Hello everyone, I'm a lot of fish.
In April 2022, I started to contact Little Red Book. It took me about two months to stay up late until the early morning every day. I paid tens of thousands of dollars to buy all the classes in the market and listen to them again. Several times, I spent all night brushing the lessons, articles and content, and the whole process of Little Red Book was run through. The monthly customer of zero launch reached 4000+, and the monthly GMV exceeded 2 million. Last year, it achieved 11 million GMV, which was relatively good. Last year we also shared in the wild, and today's text is a new supplement.
Today, I will talk about how I did the practice of getting customers for the matrix number. I hope you can learn something different from my sharing.
The content shared today is mainly divided into the following five sections:
1. Little red book must know and know
2. Choose the right keywords to dominate the small red book
3. How to efficiently create matrix number content
4. How to maximize the drainage efficiency
5. Management and establishment of matrix drainage team
1、 Little red book must know
I wonder if you have seen the picture below.
The user group of Little Red Book includes the top three: girls, young women and children. The sixth man also has the top purchasing power in Little Red Book. Especially for products with high customer price, the transaction conversion rate on the Little Red Book is very high, which should also be felt by young partners.
In fact, we are also getting customers all over the world, including Zhihu, Tiao Yin, etc. But Xiaohongshu's overall conversion rate is higher than that of other platforms, and its user quality is better. Therefore, there is no doubt that the Little Red Book must be a part of our continuous layout.
1. Basic rule of getting customers for Xiaohongshu matrix number
1) Violation detection
Before doing the Little Red Book, you must check all personal channels, and do not include marketing words such as WeChat, QR code, purchase on behalf, Taobao, and spell xx. Now the profile does not support the mailbox. Many of our account profiles have lost their mailboxes.
2) Try one machine and one number
Many children want to store a large number of accounts in a matrix way, but there are few mobile devices. It is impossible to put multiple accounts on one mobile phone.
The same mobile phone cannot have more than two accounts. It is better to have one account for one phone. All our mobile phones now have one phone, one card and one number, and the mobile phone card should be direct rather than the flow card.
3) Close Favorites and Fan List
Before doing the matrix number, I hope you can close all the favorite columns and fan lists of all accounts.
If you do IELTS or study abroad related categories, you will find that when you pay attention to some head accounts, you may receive at you from others, that is, you directly pry the user away.
4) De duplication
When you do matrix numbers, you must pay attention to the original protection, adjust the format, eliminate duplication of copy, and adhere to the diversity of content.
5) Different accounts are presented by different subjects
Why do you say that? In fact, there are two types of matrix numbers, one is parent child account, and the other is brother account.
What is the parent child account? For example, I am the official account of Xiaomi. There may be Xiaomi Preferred, Xiaomi Genuine Products or Xiaomi Selection below. All accounts are based on Xiaomi.
But how do we play? All accounts are presented in the form of different subjects.
Just like Unilever's shelf strategy, you buy Head&Shoulders, which is Unilever's, you buy Rejoice, which is also Unilever's, you buy many other related shampoo, which is Unilever's, but you may not know it.
The advantage of this is that the same customer may ask about a dozen of our accounts in a row, and our account gives him a relatively high anchor point. If he changes to another sales account, and then changes to another form of quotation, the user will be directly sold by us.
2. Recommended rules of Little Red Book
When you create a small red book account, I suggest that all copywriters check the sensitive words to avoid some current restriction problems.
I often use two tools, one is the zero gram word search, the other is sentence Yiwang.
All the contents of Little Red Book will be checked by the machine first. The machine will first check your content for sensitive words, violations, or some situations that are not allowed.
If there are sensitive words, the notes will be reviewed manually. If there are violations, the notes will not be recommended.
The following is a brief version of the recommended rule legend I drew.
The key of Little Red Book is its algorithm, CES algorithm model.
At the recent Little Red Book Grass Planting Conference, it was also pointed out that the likes are not critical now, but more importantly, comments, forwarding and attention are the three data.
All of our content revolves around these three points. For example, when we are making data numbers, we will ask users to respond to some numbers in the comment area, and then we will interact with users. The basic purpose is to pull the weight of posts.
3. About breeding number
Does Little Red Book Need a Number? You need a number! Both enterprise accounts and personal accounts need to be maintained, and the old account is always preferred.
This is a conclusion we made after testing dozens or hundreds of accounts. The weight of an old account or an account that has been maintained is relatively higher.
How can I get a pet? That's what we did.
1) Ask real people to do it, and send it after 3 days
We arrange an intern or someone else to do this. His task is to raise 10, 20 or 30 mobile phones. Reserve and maintain all accounts in advance. We will use some automated tools, such as the automatic clicker. After simulating the manual process, we can directly press Start to automatically brush.
2) The first few articles do not issue marketing, but dry goods
3) Post weight with topics
For example, if you are a mother and baby, you can post some topic posts to do account weights. For example, is monthly clothing an IQ tax? To attract vertical fans or users.
2、 Select the right keywords and dominate the small red book
In this section, I will tell you how to select keywords and how to achieve real category occupation.
What is category occupation? For example, when you search for "Changsha Decoration", in addition to the advertising space, the first 10 posts, perhaps the first seven or eight, are all your accounts, so you can achieve the occupation of keywords, or categories.
The traffic of small red books falls into two categories. The first category is recommendation traffic, that is, the traffic when we usually swipe the discovery page. The second category is search traffic, where users enter your notes through search keywords.
The conversion rate of search traffic is significantly higher than that of recommendation traffic. Therefore, selecting the right keywords will directly affect your number starting cycle and future long tail traffic.
To do SEO, or keyword coverage, the mental skill is to select 1 match N, find a core keyword, and then match N long tail words+related words.
1. How to find keywords
Key words are divided into three parts: core keywords, related words and long tail words.
What are the core keywords? For example, studying abroad in the United States or studying abroad is the core keyword, and it is related to studying abroad in Japan or Australia. These are called related words, and some words that may have related relations around the keywords.
There is also a long tail word, such as where to study in the United States, what institutions to look for when studying in the United States
There are two main ways to find the key words.
1) Free channel of Little Red Book
One is small red book search. For example, in the drop-down list of search, when you search the category of studying abroad, a row of long tailed words related to studying abroad will appear.
These words are arranged according to the search frequency, that is, "countries with low cost of studying abroad". The search frequency is higher than that of "diploma certification of international students".
There are also related words on the search page. For example, when I searched for studying abroad, the related words appeared are "applicant vlog, the United States, the United Kingdom, the Netherlands".
You can find an Excel table and record these keywords and drop-down words.
After searching for a keyword, for example, when I search for the postgraduate entrance examination, the algorithm may calculate that I am interested in some things related to the postgraduate entrance examination. A light gray font will appear in the search box. By default, a "work postgraduate entrance examination schedule" will be pushed, which may also be a relatively frequent word.
The same is true for the "Guess you want to search" section below, such as the time arrangement for taking the postgraduate entrance exam, the love between graduate students and junior college students, the job that is suitable for work when suddenly you don't want to take the postgraduate entrance exam These are pushed to you around the tags of your account. As long as you search enough, these tags will get closer to your category.
The next "search and discovery" is a hot search. If some keywords related to your category appear on the hot search, you must go to the topic in time.
How to determine which keywords we need to maintain?
When entering topics on the posting page, you can see the number of views next to each topic. By dividing the number of posts by the number of views, you can get a competitive ratio.
For example, for the purpose of studying abroad, I search for studying abroad in Britain, studying abroad in the United States, studying abroad in Japan, etc., and use the number of topic views/posts collected to get a competitive ratio. The greater the competitive ratio, the better the word is.
But here we should pay attention to one point. This is all the browsing data. It is not recent. There will be inaccuracies (previously very popular content may be cold). We should also combine multiple data to judge.
In addition, there is a keyword planning tool in the spotlight platform that can also be used. After you click it, there will be a word deduction tool, upstream and downstream word deduction tool, and note deduction tool. For example, if you search Beijing Decoration, you can see the number of searches for this word, as well as some of its corresponding bidding.
The amount of search terms is also a very important reference index.
2) Data Tools
We use 5118 and Cicada Xiaohong.
5118 is an SEO website, which has been for many years. When I first came into contact with seo, I came into contact with 5118.
It has a section called Xiaohongshu Keyword Mining. After you search for the keyword you are interested in, it will give you a long tail traffic index of the keyword, as well as some of its notes. You can use it to judge which words are suitable for you.
Another is related word mining. You can search related words on the 5118 platform and find many words related to decoration. We may have to rely on our own brains to figure out what keywords we can do. After using related word mining, we can use machines to liberate human resources.
For example, decoration and house decoration, there are more words. If you expand this word, you can learn all the words of your category, which is also a basis for selecting your topic.
5118 also has a deep mining of drop-down words.
For example, when we search for decoration, the first column is the decoration case effect picture, and there are also drop-down words on the second floor. It may be the villa decoration, kitchen decoration, duplex decoration effect case picture, and so on.
Each drop-down word corresponds to a topic and the direction of the post you can make. Around these drop-down words, you can make more posts to occupy the whole decoration posts.
The second tool is cicada Xiaohong.
It has a hot word search section. When you search for overseas students, there will be popularity, related notes, etc. It is not displayed according to word frequency, but after you export it, you will get a complete table.
Let's make a simple analysis of this table, take the interaction amount and divide it by the number of notes, and we can probably get a competition situation on the Little Red Book.
For example, in this table, it is obvious that the ratio of interaction and notes of "studying abroad" is very low, so this word is not worth our layout.
Looking at the keyword "overseas students", it is obvious that the ratio of interaction to notes is very high. Its posts are completely unsaturated. You will find that there are only 4 related notes in the past 7 days. If I put a large number of accounts in the posts, I can eat this part of the traffic.
We have been using this picture all the time. It is basically updated every two weeks or every month, and some blue ocean words are discovered for layout. To what extent do we arrange? For example, when I searched for the keyword "foreign students cooking", I brushed two screens, and most of the posts were our content.
When users search for these words, they have already entered our circle, and most likely will consult us.
There will also be some highly transformative words, such as how to choose and recommend overseas study institutions. When users search for these words, they probably want to make a strong deal. It may be a very high-quality user.
2. Keyword layout
There are 7 layout points, including cover, title, copy, comment, account name, address and label.
For example, if you do some offline categories and do local traffic, if your address tag is marked with a property or a region, you will be pushed to people around the address tag.
The same is true for the account name. For example, you are a house buyer, and your name is Beijing house buyer. Your account name also corresponds to some search keywords. Here is a small detail. Your group chat name also corresponds to the weight of search keywords.
This is what we found unintentionally. When we search for and operate this category, some people may or may not find it.
In the above picture, on the left is a company that does the promotion and agency operation of Little Red Book. The content on the right has nothing to do with the agent operation, and the account has nothing to do with the agent operation, but his group hangs the brand agent operation media.
That is to say, the group name has entered the keyword ranking, which is a very small detail that you can use.
3. Keyword occupation
Here, I want to go back to the previous keyword analysis. For example, in the past seven days, "Foreign Students Cooking" has produced eight notes, so I only need to issue 10 accounts and update 10 posts, which is equivalent to killing most of the notes in the past seven days.
Therefore, the essence of keyword occupation is to exchange quantity for probability.
3、 How to efficiently create matrix number content
Let me start with my point of view: benchmarking is a lifelong thing.
Because there is no fixed mode template that can be used all the time, users will be tired of fixed templates and content formats within 2 months. Only by continuously paying attention to the data changes of the benchmarking account, including making some micro innovations, can new growth points be brought.
1. Establish content library (title/popular content/first picture)
We can use some data tools to find the rankings of posts in the past three months, make them into a content library, and update them regularly.
For example, Cicada Xiaohong, you can go directly to search the decoration category. The time of posting is chosen to be nearly 90 days, and the low profile articles are chosen. You can see some articles that have a relatively low number of fans during this period, but the posts are still popular.
If you are working in the category of teaching materials or the category of data numbers, you will find that the popular products are basically duplicate, and most of the popular content is in those formats.
Some friends ask: How to train a new person and make sure that he can start quickly when he comes?
In fact, you need to build a benchmarking database, including what are the hot money titles of the benchmarking account, what are the hot money contents, and where to find these data.
When a newcomer comes to do an account, we will ask him to do a calculation of the data of each post to figure out which posts have good data and which ones have high liking rate. For posts with good data, we will take them to do second innovation, and then synchronize them to other accounts.
For example, when we were working on a project before, a post content exploded, and all our other accounts copied the post content to do a second creation and continued to explode. After a period of time, it may not be popular, but another account may have content that is popular again, and other accounts will go back and wash their popular content again and again.
At the stage of testing the blockbusters, we have a 531 model.
When the number starts, we don't know what we need and what can really become popular. We will set about 5 directions to test 5 types of materials.
Once we find out which direction has exploded, we will focus all content production directions around this topic.
2. Establish SOP
Whether it is management team or content production, establishing SOP must be the basic disk of commercialization.
1) Give standards
Be sure to give team partners a clear framework and clear standards to tell them how to produce content.
After we run out a popular model, we will show the molded post of the content as soon as possible.
What is the picture like; What is the copy like; What tools are used to make pictures for title production and copy production; Where does the means of production come from; What kind of information... and write down the precautions very clearly, which prohibited words are untouchable.
2) Reviewed by
The content produced every day will be entered into a table and archived.
Let's take turns to arrange, for example, five people form a group, and each of these five people will go to review, do sampling inspection, when the posts were sent, and whether they were missed.
This part is very critical, because in the production process of a large amount of content, it often happens that the content is not distributed or not distributed.
3. How to create content efficiently
There is a very core key here, which is how to ensure that my content can do second innovation?
Look at the content of teacher Bobo's account. "The whole grammar system" and "12 pages of paper can understand the grammar system" are all around one material or one content in essence.
But different accounts need to use different styles to produce content. After each account has selected a production format, it will be fixed and kept running, but the content and materials, including the title, are being replaced.
4. Continuous review iteration
Although the iteration speed of Little Red Book is not as fast as that of dithering, it is also very fast. We need to keep learning in this process.
This is a table made by one of our partners to constantly learn which posts are better, what their methods are good, and what can be used for reference.
Even if we continue to review and iterate, we will always see what we can learn and what we can learn from the invincible teacher's knowledge planet, especially the 100 bloggers. This is a very important process.
4、 How to maximize the drainage efficiency
Xiaohongshu has a private message rule. The number of private messages to strangers on the enterprise account is 20 times/day. Note that strangers mean that the other party and you are not related to each other, and you are not friends.
There are two cases:
1. You take the initiative to touch the other party's private message
1) No limit on the number of two-way attention
When we create an account, we have a requirement for our partners that the number of followers must be greater than your number of fans.
Why? Because all the fans who follow you can send messages to them directly and without limitation after you turn them off. For all the friends who follow us, we will immediately return to the fans, send a short message, ask them what they want, and what kind of help we can provide.
About 5% - 10% will be retained. Generally speaking, 10 or 20 WeChat messages will be returned, and one WeChat message will be sent to the private domain.
2) Only like or favorite users, private messages will consume times
In many cases, the number of private messages may not be enough, and we will filter them in our account of clicking praise and collecting, and determine the purchasing power of each other through user portraits. For example, when we are doing a category, the payment ability of foreign and overseas users is obviously better than that of our domestic customers.
We will screen which customers are from Australia, Canada, and New Zealand. If they are confirmed to be overseas, we will send a private letter to each other to tell them what services we can provide.
2. The other party actively touches
If it is the other party's private message, we will reply within 5 minutes, which will directly affect the V rate.
We found that if your reply speed is fast enough, the micro blogging rate will rise to a very high level. The micro blogging rate is 30% different between 5 minutes' reply and 1 hour's reply. Our highest record is about 85%.
Although we have more than 40 accounts, in many cases, the categories we make are not as large as those of some data numbers or the teaching field in a day.
Because we are a user with high customer price and high net worth, the threshold for one-time payment may be tens of thousands or so, and the transaction cycle is too long. Each user is worth about 500 yuan, so every WeChat is very important to us.
The following figure shows the volume tracking process of Little Red Book made by our partners, which is mainly divided into three rounds.
The first round was to ask the user for contact information. If the user leaves V, we will let the sales to increase it. Sales and WeChat are successful, continue to communicate.
If WeChat fails, the sales team will immediately report to our operating partner that WeChat was not added. Our operating partners will tick or cross through our CRM system.
If not, our operation will also timely feedback to the user and ask him "if the teacher has been added, the teacher is adding you, and you have failed". In this way, users will be added repeatedly to ensure that our micro credit rate will not decrease.
After follow-up, we will give feedback to the sales teacher in the fly book group, and the sales staff will receive 15, 20, and 30 quantities every day. He will check all these quantities every day and give us feedback the next day. We need to follow up on which WeChat messages have not been added.
If the other party does not reply to the message, we will carry out one round, two rounds and three rounds to remind the user to pass WeChat. We usually follow for three days. If we can't add more after three days, we will give up.
After such a set of processes, the micro credit rate increased from 60% to 85%.
This is a script that we gave to our partners to drain SOP. This script is very critical for team expansion. It's OK to say that there are few team members, but after a certain amount, SOP is more important than managing ten or twenty people.
We have many private message reply scripts, including what the follow-up scripts look like; What is the speech technique in the comment area; Pay attention to users who do not have private messages. How should we leave WeChat and messages with them; If the user who left QQ doesn't leave WeChat, how can we let him leave WeChat; Some users have to ask about the price, and you have to quote to give the contact information. What is our script
The script is iterative, and new problems will be left in the database. Once a new partner gets started, he basically knows how to reply to users without asking.
5、 Management and establishment of matrix drainage team
In fact, managing the flow team is essentially managing the sales team. This is our fly book CRM system. We use multidimensional tables to achieve an automated process.
The form on the left is a form filled in by our sales and operation department. Most of the time, we input data directly through the computer. However, sometimes the user may come at night and may have time to turn on the computer. We will use the form on the right to enter. Just fill in the customer name and WeChat signal.
The advantage of multidimensional tables is that the corresponding sales can only see some clues related to them. For example, we have customer names, WeChat, mobile phone numbers, and advertising accounts in our database, which can be used to calculate performance.
For example, the sales team's commission is roughly 30% to 70%. The performance appraisal of the operation team is carried out by stages, so we can count it as an average. For example, if the number of clues is 500, the GMV is 300000 or 500000, we can calculate the monthly performance or commission through the average line.
Therefore, through the fly book form, we can present a performance market, directly see the single data and performance, and set up automatic sending to the group. For example, we completed several posts today and ticked several posts.
Our person in charge will know which posts are OK, which posts have not been sent, what is the odd number of daily performance, and what is the addition rate. We will send them to the fly book form in the form of Kanban, and this is free of charge. It is a free Kanban.
Finally, let's talk about the expanded content: how to do with account types?
The type of matrix number is divided into three parts.
1. Data sharing number
For example, teacher Bobo's English sharing account is the data sharing account, including some information about studying abroad for postgraduate entrance examination, which is called the data sharing account. Essentially, it shares a piece of information and pulls users into the traffic pool through a hook.
We have five accounts managed by one person, and we need to produce five pieces of content a day. The content production has a certain intensity.
We will determine the format, the format style of each account, including the content, how to write at the beginning, in the middle, and at the end.
Now after the chat GPT, the speed will be faster. You only need a script to tell it to copy this paragraph and make a replacement, and it can make a very fast rewrite of your content.
The picture is from Little Red Book
For 23 years, you must use productivity tools. Note AI is also very useful. If you are interested in chatgpt, I can recommend it.
2. Experience No
You may often swipe such account: XX's study abroad notes or XX's research life.
In fact, this kind of experience sharing is done from the perspective of a simple person. This type of account is very fast in volume, and there is no too much cost for the number.
The only cost is that when creating accounts, you should carefully consider the staffing of each account, set a general standard, not reveal the truth, and do a good job in management.
3. Official transaction number
The official transaction number is very simple. You can see some chat records or screenshots of transaction cases.
This form is also common in small red books. It is characterized by a small volume but a very high conversion rate, because people are very accurate in showing your transaction cases,
These three types of accounts should be determined according to your personnel arrangement. In the early stage, I suggest that you do the data sharing number and official transaction number first, because these two types of content can be mass produced, and you can quickly deploy accounts.
But for the experience account, you need to clearly determine the staffing of each account. You can't directly force grass planting. You need to write some soft articles, which have many skills.
Generally speaking, when the flow pool has reached the top, you can do a new expansion. For example, when you find that the repetition rate of clues is very high, reaching 30% or 40%, you can turn to experience number.
