Breaking news! A9 algorithm failed, the new Cosmo algorithm caused most sellers' sales to plummet!!!

Some time ago, everyone was discussing the failure of the A9 algorithm. Amazon launched a new algorithm, Cosmo, which brought three revolutionary changes!

In fact, to put it simply - it is a simple version of a thousand faces

Amazon uses a search method similar to "one thousand faces for one thousand people", which is to recommend different products based on each person's different purchasing habits and search habits. The recommendations are more refined and the products are more in line with consumers' habits.

The core function of COSMO is multi-round navigation. Here, COSMO excels in providing multi-round search refinement through continuous recommendations. For example, a search for "camping" may lead to the selection of "inflatable mattress" and then refine to "inflatable mattress for camping". COSMO then provides various types of inflatable mattresses for camping to meet different needs, such as lakeside camping, mountain camping, or 4-person camping. This multi-round navigation allows for deeper and more precise refinement, reflecting a natural discovery process and significantly enhancing the user's search experience.

So what is the specific analysis of the thousands of faces of each person?

The full text is divided into 5 parts

1. Similar people push - offline stores

2. Low-price spiral strategy: Why can prices continue to increase?

3. Coupons are used to increase conversions, not to regulate prices.

4. Product premium is determined by the market, not differentiation

5. Some issues about product selection, you can think about them

There is a rule in Amazon that everyone knows, that is, do not change the price of products casually. You can lower it, but if you raise it, it will immediately cause a decrease in product conversion. Another famous rule is why prices can continue to rise in the spiral strategy. Let me reveal the secrets one by one.

1. Similar people push - offline stores

The search algorithm is different for each person. I don’t know if you know this. Simply put, the pages that each person searches for are displayed differently. Why is this so? It’s because Amazon has labeled you. The most important thing in the label is the price range. In other words, Amazon will classify users according to their shopping habits and then push the pages they like.

What is similar group push? It refers to users who purchase products. Amazon will push similar groups to this product based on the characteristics of the product's conversion group, whether it is advertising traffic or natural traffic. Therefore, many people raise prices when they see that sales are good, and then sales immediately drop in half, and traffic can't go up. This is the low-priced group that Amazon pushed in front, and your product price has quietly increased, so the conversion rate and click-through rate have dropped, and Amazon will not push traffic to you.

It should be pointed out here that price is a range threshold rather than a single value . For example, if a product priced at 23.99 increases by 0.99 US dollars, you will not feel it at all. But if you increase the price by 3 US dollars to 27.99, then the conversion rate will be very obvious. This is the customer's price acceptance threshold. The higher the price, the less obvious the price increase.

2. Low-price spiral strategy: Why can prices continue to increase?

Why can the low-price spiral strategy slowly increase prices? That’s because the price threshold range is relatively small. Generally speaking, products suitable for the low-price spiral strategy are those with low unit prices. The distinction between the 5.99-9.99 population in the population positioning is not very obvious, so when the order volume increases, appropriate population screening can be done. However, it is impossible for a product that costs $100 to start a spiral from $50. This is not a spiral, it is stupid.

3. Coupons are used to increase conversions, not to regulate prices

If you don’t want to push similar groups of people due to incorrect crowd tags, you can completely control the crowd through coupons, let Amazon identify similar groups of people through prices, and increase product conversions through coupons. This is completely possible. Many people feel that coupons are not as obvious as direct price cuts. That is for sure, because when pushing to similar groups of people, low-priced products will be pushed to more people.

4. Product premium is determined by the market, not differentiation

Many people want to increase product prices to increase profits immediately after differentiation. I have talked about this issue before. Product differentiation is to increase product conversion rate, not profit margin. Profit margin is determined by market supply and demand. It is obvious that there are two variants. After differentiation, the other product wants to increase the premium by 3-5 US dollars. Of course, if a bottleneck technology or aesthetic is broken through, it is completely possible, but whether this technology or aesthetic is broken through is not decided by us, but by the market. The simplest thing is the product's click-through rate and conversion rate, which are the best feedback from the market, so don't be self-satisfied and think that your aesthetics and technology surpass everyone else.

5. Some problems when selecting products, you can think about the underlying logic

Many people are keen on buying product selection software. In fact, the only problem that product selection software can help you solve is to collect and analyze huge amounts of data. However, you still need to have the underlying product selection thinking yourself. After using a lot of software, the products selected are still exactly the same as others. You buy goods at 1688 and clear them out a month later. This is the case for many people.

There are also some people who are very keen on asking for various product selection forms and various product selection dimensions. Every time I see many people asking me for a lot of forms, but when you actually use these forms, what is the point? You have made a lot of forms, but you still have no idea whether the entire product can be made.

More people also like to collect various product selection websites, such as some websites that specialize in collecting new and unique products, such as some websites that specialize in collecting maternal and child products and toys. To be honest, I also have a lot of these websites. There are also some store selection methods, cross-platform selection methods, social software selection methods, BSR selection methods, new product list selection methods, supply chain selection methods, and logistics data selection methods. There are 10 to 20 kinds of selection methods in total, but the selected products still cannot be sold. Why? Product selection is not just about what others choose to sell, but also requires a lot of things to look at. It is not a simple cut and paste copy, and 1688 takes the goods to sell.