To make money on the Internet, we must first find a breakthrough point. After making money at this breakthrough point, we will start to screen and consider doing it in batches, and win by quantity. This is the method of mining profiteering projects to make money.
For example, in bonus hunter, when you register for benefits such as pulling new products and red envelopes, you earn less by yourself, so a large number of people are invited to join in and do it in batches, and some scripts or software are used to realize large-scale low-cost unit price profit.
What the wool party lacks is an opportunity, paying for new things, sending gifts by scanning the code directly, and taking a walk on the street, earning 55,000 yuan. It used to be called speculation, but now it is reasonable and legal. Whoever sees it first can stand at the top of the project.
Another example is video porters, who often carry YouTube videos to bilibili and then bilibili videos overseas. Once and for all, the accounts on both sides can do well, and they earn a lot of money by playing volume and advertising.
Most of the fans of Zhihu Q&A platform are working people, so some people specially designed a data download station for this kind of people, including various exquisite PPT, fonts, design templates, etc. It is also a very good project to take a member at will and take care of it by one person at the back end, and even rip off Zhihu’s high praise Q&A and change it to other platforms such as Xiaohongshu to earn popularity, and then drain it into cash.
Therefore, there will never be a shortage of items on the Internet. You have nothing to do. First, you are lazy. Second, you have no way.
Today, I taught you a way to find a project. This method is straightforward, and it can be regarded as a review of my experience in finding a project in recent years. I once said before that when doing a project, we must consider the market, and everything is empty talk without the market.
Then, the key to mining the project first is the user’s needs. First, let’s simulate the user’s habits. Generally, when looking for a project, everyone likes to search for such interrogative words as “what”, “how” and so on in the search engine, right? Let’s start from here and push back the behavior of the users.
We can choose 5118, search these query words, such as “what”, “how”, etc., and export the searched keyword information. 5118 needs members to export, so that we can get a huge amount of keyword information.
However, the number of this word is too large, and a lot of worthless information is here. We can only streamline and screen out some valuable information. This screening is more important. First, we should separate these long-tailed words.
To make money on the Internet, we must learn how to dig huge profits to make money.
Screening word segmentation
Here, we need to write a script in Python, and then use jieba word segmentation module to separate a complete word. For example, “How to do gum swelling and pain” will be divided into gums, swelling and pain, how to do it.
Moreover, when these long-tailed words are separated for the second time, the frequency of each word will be recorded, and a sequential table will be generated according to the frequency of these single phrases.
For example, mobile phones and English are displayed at most, so we can continue to search for “how+mobile phones” or “how+English” in 5118, so that we can get a more accurate long tail keyword.
In this way, if we search the words with high frequency and how to search together, we will get a long-tail thesaurus with great data value. Other words like love stand have the function of expanding keywords, so it is also possible to search in love stand.
So here we already have enough such interrogative keywords, because these keywords are still massive at present, so the next thing we have to do is to duplicate them.
We don’t need to analyze the degree of competition and search data, except for the keyword column, all the others are deleted, because we just want to find out these keywords.
Next, get rid of the repeated words, and we also get rid of the too short words and long words, just the normal length.
Get rid of the shortest words, because short words don’t meet our requirements, such as “what to do” and “how to do”, which have no practical meaning. Words that are too long, such as “you don’t open the lid of the cup, what is going on here?”, are useless phrases.
Finally, get rid of the garbled characters and pure numbers.
So at this step, basically a complete and accurate long tail keyword has been screened out by us. Then, after we save it to TXT file, we also need a processing method, that is, word vector text classification, to detect the similarity of texts.
This won’t work. You can Baidu the usage method and its usage effect. I’ll give you an example. For example, the words “how to make tea delicious” and “how to make tea is correct” can basically detect duplication.
When we get here, basically some similar words will be detected, and it will be automatically classified, so that we can see at a glance when we get this keyword list.
Just like how to convert PDF into JPG, this is the problem of conversion. I would never be interested in this in normal times, but with so many long tail keywords, it is necessary to study it.
Search on Baidu and find that there are so many bidding advertisements and such a small service, but there is such huge profits. This is a typical project of selling services to others.
If you chat with them from the customer’s point of view, you can almost understand the whole project process, learn their advertising skills, find customer service kits and learn the packaging of peers’ friends.
Of course, this is only one point based on this project. We can search on Taobao, WeChat, Zhihu and other platforms, and we will find more and more projects.
Therefore, when we do projects, we are never alone. Reliable projects need data to support them. Whether we make money or not, the peers have the final say, and the search behavior of our target user groups has the final say.