Application Of Data Analysis In Apparel Inventory Management

2011 is

In April, the author analyzed the operation of several listed sporting goods companies such as Lining, Anta, XTEP, PEAK, and so on, according to the earnings report. In this year, the average stock price increased by 41.2%, and the number of stocks (DOS) increased from 28 days to 41 days.

The stock of red bean and Mei Bang clothing accounts for 60% of the total assets and 29% respectively.

2012 is a big challenge for garment enterprises, and the first thing to do is inventory management.

The inventory of clothing enterprises is an eternal topic. Agents and manufacturers have been fighting for wisdom.

The reasons for the large inventory of enterprises / agents are:

1. the sales target is extremely high, so the enterprise has to rely on pressure stock to meet the requirements.

2. order is based on feeling, excessive pursuit of explosive money;

3. the stock between agents / stores can not be pferred normally, often in the A area, but there are a lot of stocks in the B area.

This may be the cause of the mechanism. The mechanism of factory dredging is missing, and the goods are sold to the agents. The second possibility is the reason of the agents, not facing the reality.

I had talked with an agent and asked him why he did not sell the goods to the neighboring provinces. His answer was that the goods were sold so well, and I had no reason to sell them badly. I have made plans to improve the shops from all aspects.

4., there is a lack of forecasting mechanism. Replenishment is always out of stock.

This article does not talk about the big reason, but from the data analysis point of view to tell you some inventory management methods.

First,

Data preparation before

For an agent, it is necessary to determine the following data before ordering: 1, how much amount / piece of goods to buy, 2, what kind of goods to buy, 3, determine the commodity level and the number of corresponding orders.

In addition to 1, the other two are strategic plans before ordering.

(Note: the following analysis does not take into account the change in the number of stores).

1, generally speaking, the amount of the purchase is confirmed by the manufacturer. The agent should do the calculation of the corresponding quantity.

The formula is as follows:

Quantity of merchandise = purchase target / (regional sales average hang up price X (1 + planned price increase))

2, determine the amount and quantity of goods purchased: determine the proportion of the order season according to the percentage of sales of commodities in the same period. The revision principle is the commodity trend and operator's strategy this year.

This link is very important. It is a landing link for a shopping strategy.

3, determine the commodity level and quantity: we can adopt the combination of average and extreme value.

According to historical sales data, the commodity money can be divided into five levels (or three). The corresponding relationship between the level and the purchase quantity is as follows:

With this data preparation, in the process of viewing goods, buyers only need to determine the level of goods, without considering the quantity of goods purchased.

The advantage of this method is to reverse the purchase quantity according to the sales data and avoid the blindness of buying goods.

The process of buying goods is actually the process of selling goods, so it is worthwhile to spend more time on preparations.

In the process of cargo selection, the percentage of 2% and the commodity level of 3 can be adjusted according to the specific circumstances.

After buying the goods (before placing a formal order), it is necessary to detect the goods purchased, mainly testing three aspects: price segment, size, color and so on.

The data were compared with the sales data in the same period.

For example, the loss of price segment can be clearly seen.

Two. Data analysis in sales process

Clothing industry data analysis from the point of view of commodity flow is that we are well known to enter, sell, store, and the sales link can be refined into people, goods and fields.

The reason for excessive inventory is generally considered to be too much or too bad to sell. Few people work hard to track, predict and analyze the commodities. This section focuses on this aspect.

1, how to establish a tracking system for commodity data?

Commodity tracking is usually implemented in days or weeks. With the popularity of the POS system, data collection is becoming more and more timely. It is possible to track sales of goods by day.

Tracking: store inventory, sales volume in the past 4 or 8 weeks.

It can be shortened to 1-2 weeks) and large warehouse stock.

Tracking direction: if we can not cover the whole product chain, we can follow up the top 20 stores and the top 20 stocks in the big category (such as sports shoes, clothing, matching, women's clothing upper, lower, suit, accessories, etc.), the former is to sell better products, the latter is to make the inventory of large commodities digest as soon as possible.

New product digestibility is also a data that must be traced. Many clothing companies are updating this concept this year. What is new product? As long as consumers do not cross it, they should be new products.

This idea is helpful to the digestion of inventory, especially those enterprises that rely on new products more and more.

After collecting basic data through POS, the rest is to set up a tracking model with EXCEL, so that it can automatically generate analysis conclusion every day or week (the strategy part needs to be artificially formulated), and then build an automatic distribution / allocation model according to the logic of stock days.

The purpose is to monitor the main products purchase and sale status, replenish goods quickly, and let the goods flow among different customers or stores.

Under normal circumstances, do not let shop clerks come down monotonously, because clerks may be too busy to forget to place orders, and for good selling products there will be a habit of "inventory".

2, how to make sales forecasts for goods?

The weekly inventory (WOI) is a KPI value of the commodity forecast, the stock week = the immediate inventory / (sales volume / sales cycle in the cycle), the sales cycle can be 4 weeks, 8 weeks and so on.

For example, a product has 2000 stocks at present and 1000 items sold in the past 4 weeks. Its stock week is 8 weeks, meaning that according to the recent 4 weeks' sales situation, the goods will be sold out in 8 weeks.

It should be noted that if the sales cycle takes 4 weeks, if a commodity is a new product that is listed two weeks ago, the sales cycle should be changed to the actual sales week.

Sales volume tends to miss this factor.

It is a relatively rough prediction method to predict the goods according to the weekly inventory, because it does not take into account factors such as seasons, holidays, promotions and so on.

To accurately predict sales of goods, we need to introduce the concept of weekly sales weight index, which will give different weight values to each day according to historical sales records.

Limited space is not explained in detail. Interested friends can refer to a similar blog article, "retail sales data management tool: monthly sales forecast".

3, the dimensions of commodity inventory analysis

Use two diagrams to illustrate inventory analysis.

(1) the inventory structure chart (the definition of effective inventory refers to the quantity of goods sold in a certain sales cycle).

Analysis of commodity coverage

The inventory management of goods is reflected in every link. We can not simply talk about how to reduce inventory in terms of procurement and sales.

There is no need to reduce inventory. It is only meaningful to strengthen management, enhance pparency and increase early warning mechanism from every link.