Fortune Telling Collection - Zodiac Analysis - What are the classifications of crowd labels?

What are the classifications of crowd labels?

The classification of crowd labels includes buyer labels, product labels and store labels. 1, buyer label: gender, age, constellation, region, Taobao account level, Taobao reputation, etc. 2. Product label: released categories, attributes, keywords, materials, styles, etc. Forming the initial crowd label of the product. 3. Store label: The label of the store baby is combined with the buyers who browse, collect and buy in the store to form the label of the store.

1. What are the classifications of crowd labels?

1, buyer's label

(1) Specific attributes of buyers: gender, age, constellation, region, Taobao account level, Taobao reputation, etc.

(2) The buyer's recent search, browsing, collection and purchase, baby purchase records, etc.

(3) Buyer's order attributes: customer unit price, category preference, style characteristics, purchase frequency, purchase channel, etc.

2. Product label

(1) Baby release: released categories, attributes, keywords, materials, styles, etc. Form the initial crowd label of the product.

(2) After the baby is put on the shelves, the specific attributes of the initial buyers who browse, collect, purchase and purchase have already formed the baby label.

3. Store labels

(1) The store baby label is combined with the buyers who browse, collect and buy in the store to form the store label.

(2) When the store's baby label and the buyer's label overlap, it is commonly called intersection; Such a system recommends the most accurate labels, and such free traffic is also the largest.

Second, how long will the Taobao crowd label be restored?

In fact, there is no uniform regulation for this prescription.

But after a long time, for example, more than three months, it will definitely not work. Generally speaking, new products will be labeled two days before they are put on the shelves.

Therefore, we should always optimize the crowd label. If you don't label it for a long time, it will disappear. If you don't optimize them often, the labels will get mixed up.

Third, how to make accurate crowd labels?

People can be divided into five categories: region, gender, age and purchasing preference. According to users' purchasing preferences, people can also be divided into regional preferences, price preferences, category preferences, use preferences and quality preferences.