In-depth analysis: Operational strategies behind product cold start

Cold start - The state of a social product before users establish effective relationships and continue to generate content and interact. The author of this article analyzes the operational strategies behind cold starts based on relevant cold start cases for your reference and study.

Many well-known products on the market today (such as Zhihu, Bilibili, etc.) have mostly gone through a cold start-up phase. In the early days when R&D resources were tight, they all regarded restricting ordinary users from logging in and using the products as an essential function. Are there the same demands behind these product behaviors? What can we learn from this?

Products that have used the cold start strategy have a glorious history

Zhihu

Zhihu is positioned as a question-and-answer community with a friendly and rational atmosphere and high-quality content.

Zhihu began its invitation-based public beta on January 26, 2011, and launched the invitation code function at the beginning. Only users who obtained the invitation code can complete the registration process of Zhihu. Non-registered users cannot view any content on Zhihu.

In the early stage, invitation codes were distributed offline by Zhihu internal staff to well-known people in the Internet/entrepreneurship circle. In the later stage, the authority to invite new users was granted to high-quality users on the platform by giving them invitation codes every month.

It was not until March 2013 that Zhihu opened registration to the public. During the two-year cold start process, Zhihu screened out 400,000 high-quality users from all walks of life through the registration code system, and continued to output high-quality content with sharp opinions. As these contents were reposted and spread, Zhihu quickly became a most elite platform that produced the most high-quality content in the minds of users.

All unregistered users are eagerly waiting and are proud to be a Zhihuer. Zhihu's invitation code has become so popular that it is hard to get one.

After Zhihu opened registration and launched the invitation code system, the number of Zhihu users grew rapidly, from 400,000 to 4 million in less than a year. In the following years, the number grew even faster, reaching 220 million by the end of 2018.

If Zhihu opened registration to all users from the beginning, in an environment full of miscommunication and empty talk, could Zhihu's high-quality users still be able to continuously output high-quality content for free? Can Zhihu quickly penetrate into the minds of users and become an elite, high-quality question-and-answer content platform in the minds of users?

The author believes that after Zhihu screens out high-quality users through the invitation code registration system, it has greatly reduced the intensity and difficulty of content operations, avoided the shortcomings of the content review mechanism of start-up products, and laid a solid user base, content base, and word-of-mouth foundation for the subsequent explosive growth in the number of users.

Bilibili

Bilibili is positioned as a barrage video website dominated by ACG culture (Animation, Comic, Game).

On January 24, 2010, bilibili was officially created. In the early stage, a limited-time opening system was adopted, and registration was only open on certain holidays (from 2009 to May 2013, there were only 13 days when B station was open for registration). The website was only open to members, and non-registered users could not log in.

On May 20, 2013, bilibili opened registration, but new registered users were only registered members, not formal members. Except for the video playback function, the functions enjoyed by formal members such as barrage, video submission, comment, coin insertion, etc. were restricted.

If a registered user wants to become a formal user, he or she can obtain the rights of a formal member by answering questions (100 questions, 1 point for each question, 60 points or above), or complete the registration through an invitation code obtained from a formal member who has registered for more than three months.

Bilibili is still using the question-answering and regularization mechanism to maintain the atmosphere of the two-dimensional culture and the tone of the barrage content by restricting informal user interactions.

As shown in the figure below, after Bilibili opened registration, the number of users grew exponentially. It can be said that the relatively niche two-dimensional atmosphere can take root in bilibili, and the system of answering questions to become a regular employee has played an indispensable role in this.

Here is an interesting thing I found when collecting information:

Regarding the Cai Xukun playing basketball incident, Cai Xukun’s fans wanted to “flatten” Bilibili, but they were stuck at the question-answering stage and couldn’t even register as a member, making it impossible for them to post comments or barrages.

At that time, some Bilibili users sold their premium accounts on Cai Xukun’s Tieba. A level 4 or 5 account could be sold for 4,000 to 5,000 yuan…

The Guiding Principles Behind Cold Start

The author believes that there is actually a common goal behind these products, which is to verify the value of the product and acquire seed users. The above-mentioned method of restricting the registration and use of non-target users through various product functions is an important method to achieve this goal.

Why do these products restrict user usage during the cold start period?

In fact, the product in the cold start period does not restrict the use of all users. Its essence is to raise the entry threshold and restrict the use of non-target users, because the influx of non-target users will affect the realization of product goals at the current stage.

1. The influx of non-target users will disrupt the formation of product tonality and be detrimental to the retention of target users.

For products containing UGC (user-generated content), the massive influx of non-target users will produce the effect of bad money driving out good money, interfere with the user experience of target users, and lead to the loss of target users.

Generally speaking, depending on the personality of each product, it attracts different groups of people, evokes different emotions, and produces different chemical reactions.

On NetEase Cloud Music, we are emotional and warm while listening to beautiful songs and reading wonderful comments; we are rational and curious while checking professional and serious answers on Zhihu; we are happy and funny while reading Encyclopedia of Embarrassing Stories.

Different product personalities are shaped by both product mechanisms and user content, but a large influx of non-target users will disrupt this chemical reaction.

If users often see jokes in the comments of NetEase Cloud Music, they will often see professional film reviews in Qiushi Encyclopedia.

This will make users feel dislocated and is not conducive to the formation of product atmosphere. It may even cause conflicts and abuses between users with different values, affecting the target users’ impression and reputation of the product.

In the early stages of a product, limiting usage by non-target users through cold starts can better foster a product atmosphere, strengthen the product's positioning in the minds of users, and improve retention and conversion of target users.

1. Non-target users will interfere with key product indicators and affect the verification of product value.

In addition to subjective personal perception, product managers must use data indicators to measure the current health status and stage of the product on an objective level. The influx of non-target users will seriously affect various key data indicators and affect the judgment of product managers.

In fact, the idea of ​​screening users has always existed in real life. For example, if the product is a restaurant built by an operator, then the users are the customers who walk into the restaurant to eat. In fact, the operator screens the customers through factors such as the decoration of the store, the price of the meals, the environment in the store, and the star rating of the hotel.

However, there is no threshold for downloading and using Internet products. As long as you have a mobile phone that can access the Internet, you can complete the visit.

For product managers, if they want to analyze the current status of the product to decide the next strategy, they usually set up key statistical indicators such as retention rate, daily active users, praise rate, sharing rate, etc.

If a large number of non-target users flood into the product, the ratio-based indicators will definitely be lowered, that is, the key behavioral data belonging to the target users will be "diluted". If the product is evaluated based on these inaccurate indicators, it is very likely to reach the opposite conclusion.

Here’s another real-life example:

An IM product targeting internal collaboration scenarios in the TOB field was directly put on the application market after it was launched. As a result, a fake order studio accidentally discovered that this product has both the instant messaging and payment capabilities that they urgently need.

Frustrated by the bans on WeChat and YY, fake order studios have adopted this IM software as their designated order dispatching platform. After recruiting a large number of order takers, their peers quickly followed suit, resulting in a surge in product users, and high user retention rate, activity and other data indicators...

But are the users brought by this group of fake order studios really the target customers that the IM team wants to serve?

Can this group of "users" bring the commercial value expected when the product was first designed? Absolutely not! Even due to the illegal nature of the fake order industry, after the fake order studio has absorbed a certain amount of deposit from the fake order workers, it will pack up and leave, making the deceived fake order workers very angry, and even causing serious impact on the reputation of the IM software.

In the early stages of a product, when there is a lack of sufficient user portraits and inability to conduct targeted data analysis, product mechanisms can be used to raise the entry threshold, restrict use by non-target users, and verify product value through data indicators while ensuring the degree of fit with target users.

Gains and thoughts

Gains

1. The influx of non-target users will disrupt the formation of product tonality and be detrimental to the retention of target users.
2. Non-target users will interfere with key product indicators and affect the verification of product value.
3. When product managers use data to review the status of their products, they must be careful about "pollution" at the source of the data.

think

1. Why did Zhihu (registration invitation system) and bilibili (answering questions to become a regular member system) choose different mechanisms to screen target users?
2. Why have Zhihu and bilibili gradually relaxed the difficulty of screening target users after their products have developed to a certain level? What advantages and disadvantages will this bring?

Author: cc

Official account: Product Wang's Green Grass Pasture