Alibaba’s traffic methodology!

The first thing I want to share is a very mainstream traffic channel in the past – search engines, and see how we can still obtain traffic in this era of mobile Internet and new media.

The first topic is: Is it still worth doing search engine marketing?

The conclusion is: it is worth doing, but if you don’t understand the current ecological situation of Baidu in China, 99% of people will definitely have a poor ROI. As for Google overseas, it is still the largest source of traffic for most people.

First, let’s talk about why it’s worth doing.

What you may not know is that on search engines, more than 65% of traffic comes from mobile devices, so don’t think that search engines represent PC traffic. Baidu's DAU has not decreased and is still over 300 million. People also search on their mobile phones.

I like to divide traffic into "search traffic" and "social traffic", which correspond to "people looking for information" and "information looking for people" respectively. My point is that there is no priority between these two modes of obtaining information.

However, in China, Baidu's relative market share has been declining. First, Taobao, Qunar and others took away a lot of vertical search traffic. Then social platforms such as Weibo, WeChat and Douyin rose one after another. Baidu is no longer as important as before.

For several leading Internet companies we serve, more than a quarter of their total daily UV (including APP) traffic comes from search engines. The client cannot be used as a case study, but there is an example that can be found through public channels: during a certain period of time, Zhihu had a total of more than 40 million daily UVs, of which 13 million came from search engines.

The conversion rate of search engine traffic has always been the highest among all traffic channels. Because the nature of social media traffic is "browsing", while search engine traffic is active "searching". There are many examples of search engine traffic achieving a conversion rate of around 10% when the keywords and content match.

Then the search engine traffic is relatively stable and controllable, and traffic like SEO natural ranking does not cost money, which is why leading Internet companies pay more attention to this area. Not to mention them, those medium-sized Internet companies whose business model is to sell traffic, such as Rong360 and Tubatu, are representatives among them. They obtain more than one million UV free SEO traffic from Baidu every day.

Finally, there are still some bonuses that few people knew about before, which will be the focus of today’s discussion.

So why do most people who work with search engine traffic have a very poor ROI? We need to look at the so-called inside story from the perspective of natural ranking (SEO) and paid ranking (PPC).

Search engine marketing is divided into two parts: natural ranking and paid ranking. Natural ranking is non-advertising ranking, and the system decides who should be ranked first according to a certain algorithm. Paid ranking means buying ranking (PPC, or Pay Per Click). Your ad will be ranked higher based on how much you spend and whose ad the user prefers. Of course, in Baidu, the main thing is to see who spends more money.

For natural ranking, its payback period should be one year later, but many people expect good results in a few months; then, the entire industry, whether it is the client company or the contractor company, does not have very professional people, and the search engine marketing service provider industry is generally a reverse elimination industry. In addition, Baidu's own pages account for almost one-third of the search results, and the remaining traffic has obvious head effects, that is, only large websites can obtain good traffic; in such a bad situation, Baidu's judgment of excellent content is seriously out of control, and many people quickly obtain a lot of traffic through some cheating methods.

However, the method of paying for ranking by paying for bidding ranking is useless to many people. I personally feel that Baidu’s values ​​are not very positive. They used the wrong methods from the beginning to guide everyone to place advertisements. This may be a bit alarmist, that is, accusing Baidu of misleading everyone to spend more money for its own benefit over the past decade.

This is determined by the bidding ranking model. Only when everyone agrees that certain keywords have a lot of traffic will they compete for this word. The public decides which words are important or not based on their own guesswork. For example, anyone selling lipstick thinks that if they don’t bid for the word “lipstick”, they won’t have to attract search engine traffic. Baidu then goes with the flow and lets everyone use this mindset to advertise.

Specific example: I will give you two pictures, assuming that the customer is selling "Ejiao":

In order to give everyone an intuitive feeling, let me test you first: Suppose you are a merchant selling donkey-hide gelatin, and you go to Baidu to place an advertisement, what keywords would you decide to buy?

This is the keyword selection method that Baidu once trained its customers:

This method is to define some core words related to your business, such as donkey-hide gelatin, tonics, etc., and then define some prefixes and suffixes, and then arrange and combine them. The words that come out are:

Shanghai Ejiao Telephone

Beijing Ejiao Telephone

Shanghai Ejiao Website

…

Since everyone chooses words in the same way, competition will arise, and the prices of the few words that everyone can think of will definitely be raised.

This is the keyword selection method we have been using:

Here, we use a lot of data mining tools to first list what users in the entire industry are searching for. We must exhaust the entire industry, the more the better, and then divide them into different groups according to core words. At this time, the user's needs will be clearly listed in front of you.

From this data, we can see that the word "ejiao", which represents the entire category, has a daily search volume of only 8078. This is the case for all industries on search engines. 80% of the words searched by users have a lot of words. In fact, it is not difficult to understand if you think about it. Even a novice netizen knows that he cannot find what he wants by using words like "ejiao". Most searches on search engines come with clear needs, and most people's needs are personalized and unique.

However, when businesses placed advertisements, they fell into a misconception, thinking that the word "ejiao" was extremely important. In fact, the importance of a category does not mean that a certain industry keyword representing this industry is important.

For example, the search volume for the term "iphone11 price" far exceeds that for the term "mobile phone"; the search volume for the term "163 mailbox" is more than 100,000 a day, while the search volume for the term "free mailbox" is more than 3,000. The search volume for the word "novel" is not high, but the search volume for the word "TXT full novel download" is over 1.6 million per day.

Baidu did not correct everyone's misunderstanding, allowing everyone to buy keywords according to this mindset. For more than ten years, many people have been doing SEO according to this misunderstanding.

The second picture is the data I obtained from Baidu on October 28, 2019. The real searches for keywords related to "ejiao" on Baidu every day are basically here. No matter how hard you try, you can’t imagine that this is how users search. The data in Figure 2 is something that even some people who have been engaged in search engine traffic operations for 10 years have never seen before.

The traffic dividend I mentioned earlier is that the industry has been using the wrong methods to generate search engine traffic for so many years, so that even today, more than half of the traffic on search engines is still unwanted. I'll post a few more pictures and you can find the answer by searching on your own.

(The daily search volume here refers to the total traffic volume 30 days before the day of querying this data divided by 30 days, which gives the average daily search volume in the most recent month. Data is time-sensitive, so the latest data should be used.)

Here are some slightly older data.

Investment and Financial Management:

Automotive Industry:

The second topic: What kind of traffic world is search engine marketing?

A group of people have been doing search engine marketing for more than ten years. What kind of traffic world is this now?

First of all, search engines are the best proof of the “Long Tail Theory”. To generate traffic on search engines, you need to have a deep understanding of the long tail effect. Google once released a piece of data: among the billions of searches conducted on Google every day, if the words searched by the user are exactly matched, 15% of the search words have never been searched by anyone in the past year. For example, when searching for word combinations like “how to make an automatic door in minecraft”, 15% of the words searched every day have never been combined like this in the past year. This situation is not only the case with Google and Baidu, but also with Taobao, Xiaohongshu, and WeChat searches.

If this is the case, you simply can’t do something like “ranking for a few hundred words” because the words change every day. I have seen too many teams that actually just use dozens of keywords they came up with to do advertising and SEO. This situation was like this ten years ago and it is still like this now.

Many people have asked me how to find people in search engine marketing. My advice is to end the interview as soon as they ask you a question. That is to ask the other party how many target keywords they use when doing search engine marketing.

There are several levels:

• Less than 1,000 words—not yet started;
• 1,000 to 30,000 words—mostly trained through standard tutorials;
• 30,000 to 100,000 words—only at this stage can you be considered for employment;
• 100,000 to 500,000 words – This is acceptable for a very small business category.
• 500,000 to 5 million words - you probably won't be able to interview one such person in a year.

On search engines, finding 500,000 words belonging to a certain industry is definitely a powerful ability. In fact, it is not enough to just find 500,000. If you want to do a good job of placement, you also need to find the most cost-effective and freshest words. If there are people here who are placing search engine ads, you can compare how many keywords your company is placing in the background to know what stage you are currently in.

Search engines actually have a broad matching mechanism, but if you apply too few words, the broad matching mechanism cannot save you, and there will be a lot of words that do not convert and waste money.

Most people currently working on the scene are in the new media industry. Suppose you are in the beauty industry, your goal is to promote your products through beauty KOLs on Xiaohongshu. Suppose Xiaohongshu officially collects a statistical data of all beauty influencers in the background and there are 20,000 of them, and you only have more than 20 KOLs investing in you, but your competitors have more than 1,000. (Let’s ignore the operating costs of more than 1,000 KOLs here. Unlike celebrity KOL marketing, the operating cost of buying traffic for a certain word on search engines is very low.) This is exactly the same as placing an ad on a search engine.

Similarly, I don’t know if you agree or not, but we are still looking for big Vs to cooperate with. If we are doing performance advertising, they are the kind of people who used to buy keywords like “ejiao” on search engines. The ROI is very low.

Alibaba 's traffic growth team was established in 2002. The current level of search engine placement can be viewed on this website: https://www.alibaba.com/showroom/showroom.html

Please click on the letter A. Here are the keywords starting with A. There are 120 words on one page, and there are 22,429 pages under the letter A. There are more than 20 million words from A to Z.

This was my responsibility in the past, and I could make sure that each word had a certain amount of search traffic every day, which meant that they were all relatively important words. At that time, our catchphrase was "adding words equals increasing traffic." I remember that we bought more than 30,000 high-quality words from a foreign company that specializes in selling words, and then we completed our KPI for one quarter.

By the way, this method can also be used to make products.

For example, in the tourism industry, there was a well-known website at the time. We mined about 5 million words in the entire tourism industry, and then grouped and classified them as before, and we knew the demands of most users in this industry.

When the number of words you can analyze on a search engine reaches 5 million, you have basically obtained all the data, which can at least represent the demands of more than half of the people in this industry. This can be a great application of big data. Later, through this batch of data, we found that in the tourism industry, as long as you make guide-type products, you will not be able to survive, because the demand for guides ranks very low. As a result, we can see that some strategy apps are still struggling on the edge of profitability.

Then let’s look at the vocabulary data of the automotive industry mentioned above. Since 2015, “SUV” and “seven seats” have always been the real hot spots in this industry, but many automotive Internet companies still don’t know this.

Then look at a page on Alibaba: https://www.alibaba.com/showroom/mp3.html

Some attributes of MP3 are listed here, including "function", "screen", "charging time", "style" and "color". The attribute categories here and the items in each category are completely sorted from high to low according to the search volume of users in the search engine. The result of this is that the conversion rate and stickiness are very high, which is 4 times higher than the page made by the product manager at the time.

Many specific details can be shared at offline gatherings.

The third topic: How to do SEO and PPC well

The most important thing to do well in SEO and PPC is keyword mining. Without doing this, it is basically impossible to improve search engine traffic. Everyone should understand this scenario: on search engines, your traffic is spread through keywords as carriers. Without too many words to spread your traffic, how can the total traffic be high?

As for specific keyword tools, I will first introduce one we developed:

https://mp.weixin.qq.com/s/M1VCCFqNb9xSBnmD2hktFg

In an industry, the human brain alone cannot come up with so many words. This tool can help you automatically pick out many words. Lightyear Labs is developing a SAAS platform, and more keyword tools will be released.

One thing to note is that the word data must be timely, preferably from the last few days. As mentioned before, 15% of Google's search terms have never been searched in the past year. If the data is not timely, it is meaningless. Many words that had high traffic a week ago may not be the case now.

For many people, the most feasible way is to do PPC advertising on Baidu. If you can do a good job of keyword mining, Baidu is still a very profitable traffic channel for you.

However, there are some basic rules to follow:

1. In order to avoid not being able to recover your costs, the first batch of keywords you place must not have anyone competing with you.

On Baidu, if there is no one competing with you for this keyword, the click price can be as little as 0.3 cents.

So our initial strategy was to buy words that no one else had invested in. We would check whether anyone else was investing in any of the words we had invested in. If there was a competitor, we would not invest in them. What you may not expect is that even so, you can still submit a lot of words.

1. Optimize ad groups and landing pages

I personally think that search engines were the first to teach domestic Internet people about basic concepts such as conversion rate, landing pages, and repeat purchases. The mature Taobao after it simply further popularized it to more people. North America had its first wave of commercialized Internet companies in the mid-1990s. They studied many things earlier and more systematically. You can go to them to make up for the shortcomings, such as reading the most basic book "The Authoritative Guide to Landing Page Optimization".

Many Taobao training courses, since it is a closed system, are mostly quick-learning skills and not many systematic things. Optimizing ad groups and landing pages and making some scientific improvements may increase ROI several times.

As for those companies that cannot do even the most basic things, for example, their advertising staff uses less than 100 keywords, and traffic from any keyword is redirected to the homepage, etc., they should immediately shut down search engine advertising and expand current channels such as WeChat, Kuaishou and Douyin. Without professional placement, you can only give up this channel, but if you ask me, there is really a lot of traffic here that no one wants.

I once saw two Chinese Internet companies competing overseas. For the same APP, one’s customer acquisition cost was more than $1, while the other’s was more than $10. The second company always thought that it would cost so much, which directly determined the latter company’s elimination from this track a year later.

Unprofessional investment is bleeding the company. I have seen many unprofessional pitchers kill companies.

1. Don't look for outsourcing companies, rely on your own team to build your own traffic competitiveness

Especially those companies that operate accounts on your behalf. This statement may offend many people, but it is true. Traffic acquisition and user growth should be a company's core competitiveness. How can it be possible to rely on a third-party company to obtain this competitiveness? Search engine traffic outsourcing is a reverse elimination industry, and good companies have been eliminated. (Consulting companies like ours are only suitable for mature traffic teams)

For example, there is a traffic outsourcing company that is now listed. We once circulated their plan in the office as a joke. When their income comes from a percentage of the amount of money you spend on advertising, the more money you spend, the higher their income. Their interests are contradictory, so how can they have the motivation to make fine-grained optimization? Since Party A has too many accounts and most of these accounts are operated by interns, there should not be too many outsourcing companies.

Based on my limited entrepreneurial experience, products and traffic are probably the areas where company executives spend the most time. I've seen a lot of great user growth teams over the years, and they've largely made those companies successful. For example, I knew Zhang Yiming of Toutiao earlier than everyone else. When they were developing Kuxun in 2006, they achieved excellent traffic and even used a method that became a common practice in the industry.

At that time, Kuxun was doing train ticket inquiries. They generated hundreds of thousands of different web pages from a fixed train inquiry data by combining many station inquiries and train number inquiries. These web pages all hit a lot of traffic from queries such as "Shanghai to Nanjing train tickets". From the previous cases, we can infer that this type of traffic happens to be the bulk, and the traffic for too general words such as "train tickets" is not as high as everyone imagines, so Kuxun's search engine traffic is very good.

Even after this practice was copied by Xinxin Travel Network ten years later, it still attracted a large amount of traffic.

For detailed operation instructions, please refer to: http://huoche.cncn.com/

This method was later summarized as: when your industry has some data that can be queried from an existing database, please make it a web page, hit the search engine keywords, and then the traffic will be yours.

Later, someone turned QQ numbers, postal codes, mobile phone numbers, weather forecasts, etc. into web pages, which generated a large amount of traffic.

When Toutiao APP was first launched, it had 9 million installations in just over a month, a performance that could only be achieved by a team that is good at traffic. The early practice of the "Toutiao" APP to crawl content from more than 5,000 websites and then generate traffic is very similar to the site group methods used by early webmasters. Think about how the TikTok app itself became so popular. It is impossible that there was no traffic growth team doing a lot of work behind the scenes.

In terms of SEO, we should make full use of the fact that Baidu gives nearly one-third of the traffic to our products. Your content should be included in all the Baidu Q&A, Baidu Tieba, Baidu Wenku, Baijiahao and even Baidu Netdisk. When you have a vocabulary library like the one above, it is full of users' search needs, and it will tell you which ones have high and low search volumes. Then, to spread content on channels like Baidu, you only need some common sense and just figure out how to put your content there reasonably.

The most recent traffic bonus is Baidu Mini Program, which, like the previous Baijiahao, is given priority ranking. I personally think this is Baidu’s last wave of dividends.

Baidu also gives high weights to many large sites, such as Zhihu, Sohu Self-Media, CSDN (you read that right), etc. This is also the trend of Google overseas. This is not because the search engine is biased, but after a long period of evolution, more good content creators will gather on it. We can put content on it and convert users.

Everyone should give up the idea that they must use their own website to generate traffic. Traffic will definitely be increasingly concentrated at the top. The idea of ​​turning your own website into a traffic center is becoming increasingly unrealistic, especially for small sites that have no advantages. Therefore, you can rest assured to be a content contributor on other people's platforms, contribute content on these platforms, and then take these pages to any other platform to drive traffic. The method I have always said before is to "let the content flow in the middle of the channel."

Let’s take “Ejiao” as an example. You can do it like this:

From the above vocabulary, we can see that "Who is suitable to eat donkey-hide gelatin cake" has more than 20,000 searches a day. Go to Baidu and search for this word. I am searching in Hangzhou now and found that no one has placed an advertisement. (It is not ruled out that it may be available elsewhere or when you look at it)

At this time, you should have a soft article appear on channels such as Zhihu, Baijiahao, Sohu self-media, Baidu Knows, Baidu Mini Programs, etc., recommending your product. After doing this, you will have to compete for rankings with the current web pages ranked at the top. With some simple SEO techniques such as "anchor text bombing" or "external links", or even the most primitive "keyword density", your chances of ranking up are very high, and then the traffic will be yours. Of course, in the end, if this word is discovered by others, it will also become a red ocean, so your way of dealing with it is to discover more other blue ocean words, or just grab them directly since there will be few people competing. Also, if you answer the top three Baidu questions in the picture now, you will still get a lot of traffic.

After so many years, Baidu still has a lot of traffic that no one wants. Baidu is actually killing itself in this regard. You see, although Baidu let many people buy keywords at high prices and earned a lot of revenue, a lot of Baidu's traffic was not sold. This means that there is still a lot of traffic inventory, the traffic has not been fully commercialized, and advertisers have not obtained a good ROI from high-priced traffic. This is a lose-lose situation. Toutiao has done much better in this regard. Its information flow is more concentrated than Baidu's. Toutiao's massive engine is now beginning to develop many functions to allow everyone's delivery to be staggered as much as possible to avoid collisions.

If you have an APP with a lot of content, you must find a front-end engineer to convert the content into a web page and do basic SEO, which can help you get a lot of free traffic. For example, in my opinion, Toutiao’s SEO is not good enough and its content is repetitive, but it should not be a problem for it to get millions of daily UVs on Baidu.

There are still many techniques that cannot be explained in a short time. There are dozens of factors that affect SEO rankings, and none of them have been mentioned now. But these are not important. The three mentioned just now are techniques that can be found with just a simple search. Over the years, there have been both technical algorithm factions and strategic factions in the field of traffic, and in the end the strategic faction always wins. This can also provide some inspiration for those who are working on Douyin, WeChat, and Kuaishou.

By the way, if you are confident that you can recover your investment in the term "Who is suitable to eat donkey-hide gelatin cake" with a PPC cost of 0.3 cents, you can directly buy a round of traffic first.

Fourth: Traffic methods based on keywords and relationship chains

Next, I will share with you a traffic system that I have accumulated over the years. This is what I want to share most, and it is the source of many methods.

In my personal opinion, there have always been only two carriers of traffic: keywords and relationship chains. Traffic transmission in many channels is one of these two basic carriers, or a variation or mixture of them.

This is the methodology we used in the Alibaba traffic team. Its derivation is as follows:

Behind the traffic is our common need for information, whether you are looking for a joke, reading an article that teaches you how to cook, or just scrolling through it out of boredom.

The mode of information dissemination that humans have adopted so far is either through text or through relationships between people.

This was the case in primitive society when there was no written language. For example, when a tribe discovered prey and wanted to hunt it, the information would be transmitted through the language that had long been formed and the network of relationships between people.

What if we go back to more than 20 years ago when there was no Internet? It's the same.

I was born in the 1980s and lived through an era without the Internet. Let me give you an example. At that time, I had a relative who worked in a shipyard. He heard about an underwater welding technology, but he didn't know how it was implemented specifically. If you face such a situation today, you can find out by searching.

But in an era without the Internet, he also had two ways to solve this problem: one was to go to a library as large as possible to look up information. In the past, an important function of libraries was to be a place for people to look up information; the other was to ask his master or other friends and find information through their network of relationships. Interestingly, the Google search engine was born in 1996 when its two founders built an electronic library for Stanford University.

When pictures, texts and videos are spread, they must be indexed with keywords. For example, the algorithm of TikTok is to label each video and each person. The labels are keywords, and then the keywords of the person are matched with the keywords of the video. This is a variant.

Let’s first look at the traffic channels based on keywords.

It goes without saying that search engines, Taobao and APP stores are keyword-driven traffic channels. Even WeChat, a traffic channel based on relationship chains, has a very large amount of search traffic on the first screen, and the so-called WSO (WeChat search optimization) has also had some successful cases. In the WeChat mini-program search, if you search for "stocks", there are thousands of mini-programs with stock titles, but if you search for "stock price", there are only 3 mini-programs with stock price titles. As long as you are the fourth one, you can still attract many users every day. In the entire stock and securities industry alone, there are at least tens of thousands of such words that no one pays attention to.

In the United States, an organization has counted all search traffic and all social traffic, and it is a very coincidental 1:1 ratio. As mentioned before, there is no primary or secondary distinction between the two modes of obtaining information: "information looking for people" and "people looking for information". Here is the first data proof.

For those of us who are trying to acquire traffic: channels with a lot of people should attract more people, and channels with a lot of words should occupy more words.

I will use an example to open everyone's perspective on using keywords to generate traffic. There is a company that does home improvement internet business and acquires customers in this way: they collect thousands of the most popular keywords in the industry on search engines, buy more than ten QQ numbers, and each QQ number can create up to 500 groups. The names of the large number of QQ groups they created are all one of those thousands of keywords. There is software that can maintain the activity of groups in batches, so when QQ users use these keywords to search for QQ groups, their QQ groups are ranked relatively high. In this way, more than ten QQ numbers can attract about 8,000 to 10,000 QQ numbers to join their group every day, and then use the WeChat private domain method to convert users.

The traffic carried by keywords can be summarized into such a methodology: When doing traffic in any channel, you can observe whether some of the traffic in this channel is distributed through keywords. If so, go and grab the traffic using keywords as carriers.

As for the source of keywords, Baidu is a very popular search engine. As long as the keywords are hot there, they will most likely also be hot in the channel where you want to generate traffic. You need to occupy a large number of keywords in your industry. Our traffic team at Alibaba used to call this method "keyword occupation."

For example: WeChat Good Product Circle, this is a channel for distributing traffic through keywords; if you occupy a group of circles named after good keywords, you will naturally gain a lot of traffic.

For example, Feichat is also a channel for distributing content through keywords. If you think this platform will rise, you must create many groups with titles that include popular keywords in this industry. If the groups become popular, the traffic will be yours. Currently, Feichat has several million DAU.

For example, on an algorithm-distributed platform like TikTok, if your videos are tagged with many popular industry keywords, it is not a problem to get a lot of extra traffic, not to mention that TikTok is searchable. If you search for some popular keywords, you will find that many of the videos at the top have a lot of traffic, but they may not get the traffic by having great content, but simply by having the tags hit the popular words.

You just need to be certain of one thing: as long as humans continue to use text to spread information, the method of "keyword occupation" will always be effective.

Even if you spend money to buy traffic channels, there are still channels where keywords can play a role. For example, the background of Toutiao's massive engine is open to targeting the crowd with keywords.

Regarding the delivery of information flow, many times when we deliver information flow, the time points are very close, and the population attributes may be 25-45 females, high-income age and so on. Everyone's targeting conditions are actually highly concentrated, so advertising display depends on price increases. But if we use keyword targeting, there won’t be such a serious collision, and we can reach many users that we can’t reach normally. The platform traffic can be fully utilized, and the advertiser’s ROI will be better, which is a win-win situation for all parties.

Regarding the traffic of the relationship chain, I have heard many people share many good methods offline, especially content-driven and private domain community methods. Sometimes I feel very emotional. A lot of the content you are creating today was produced by a group of people who had the same mentality and state as you more than ten years ago. Only the platform has changed, but a lot of the content discussed is exactly the same. In the past, Taobao used the mentality of looking up to other platforms (163/NetEase/Sohu) to attract traffic. Today, there are so many merchants who use the mentality of looking up to Taobao to obtain traffic. It is very interesting to think about it.

Unlike the way of playing with keywords, I think the way of playing with relationship chains in China is far ahead of that in foreign countries.

However, there are a group of foreign peers who are also engaged in social relationship chain traffic. They systematized the social relationship chain methods earlier than us. How did they do it?

Let me add here that I think platforms like WeChat public accounts are actually copied from Facebook. This is a very alternative point of view, and I would like to take this opportunity to talk about it.

Let’s take a look at a screenshot. Facebook launched a product called Facebook Pages in 2009, such as: https://www.facebook.com/pages/ (need to climb over the Great Firewall)

Businesses, companies, public figures, and communities can create a page for themselves, which everyone can follow. If the business releases any news, it can be updated on your information flow. This has the same positioning and functions as the WeChat public platform, but it was launched more than 4 years earlier. Many of their businesses are also working hard to increase their followers when doing marketing on Facebook. For example, Amazon’s Facebook Pages has nearly 30 million followers.

In 2013, the WeChat public platform with its current positioning officially came into everyone's view. Therefore, if we learn from and absorb the methods of our foreign counterparts, there are some interesting angles.

Basically, the methodology at that time was this: if the carrier of communication traffic is the relationship chain, then optimize these four points: namely, the number of people who spread information, the influence of people who spread information, the speed of spreading information, and the reputation of the information spread.

The speed of information dissemination in China has been maximized, and I believe there is no need for me to repeat those methods. Then the people who create various account matrices have also done a very good job of increasing the number of people spreading information, such as the traffic operation teams of many celebrities on Weibo.

The harsh domestic competitive environment has given rise to all kinds of strange tricks. I will share a few of our ways of generating social relationship chain traffic, and see how we build a domestic platform based on foreign social traffic methodology.

Regarding the “number of people spreading information”, I’ll give the example of the Kuaishou APP.

The official logic of Kuaishou APP is to distribute traffic evenly, that is, when you open Kuaishou, some people near you will be recommended to your Kuaishou homepage. In order to cater to this distribution logic, we must do multi-account distribution. For example, live streaming works like this: We developed our own multicast system, which looks like this:

Its video signal is captured by a camera. After the video content is input into this system, the video can be distributed to countless other Kuaishou accounts. Each account simulates a different geographical location, so that only one live broadcast content can appear in different places across the country.

Of course, it can also broadcast the same video signal simultaneously across platforms such as Kuaishou, Douyin, Tencent, Douyu, etc. Kuaishou doesn’t like the same content, so the background of the picture taken by this camera is a green screen. This system can render different backgrounds in real time, so that it looks like different content. (No illegal operations such as cracking the app.)

In the social relationship chain communication method, the environment is very bad and you are forced to do this. Everyone present should know this because others will plan more aggressive things. Some big companies that we all know also have many similar operations in private. With other conditions being the same, if the number of people spreading your message is dozens or hundreds of times greater than yours, you will not be able to rely on content-driven growth with peace of mind.

In many successful cases of growth hackers, as well as the successful products we see on the market today, there are a lot of examples of winning by relying on numbers.

So how do we achieve “the influence of people who spread information”? We have created such a system. The principle of this system is similar to that of "Qichacha", which is to create an aggregate crawler (legal crawlers crawl data from different data sources and aggregate them together). The result of this is: specifying any industry, you can find thousands of high-quality KOLs in this industry every day, and have their personal WeChat QR codes and other contact information.

The following figure shows the data captured every day in the financial field:

Each KOL will be scored according to various indicators. If the score is lower than a certain value, the contact will not be considered for the first time. But there are still so many KOLs every day, so how can we get a lot of contact? We also stole the laziness and used corporate WeChat to contact us in batches.

Since it is not a problem to add tens of thousands of people to a day on the Enterprise WeChat, we have built a group control system for Enterprise WeChat, which can add these people as friends in batches, and then communicate with each other in groups, and then use them to siftly sift it over again and then communicate manually. When others are batch "washing" end users, we are batch "washing" KOLs in the industry.

The following is a demo version of our WeChat group control software, similar to Enterprise WeChat:

Most people don’t pay attention to the functions of corporate WeChat. Everyone must pay attention to this corporate WeChat and it is very useful. After a while, we will launch a corporate WeChat tool.

Many people don’t know that the official WeChat of the enterprise has a mass sending interface . Don’t use those fake mass sending functions anymore. Then, adding people every day will not be easily blocked. There is no limit on the total number of people added. And can you believe that Tencent’s official has developed something like “live code” in the enterprise WeChat?

From a strategic perspective, if Tencent official specifically hopes that you can use corporate WeChat to create a private domain, you will not have the truth that you can use it quickly, and it will only adapt to everyone's needs more and more. Just note that there is no circle of friends for corporate WeChat, and the functions of WeChat group are also limited (there are reports that they will be released at the end of November), so we must explore a new set of traffic gameplay that is different from the private domain operation of personal WeChat.

This gameplay is also particularly suitable for cold start of products. What everyone lacks most when cold starts is seed users. This method of aggregating crawlers can quickly obtain the users you want and thus successfully start the cold start. We have had many successful cases of cold start.

Speaking of this group control system, we have made two systems with different technical principles. As we all know, the group control system based on xposed was severely cracked down on WeChat during 618 this year. We used to think that the group control solution of xposed is too invasive. If I were WeChat official, I would really dislike others changing my APP in a mess. The two systems use other solutions, just use it instead of manual labor, just don’t crack and use it illegally. At present, both solutions are relatively safe.

Sometimes we can create a relatively large advantage by using these systems, but the more we do this, the more we will find ways to do the content. After all, we come from big platforms, and we don’t want to create content garbage on any platform for traffic demand.

In order to prevent everyone from being deviant, I will say something more. For many methods of doing traffic like this, it seems aggressive, but there is a bottom line in doing it. It should be like Alibaba's internal style of doing things, that is: the original intention must be positive, but the behavior must be wolfish.

The Internet industry we are engaged in is an information industry, so we must "transmit high-quality information" as much as possible and make the content as good as possible. But when distributing these contents, you must do everything possible to spread them.

Finally, let’s talk about the “reputation of spreading information”. Our country really does not pay attention to this, and our colleagues in Europe and the United States attach great importance to this and do a very good job in this area.

For example, I went to Geneva, Switzerland to train a company in 2011. They had a business that specializes in helping the luxury watches in Switzerland spread on Sina Weibo. One of their jobs is to analyze all the comments from everyone on a Weibo account and use Excel to count them all. It is divided into positive evaluation and negative evaluation, and positive and negative evaluations can be divided into several dimensions, which angles are positive and which dimensions are negative.

What surprised me was that all this was actually operated by human flesh at that time, and the person who operated this only learned Chinese for 3 years. In order to understand the content of the text, he had to translate the dictionary.

Social media is the most important thing to pay attention to reputation . Now that we have the concept of depth of communication in China, the reputation determines whether your potential energy increases or decreases by several tens of percent when each layer of transmission. Calculated according to mathematical formulas, the subsequent results are very different.

That’s all for today’s sharing, I hope it will inspire you. Finally, I suggest you watch a movie called "Social Network", which tells the history of Facebook's founding. The first ten minutes of the movie actually talked about Mark Zuckerberg's method of growing hackers. If you can understand it, it is very interesting.

In fact, the growth hacking culture I know comes from Facebook. There are many hidden ways to obtain traffic that cannot be said in Facebook's history.

Author: Guoping

Source: Light Year Laboratory (ID: gnlabs)