How can operations break through product bottlenecks?

How can operations break through product bottlenecks?
The product has hit a bottleneck and the order volume is neither increasing nor decreasing. How can we make a breakthrough? Let me chat with you guys today and answer this question at the same time. S Curve vs. J Curve In the product development cycle, you often see discussions about growth curves. Most products follow an S-curve growth pattern: 

 But some products may produce a J-shaped curve: 

 Taking the two most familiar communication tools as an example, the development of QQ follows the S-shaped curve, while WeChat can follow the J-shaped curve. Many people would think that WeChat is more powerful, but my view has always been: If WeChat was not a tool produced by Tencent, then it could be compared with QQ in terms of whether it is awesome or not. Since the two products have the same father, leveraging each other's strengths is an important reason why WeChat can show a J-curve performance. Not only user growth, but most transaction growth also follows this curve. Therefore, in the financial reports of traditional businesses and e-commerce , you will often hear the comment "affected by seasonal factors". This comment actually refers to the situation when the most common objective factors in business cause growth to fall short of expectations. There is actually no good or bad in S-type or J-type. They will all encounter bottleneck periods, or we say, stable periods. No matter what the growth curve is, the demands of shortening the stable period and breaking through the bottleneck are the same. Methods to shorten the plateau period All growth relies on breakthroughs at the data level, and all means also rely on the support provided by data. Many people are familiar with the story of "beer and diapers" and "Target pregnancy index". Whether it is true or not, it actually gives inspiration: If you understand your users, you have plenty of opportunities to try to convert them, and once a method is proven to convert, you can do in-depth research around this method. Here, you can actually understand it by using the mobile phones you use every day and the traffic discounts provided by mobile operators . When users apply for mobile phone card services, operator A asks users to choose a package and tells users that they can participate in an activity. If they promise to spend a certain amount, they will receive a double data package every month for a period of six months. Operator B will recommend corresponding traffic packages to users based on their monthly traffic usage, encouraging them to switch to new packages. Both methods are simple, but behind them are operational strategies tailored for users. Operator A's strategy is to assume that a target of 100 users accept this plan, then analyze the monthly traffic usage of all users, and determine whether this sample group can adapt to the package with half the traffic after the six-month discount period ends. If the sample cannot adapt, what proportion of this sample group will purchase a higher package. Operator B's strategy is to assume that the target number of users is 100 and to observe the proportion of these 100 users who accept the recommended packages. It also analyzes the actual package usage of the recommended users and their repurchase patterns, and continuously optimizes the recommendation model. In "Freakonomicon", some very interesting examples are mentioned, for example, whether the choice of cheap chocolate with average taste and expensive chocolate with better taste will be affected after applying different strategies to consumers. The answer is obvious. Even consumers who prefer better-tasting and more expensive chocolate may shift their choices in certain scenarios due to the strategies imposed by marketers , and may form new decision-making habits after long-term strategy implementation. When it comes to consumer users, operations need to consider applying different strategies. I once said that game operations have greatly inspired Internet operations, and traditional business operations have also greatly inspired Internet operations. Take mobile games for example, especially domestic mobile games. You must have seen many prop packs that can be purchased for a "limited time", or "limited quantity", or even "limited times". The reason why players buy them is because they think these prop packs with restrictions are more affordable. Even if the prices are higher, they still have the urge to try them. However, after several limited purchase promotions, some users will solidify their habits. So, when we evaluate mobile games that make a lot of money, we often joke, "If there is a problem that can't be solved, then spend 648. If it's still not solved, then spend another 648 and it will definitely be solved." This is a joke, but judging by consumer behavior, it is no joke at all. The game between operators and consumers has always been going around in circles of "giving benefits - cultivating habits - converting a group of people - changing the method and giving benefits again - continuing to cultivate habits - continuing to convert". In the end, there will always be a group of users who will be converted, and there will always be a group of users who will not be converted. Let me give you an example of what is going on. A certain business module specifically invites some users to enjoy certain benefits. Initially, users are required to complete three steps, which are optimized to one step. Through targeted invitations through various channels, the operation has increased the conversion rate from 1% to 7%, but it is still not up to expectations. Data shows that some users do not take any action even after repeated invitations. What should we do at this time? The first step is to provide relevant attribute and behavior data for the sample request data of users who accepted the invitation and those who did not accept the invitation, and then analyze them to see where the differences lie. Is it because of age issues that prevent them from operating the system, or is it because their activity level is relatively low and they are not affected by soft or hard tactics? The second step is to extract the analyzed samples, separate out the users who do not accept the invitation but have the same attributes and behaviors as the samples that have accepted the invitation, and try to convert them again. At the same time, optimize the business modules to make the invitation reach more efficient. The third step is to take out samples with larger differences and try to invite them in a different way. After completing the three steps, look at the conversion. We are currently doing the first step. If you are willing to study the work that has been completed, you will find that the improvement from 1% to 7% was not achieved overnight. During this period, three rounds of attempts were conducted. The first round of attempts was not successful, but after modifying the copy , the conversion rate increased significantly. The work of operations always involves repeatedly proposing hypotheses, repeatedly verifying hypotheses, and ultimately finding the driving module to break through the bottleneck. In fact, the principle is very simple to explain, but it is very difficult to put it into practice. The logic of different products and businesses is not the same. Although both are e-commerce companies, the issues that Tmall and JD.com have to consider are different. Therefore, the actual operation must be completed through practice. Methods for testing hypotheses I understand that many operators hope others can give them guidance, but the dilemma is that sometimes the experience of others in different fields may not be able to provide practical support. Therefore, only reasoning can be given. The proposal and verification of hypotheses is actually a method that was repeatedly demonstrated in the popular book "The Lean Startup " last year. This method is actually universal, and with the help of this method, operations can better save costs and improve results. The key lies in how to do it. Let me give you a practical example. If I have 1 million users today, I have a hypothesis, and the conclusion of this hypothesis is that either user activity can be increased by 100%, or it will have no effect. It’s so simple, I might divide the verification of this matter into 3 steps. The first step is to select 1,000 users (this can be done randomly or with certain conditions), and conduct an activity or iteration that only they can see. Then observe the behavioral performance of the 1,000 users over a period of time (such as 1 week, 2 weeks, 1 month) to see if there is any improvement compared to before. If so, find the sample with the largest improvement and the sample with the smallest improvement among these users, and look for the commonalities between these two samples. In the second step, the selection was expanded to two types of samples with 10,000 users each, and the effects were continued to be observed. Here we can basically determine whether the effect will unfold as expected after opening it to all users. If so, then go to the third step and open it to all users. If not, return to the second step, re-sample, and reorganize. Many times, it’s not that we don’t know how to do something, we’re just lazy. But if you are used to being lazy, it will be difficult to achieve refined operations. In the current environment, an Internet company that cannot achieve refined operations is almost 24 hours away from closing down.

Mobile application product promotion service: APP promotion service Qinggua Media advertising

This article was compiled and published by @张亮 (Qinggua Media). Please indicate the author information and source when reprinting!

<<:  What can Baidu Mini Programs do? What can mini programs do?

>>:  How much does it cost to develop a Baidu applet?

Recommend

The price of dog meat in 2020. How much does a pound of dog meat cost in the market now?

The price of dog meat in 2020. How much does a po...

The number of mini program users has exceeded 400 million. Do you know how to promote it?

WeChat has made a lot of moves recently, and each...

A brief analysis of the successful cases of leveraging marketing on New Year’s Eve!

New Year's Day is the beginning of the new ye...

An article reveals the social gene secrets behind Zhihu's $1 billion valuation

Even a community like Zhihu, which is full of hig...

Cracked version of the King of Glory Luna teaching video, Luna gameplay skills teaching video!

Cracked version of the King of Glory Luna teachin...

How UGC establishes content order!

I have said before that we should consider the lo...

3 aspects to analyze the art of channel operation

In the Internet environment, it is said that traf...

5 ways to express your ideas in short videos

As a merchant selling goods, when we publish shor...

UGC community operation: From cold start to mature operation, only these 3 steps are needed!

I first came into contact with community operatio...

Cross-border marketing, what exactly is it crossing?

Nowadays, young people have become the main group...

China Mobile APP Advertising Fraud Analysis Report

PC is disappearing from our daily life scenes at ...

Zhihu video income project, simple operation to earn 10000+ per month

I believe many people have heard that Zhihu could...

Brand Marketing Promotion | Where will the “viral” spread of vulgar marketing lead to?

Introduction: I don’t know when it started, but t...

Tik Tok practical operation notes!

Tik Tok is a new world, and every world has its o...

Want to quickly gain 10,000 followers on your new public account? Take this formula for increasing followers!

Let us first think about what resources are avail...