|
In daily work, everything must be done in detail. In the Amazon battlefield, the competition is about the ability to carry out refined operations. Fine has at least two meanings: The first is to disassemble and subdivide, which is to break down a thing into smaller granularity, which is exhaustive, independent and operational. The second is details, which is to do things perfectly from the small details, improve the overall quality through local experience, and enjoy the sensory premium brought by the details. Today, let’s talk about Amazon’s “word selection” and how to select and use keywords carefully. This article elaborates on five aspects: keyword library, etymological objects, traffic structure, traffic stratification, and traffic generalization . Refinement of keyword library The first step to refinement is to refine the underlying foundation. For word selection, the underlying foundation is the vocabulary. What kind of vocabulary is an effective vocabulary? The vocabulary is composed of keywords that consumers search for. The solution is familiar to everyone, which is to use ABA data. ABA is the official backend data of Amazon, and its authority is unquestionable, but it has two shortcomings: First, if you use the exported words directly for filtering, your computer may not be able to handle it, because processing a file of more than 400 megabytes will cause you to crash. Second, if you reverse search by entering ASIN, you can only get the keywords that are in the top three in click rankings, and many keywords that are not in the top three in click rankings will be ignored. Based on the above problems, there are many reverse keyword search tools on the market, but the quality is also uneven. Here is a simple identification method: if the ABA data can reverse the keywords that the tool you use does not have, especially the big words, it means that the data of this tool has problems.
The refinement of etymological objects After determining the appropriate vocabulary, how can you further narrow the scope to keywords that match your products? A common method is to find similar competing products based on your own products, then organize the keywords of multiple competing products, and then select appropriate words from them one by one. This is easy to say, and everyone seems to understand it, but when you actually start doing it, you’ll find that there are still many details that need to be considered. For example, many products have variants. So when obtaining traffic words, what criteria are used to determine which variant is more appropriate? It's easy to think of using the best-selling variant, but it's difficult to do it. It is not just operations that are troubled by this problem. In the later stages of product selection, especially the production and stocking stage, knowing which styles are more popular is also very critical to cost control. There are many sales estimation software on the market, but basically they can only measure the sales of the parent entity. Especially now that Amazon has basically changed all categories to variant shared rankings, it is more difficult to measure the sales of variants. Since there is no direct method, is there an indirect method? Today I will share with you a new idea: indirectly confirm the sales of different variants by comparing the traffic words of the variants. Just like the picture below, sort out the traffic words of different variations under the same Listing, especially the natural traffic words (because the number of natural traffic words can better explain the situation of natural orders), it will be much easier to choose which variation. As for the advertising words, since the display of the product is linked to "profit" and is controlled by the budget, they cannot naturally reflect the degree of fit with the product. There is a certain degree of risk in choosing these words, because your competitors are still trying, and the effect is still unknown. Detailed traffic structure In the first picture comparing the variant traffic words, I also revealed another very important information - the traffic structure of the product. Some people like to try the whole package, including the whole set of advertisements (SP product advertisements, SB brand advertisements, SVB video advertisements) and the whole set of recommended traffic; some just try SP and SB; others are more conservative and only play SP; the last one is purely Buddhist and doesn't play any of the money-spending ones. After getting the candidate keywords, the question before everyone is how to allocate these keywords? How to plan the traffic structure? With a limited budget, how to balance the choices between SP, SB and SVB ads? These are all very complex issues, and each deserves its own topic. Due to time constraints, we are only giving you some ideas here, and we will discuss in detail how to plan your own traffic structure based on the attributes and development stage of your product when we have the chance. Refinement of traffic stratification If different types of traffic are the "horizontal" of the matrix, then reasonable traffic stratification is the "vertical" of the matrix. Specifically, each type of "horizontal" traffic will be composed of which "vertical" keywords, including main keywords and long-tail keywords, and the budget will be allocated and the rhythm of the strategy will be planned accordingly. We all know that after a product enters the mature stage, most of the traffic actually comes from a few main traffic words, generally no more than 5, but the proportion of orders is as high as 80% or even 95%. Sellers who have no idea about this data can purchase an ASIN report from a service provider. Most ASIN reports show this highly concentrated traffic distribution at the top (but it’s a bit expensive). This is the interesting part about all e-commerce traffic. Although the main traffic words make the absolute contribution, without the early preparation of long-tail words, there would not be the glory of the main traffic words today. Basically, everyone uses big words as the main traffic words. This kind of words generally represent a subdivided category or a subdivided category with specific modifications. Usually, there will be no problem in judging this. What I want to talk about now is precise long-tail words. Because long tail is easy, but precision is not necessarily. Precision requires ensuring that the modifiers attached to the root word strictly conform to the attributes or positioning of your product. Don’t say you know the attributes or positioning of your product very well. Everyone who has made a good product will be amazed at the various usage scenarios and methods of customers. It is important to stay humble. How to be precise in this regard? Experienced sellers may ask, aren't there many keyword mining tools on the market? Wouldn't it be great to use them? Isn't it a bit too much to spend all this effort? Sorry, this may not be so great. So what is the best way to do it? It is very simple. If we want to cross the river by feeling the stones, our opponents are our stones. The refinement of generalization strategy The entire process above is a process from broad to specific, from open to convergent. For the last point, we will "go the other way" and talk about the generalization of keywords. What is generalization? It is to obtain the traffic of related keywords based on the target keywords. In Amazon's system, it is divided into the expansion matching of related keywords and phrase roots. There is a progressive relationship between the two. Related keywords refer to the existence of two or more central roots for a product. For example, there are at least three roots in English to describe power banks, namely: charger, power bank, and battery. These roots will extend into many keywords, and even these roots can form new keywords, such as battery charger.
Second, when determining whether a keyword ad needs broad matching, use the data obtained in the first step as a basis. Group the broad matching words obtained in the first step, estimate the approximate total search volume of these words, roughly evaluate the possible effect, and plan the budget accordingly. If you cannot find a suitable broad matching word, the broad matching traffic is not accurate, or the broad matching traffic is too small, you can directly choose the precise matching method to place it.
|
