 1

Method 2

There are generally two types of registration processes on the market: Method 1 has one more step than Method 2, which reminds users to register and log in before they see the product (Method 1 also has a "Decision 1" that does not allow users to choose, forcing users to register and log in before using the product, such as chat products). Method 1 emphasizes that the continuous experience of the first use of the product cannot be interrupted, and it is easier to get user registration, while Method 2 can obtain more accurate but registered users;

It is recommended to determine which one to use for the product through user testing. The test measurement criteria can be selected as: "registration success rate" cross "new user's first order success rate", and the company's stage goals must also be considered (the registration volume looks better during financing , but user accuracy is more helpful for later operations).

(2) Registration page design

Try to use universal icons, add text where possible, and do as many user tests as possible. A page that you think is very clear may be completely incomprehensible to users. Do not set a jump if the registration page can be used on one screen. Even a "Next" button may result in an additional 10% bounce rate.

Don't let users think. I recommend referring to the registration page of Mobike . The new user registration scenario is usually standing next to the bicycle and eager to use it, so there is no need to switch between registration and login, and even no need to set a password. It only takes a few seconds to complete the registration and use without thinking (pay attention to the online customer service in the upper right corner. If there is any problem with the registration, you can contact customer service immediately to resolve it. Mobike does a good job and it is not just about spending money). On the right is XXX's registration page. It takes three pages just for the phone number, password, and verification code. Let's compare them. I won't say anything.

(3) Fill in registration information

The amount of information collected should not affect the next step. It is strictly forbidden to collect sensitive information such as ID card number and bank card number right away. User attribute information can be collected later through guidance, incentives, etc.

(4) Account association registration

It is enough to choose WeChat /QQ to register and log in. Domestic Internet users must use at least one of these two. However, attention should be paid to the operational settings for account merging. Some users will register their accounts via mobile phone and WeChat separately. If this issue is not properly resolved, subsequent active operations will be affected.

2. Registration data monitoring

My commonly used dimensions include: weekly/monthly user registration conversion rate trends in different channels, user activity rate trends, and the choice between average and median.

Here are two common misunderstandings when doing data analysis:

(1) Different dimensions are represented

The low conversion rate on the Android side does not necessarily mean that the conversion rate of Huawei Mall and Xiaomi Mall channels will also be low. While the conversion rate of iOS is high, it is also possible that the conversion rate of a certain model is low due to untimely optimization (such as iPhone X).

The mutual representation of specific channels and overall channels should be cautious, and it is recommended to develop the habit of multi-dimensional small data verification.

(2) Problems with mean and median

When the trend fluctuates greatly, the average is the best fig leaf. At this time, it is best to choose the median as a reference for judgment. For example, the average wealth of me and Jack Ma’s father is a number that I dare not imagine in my lifetime.

When the data differences are similar, the average is relatively simple and objective. For example, the average appearance of me and Daniel Wu does not seem to be too outrageous.

III. A: User Active Operation

After users complete registration, they will usually experience the product for a while before leaving. It is important to capture users during this "moment" and keep them active by allowing them to continue using the product. It usually requires efforts in the following aspects: guiding new users to their first order, marketing awakening design, user incentive system , and user education (cultivating usage habits).

1. New user first order guidance

(1) Newcomer Subsidy

It is often more direct and effective to enable users to quickly discover the value of the product (what it brings to users) and use marketing techniques to grasp the psychological characteristics of users. For example, e-commerce and financial products usually send combination red envelopes to new users to stimulate them to complete their first order, allowing users to feel the value of "getting a bargain."

(2) Reduce user cognitive costs and provide positive incentives for key steps

It should be ensured that users can quickly understand the meaning of the product's commonly used icons, and certain text guidance should be added to new functions to minimize the user's cognitive cost and reduce user thinking. Appropriately add incentive functions, such as a celebration page when a bank card is successfully added, to increase users' sense of accomplishment and pleasure.

(3) Strengthen users’ perception of product value

After a user completes an operation, design should be used to reinforce the user’s perception of the benefit. For example, after using a credit card management product to complete the first credit card repayment, an animation suggesting stress relief appears on the page. These actions psychologically enhance the user's experience of the product's value and make the user feel "cool" using the product.

(4) Guide users to choose product usage preferences

Many products encourage users to fill in more information to assist in subsequent operations. Here we recommend proactively designing several product models for users to choose from. For example, content products often require new users to choose their favorite topics. "Choose what you like" is more accurate than "Guess what you like".

2. Marketing awakening design

Build a user churn warning based on the user's minimum usage. There may be several reasons for user dormancy: the product value does not meet expectations when used for the first time, the product value is not found, the usage interval is long, the user has forgotten about it, or there is no need to use it.

(1) The product value is not expected or the product value is not found

Interest incentives and large-scale promotional activities can be used to awaken customers, but the core should be to optimize your own products and iterate product functions to optimize user experience .

(2) The product has been used for a long time and has been forgotten or has no need to use it yet.

You can launch more frequent functions based on collected user characteristics or integrate tracking and reminder functions into product features. For example, financial management products are launched for salaried users, and monthly financial management income statements for more than 6 months are combined with financial management briefings, financial management knowledge classes, communities and other methods to promote user use of products.

(3) Tracking user behavior

Usage does not mean just opening the product, completing a task, or browsing a few pages. Therefore, marketing awakening is not completed when the user clicks to open the app. We should pay more attention to the user's behavior after opening the app, so as to plan the creative ideas for the next marketing awakening.

In fact, for each of the four reasons for sleepiness above, user research should be conducted to further narrow down the reasons, but I really don’t know how to write it. I feel like the workload for this marketing awakening is so much that it deserves a separate article. I’ll do it later when I have time.

3. User incentive system

The best user motivation effect is to make users think about and look forward to using the product next time. There are two types of incentive systems: quick cash-out and cumulative achievement.

(1) Quick cash-out

A reward is given for each completed task, and the reward will "benefit" the next time a task is completed.

The best example is the offline payment war between Alipay and WeChat Pay . The method of scanning QR codes to pay and getting bonuses appeals to users’ mentality that “I’ll lose out if I don’t use it next time”.

(2) Cumulative achievement type

Small rewards are given in batches for completing tasks of different levels, and "benefits" are obtained after reaching a certain threshold. The member points system usually belongs to this type, and the evaluation levels of Dianping.com also belong to this type. However, it should be noted that as the time distance to "benefit" increases and the difficulty is too high, users will psychologically discount the value of the rewards, so other "benefits" should be added in a timely manner, such as social interaction , a sense of honor, etc.

When many people talk about user incentives, they think of points and red envelopes. In fact, social interaction and a sense of honor can also motivate users to use the app, such as Alipay's "Chicken Game ".

4. User education

(1) Cultivate user usage habits

Understanding the time characteristics of users' use of products and pushing them is a common way. Another way is through a series of long-term activities (about 3 weeks is best). For example, daily news products can regularly push news topics during the morning and evening commuting time to cultivate people's habit of using products during commuting.

(2) Increase user awareness of the product

In situations where users are prone to cognitive bias, we should remind them of the risk of possible harm to them. For example, financial management products, money funds and funds have very different risks and returns. If you are not reminded during operation, users are likely to place their needs on the wrong products. The result can only be that users will suffer losses and complain that the product is a "scam."

4. R: User fission

User fission is generally managed based on work tasks, and is generally divided into daily and active types.

1. Daily fission

It is usually implemented using product functions, and the display moment is usually when the user completes a product use and their needs are met.

Common sharing scenarios are WeChat, QQ, and Weibo. The page design after the invitee opens the link should be considered, including the implementation of login and registration functions, jump to download functions, etc., which can usually be very useful for awakening and retaining users. A common example is the red envelope sharing function for food delivery services like Ele.me.

2. Activity fission

It’s usually a viral marketing campaign . There are a lot of tutorials online about the specific methods, so I won’t repeat them.

In fact, there are only a few types of creative ideas: "Sharing and invitees gain benefits at the same time", "The theme of the sharing copy should meet the positive publicity needs of the sharer". These two are basic principles and there is no guarantee that they will not be done well.

In addition, it is necessary to optimize the convenience of the sharing function and the convenience of use by the invitees. Every additional step of operation or scene switching will reduce the conversion effect (for example, Pinduoduo’s one -stop service of registering, placing orders and paying for group purchases on WeChat).

5. R: Demand Operations

User demand is expanding, so product functions are also required to change accordingly. Product function development mainly has two directions: horizontal expansion and vertical optimization. The changes in both should be based on business models and user demand acquisition.

1. Horizontal expansion

The product focuses on core functions, and adds functions by improving other related needs of users. However, it should be noted that functions should not be grown in a barbaric manner, but should be closely integrated with a series of ecological functions under the core functions.

For example, Alipay was originally just a payment tool for Taobao transactions. It added the transfer function and Yu'ebao, completing the three major functions of payment, transfer, and financial management. In order to meet the needs of offline payments, it added the code scanning function, public transportation payment function, online water and electricity bill payment function, appointment registration function, and even the recently launched provident fund withdrawal function, giving you no reason to uninstall it.

2. Vertical Optimization

For example, if I switch from Youdao Cloud Notes to Graphite Documents, the typing function of Youdao Cloud is really crushed by Graphite, not to mention the online table function. If you want to retain users, you must better meet user needs.

Generally speaking, products need to conduct regular user satisfaction tests and open user interviews (user meetings), make analysis reports on future user demand directions through research, and evaluate the direction of horizontal expansion of the product. As for vertical optimization, it is inseparable from satisfaction surveys, and there is also a little skill: search on Zhihu for "Which experience is better, XX (your product) or XX (competing product)?" After removing the water army, it is basically the feedback from high-intensity users, which is simple and real.

6. R: Value-added Operation

My understanding of Revenue is that users are willing to pay for better use of the product. For example, a user's normal order to purchase goods on Amazon does not count as Revenue. Only when a user purchases an Amazon membership does it count (although GMV will also bring revenue to the company, I believe this work has been completed during the user active operation stage). Users will only pay if you provide more value-added services (concessions, privileges, advanced features).

Therefore, the core of your work is how much your value-added services are worth in the minds of users, rather than getting users to pay for them.

1. Value-added through profit sharing

Users pay to obtain continuous preferential services, which are usually limited in frequency and time period. When calculating ROI of preferential value-added services, we must not ignore its great help in user retention (very few users will churn during the preferential period), so user marketing costs should also be taken into account.

2. Privileged value-added

Privileges mean resource scarcity. The more frequently used a privilege is, the higher the payment conversion rate will be. For example, exclusive live broadcast membership during the World Cup is obviously more attractive to housewives who don’t like watching football, and priority booking privileges are obviously more attractive to business travelers.

3. Functional value-added

Higher storage space, faster download speeds, other functions, etc., must first ensure the user experience of non- paying users . If the paid value is forcibly created by castrating the needs of non-paying users, it is tantamount to killing the goose that lays the golden eggs.

4. Promotional Trial

Promoting paid value-added services and inviting trial users are good ways to increase the number of paying users, but attention should be paid to the pricing strategy of "sharp drop and slow rise". The huge number of new paying users brought about by a sharp price cut may also leave due to a sharp price increase. A slowly rising price can reduce users' sensitivity to price.

VII. Conclusion

AARRR was not invented by me, but due to my limited comprehension and my bias, my "earthy taste" may be different from what you see elsewhere. As for what is right or wrong, everyone is welcome to leave a message to express their own opinions.

In order to get a raise, I have persisted in writing two articles. Starting next week, I will try to write an article about my experience in user research work. I hope you will encourage me! (I will go to Liangzhu Museum on the 30th, and I expect to finish writing it on the weekend of early July)

Finally, I would like to do a small survey with everyone: "Does your company organize activities to take advantage of the World Cup? If so, what is the purpose of the activity?" Welcome everyone to leave a message. If there are more than 50 answers, I will release a photo of me and Daniel Wu in the next issue to prove the rationality of using the average~

Author: TAiFin, authorized to be published by Qinggua Media .

Source: Everyone is a Product Manager