As we all know , doing evaluations on Amazon is one of the most effective means of promotion, but there are great risks in evaluations, so where do the risks of evaluations come from? How does Amazon analyze your operation evaluation, and what is its logic? How can I do the assessment safely? Let me give you a detailed analysis based on my experience.
First of all, let's take a look at how Amazon checks these non-compliant reviews. What is its logic?
We need to understand that Amazon's catching and brushing rely on its big data + seller's suspicious evaluation records + buyer's purchase history ( like a treasure, everyone's account will have its own label, which will record your shopping habits ). First of all, Amazon can use the basic conversion rate to leave reviews (for example, the conversion rate is obviously abnormal) , and then wait for big data to catch some sellers who obviously manipulate the ranking , such as catching seller ( A), then Amazon will ban seller ( A) at this time. Processing ( review these suspicious accounts, and then ban the account after confirmation ), and then ask the seller ( A) to write a complaint report, requesting a detailed listing of each abnormal VP and review. It means that you have to confess and strive for leniency. (If the seller does not confess, then the store is basically cold . If he confesses , the store will restore the sales authority. Of course, there are problematic reviews , and even the products will be removed ) . At this time, Amazon will get a new batch of data, which is Buyer A , Buyer B, and Buyer C provided by the seller . Then, according to the comprehensive analysis of the VP and review historical data of these buyers , follow the clues, and catch the seller B , the seller . C , then seller B and seller C are banned from selling , and the above steps are repeated again . This is called being associated, and one letter is a large piece. This is just one point of Amazon's risk control.
Let's talk about where the risks of evaluation mainly come from ?
In fact , the risk of evaluation mainly comes from these trading numbers . For example, if you find a buyer to help ten sellers to evaluate, as long as there is a problem with the data of one of them, according to Amazon's logic, the other nine sellers may be implicated.
There are two main methods of evaluation
1. Find a real person to evaluate
( 1) Find reviews on FB
Many sellers will look for real-person evaluations on FB . It is very difficult to find real -life evaluations on FB. Many foreigners raise their own trumpets to find "free lunches" , so when looking for evaluations, they must ask foreigners to post their profiles, and then keep them. The review account can be correct. Many foreigners show you the old account, and then use the small account to make purchases and leave reviews to reduce the risk of their main account. But for the seller's store, it increases the risk of the store!
( 2 ) The intermediary pretends to be a real person for evaluation
Many evaluation intermediaries will pretend to be foreigners to receive exemption orders, and use their own accounts to place orders . If you see that the Facebook of the reviewers is relatively blank, most of these people are usually intermediaries or trumpets pretending to be foreigners and looking for you to evaluate .
( 3 ) Looking for foreign students for evaluation
In fact, real foreign students are not short of money, and it is difficult to find them. Some people who claim to be foreign students are also pretending to be evaluation intermediaries . Generally, they ask you to transfer money through WeChat. Most of these are intermediaries, because they may not have a foreign paypal account.
( 4 ) Evaluation studio
There are many evaluation studios on the market, and they have their own batch of buyer numbers to help sellers do evaluations. Many studios I know either use a poor environment, use VPS or virtual machines to open more loopholes, or buy a batch of numbers for evaluation, the labels are confusing, and problems are prone to occur.
Therefore, it is too difficult to find high-quality reviewers on FB . British evaluators are relatively easy to find, and American evaluators are really scarce! To sum up, good real buyers are hard to find, and most of them receive a large number of orders. Referring to the above Amazon logic, the risk is very high.
The second is the self-support number assessment
The advantages of the real self-supporting number assessment are very large
(1) You can control the time and quantity of the evaluation by yourself, and you can measure it anytime, anywhere.
(2) The evaluation cost is very low, the cost of registering an account is more than ten yuan, and it can be used for a long time.
( 3) Don't worry about being refunded maliciously or using a black card to place an order
( 4) One device can support unlimited numbers, the mode is easy to copy, and a professional evaluation studio can be opened
The disadvantage is that it has high requirements for environment construction and technology, and the success rate depends on the environment you build. Our team has been researching and developing Amazon's assessment technology for 6-7 years, and has a very deep understanding of Amazon's risk control points . System cache, device name, UDID, iMEI, MAC address, IDFA , area code, regulatory code, motherboard code, WIFI address, fingerprint code, payment card and other dozens of parameters to read, once the parameters are associated, it will trigger Amazon's machine review, resulting in the cancellation of orders and bans.
Below I will give you an analysis of the following common evaluation environments:
Operation method: Install /Virtualbox virtual machine software Install in the virtual machine/Win7 image to register a 911 account, install and run 911 in the virtual machine Advantages: Easy to operate, low cost Disadvantages: When Amazon obtains virtual machine hardware information, there will be some special signs (For example, the hardware name of the VM will start with vm), which is easy to identify. 911 shares the IP of the computer room. If the same IP is evaluated by multiple sellers at the same time, it may trigger a high-level review by Amazon. The result is needless to say.
Alibaba Cloud , Google Cloud , Tencent Cloud, Huawei Cloud
Create a pay-as-you-go US cloud server to operate using Remote Desktop. Advantages: Pay-as-you-go, low cost, foreign real computer room Disadvantages: Pay-as-you-go IP needs to be changed frequently, not suitable for raising accounts. And when searching for IP, the IP number segment will be displayed as Alibaba Cloud, Google Cloud , Tencent Cloud and Huawei Cloud . DNS is easily exposed in China, triggering risk control.
Foreign VPS partition operation
Buy a large VPS server from some service providers , then divide it into multiple small VPS, and use virtual network cards to virtualize multiple IPs. Advantages: real computer room, fixed ip, can keep the account logged in for a long time, suitable for raising accounts The cost is high, and you need to buy a larger VPS server first.
4. iPhone+Luminati or srocks line
Why use IPhone, because IPhone is the best environment at present, because it will prohibit the APP from obtaining the basic hardware parameters of the device, specify the version, and use AWZ, AXJ and other one-click new machine tools, and the parameters need to be debugged and hidden. , before it can be used in Amazon reviews.
There are too many people using Luminati , it is only suitable for raising accounts, not for opening new accounts, and the purity is not enough. The success rate of placing the first order is very low, which can only reach about 6-70%.
The srocks dedicated line is currently very stuck, too many people use it, and there are many blacklist blacklist IPs. It is easy to be associated to lead to a machine review.
There are many commonly used environments on the market, such as: Virtual Machine + V2, Purple Bird, Endless Road, /Lincoln Sphere/Migratory Bird, Luminati/Lincoln Sphere/Migratory Bird, Poster System +911, lumit . I have said a lot in my previous articles. There are always various problems when doing Amazon evaluation. If the IP is not a public IP, it can be reused and easy to associate, or it is not suitable for raising an account. The fingerprint browser can only To solve the problem of association between cookie caches, the evaluation also requires debugging and hiding of device parameters.
At present, the IP with the highest degree of purity is the IP of the foreign independent home line . The IP pool has more than 5 million IPs, and it is very stable. It does not skip DNS and has accurate positioning. It can be independent and unrelated to each other with our environment and the use of fingerprint browsers. (Even if individual accounts are unstable and abnormal, it will not affect each other. other accounts).
After a lot of testing by our team, we need to do the following to pass Amazon's various risk controls.
The first is the computer side :
On the computer side, a foreign server needs to be built, equipped with a firewall, and then a foreign independent home line IP is built through a remote terminal , and finally a fingerprint browser is used with a privacy plug-in to run.
The mobile phone side needs a specified version of the jailbroken mobile phone + one-click new machine tool, and then debugs and hides the underlying parameters, and then matches the foreign independent home line IP.
The mobile terminal is responsible for opening a new account, placing the first order , and the success rate is higher. Later, it is moved to the computer terminal for batch operations, which is more efficient.
We have been doing evaluations for so long, especially for our own stores, the security of the environment, the account maintenance cycle, the number of orders, how to maintain accounts safely, how to control evaluations in stores, and how to open advertisements. The whole process and system must be Yes, it's not a blind test. At present, the success rate of our environment construction plan can reach more than 95%, and basically zero orders are cut. Those who have been banned for evaluation are basically outsourced to service providers or caught on third-party platforms. I have never seen a seller who raised an account by himself and killed himself. Doing the self-supporting number evaluation requires relatively high technical requirements, and it needs to have the ability to solve the risk control at any time. I will share it here today, welcome to comment and exchange, I wish you a great sale ~! Edit: BWM9992