Developers Club geek daily blog

2 years, 12 months ago
Proverbs in itself do not appear … Sometimes you get into such jungle of analytics that necessarily the hand with the hot reaches for locker (all right, we he knows is at each office).

As we the analysis for retail did ABC, or "without half of liter will not understand"


But we will speak a little about other.

In retail, logistics, management of warehouse and stocks there is such thing as ABC analysis. About it many theoretical publications are already written. And, it seems, everything is rather simple and clear, but whether so it actually?

When the category manager or the marketing specialist of distribution network closely approaches carrying out ABC analysis at it inevitably there is the whole lots of questions, fluctuations and doubts. We will also work with them in this article!

Let's walk on algorithm of actions at ABC analysis in product distribution networks, exceptions to the rules which should be considered, we will set example of carrying out the analysis on commodity group of Alcoholic beverages (yes, those half of liter).

If someone hears about ABC analysis for the first time, here
as it becomes.
ABC analysis is the most widespread method of studying of the range. Is its cornerstone, applicable to many aspects of life, Pareto's law. Its essence for retail that 20% of goods give 80% of efficiency, and other 80% of goods – only 20%.

ABC analysis is method by which it is possible to define contribution of each goods to turn and profit of shop, to distribute goods on categories for effective management of the range.
For this purpose it is necessary:
  1. To sort all goods by the selected criterion (for example, to turn).
  2. To count how many percent the turn of each goods makes from total turnover of commodity group.
  3. To count cumulative (or accumulative) percent by addition of percent to the sum of the previous percent.

Goods turn Goods turn percent from total turnover Cumulative percent
Goods 1 100 UAH. 10% 10%
Goods 2 92 UAH. 9.2% 10% + 9.2% = 19.2%
Goods 3... 80 UAH. 8% 19.2% + 8% = 27.2%

We select categories, for example
category A — the priority goods bringing to 80% of total turnover;
category B — normal goods, from 80% to 95% of total turnover;
category C — goods outsiders, from 95% to 100% of total turnover (everything that remained).

We define borders of categories which have to differ significantly among themselves.
  1. We build cumulative curve.
  2. We connect straight line extreme points of curve.
  3. We find point of contact of the line of the parallel received straight line. This point will define borders of category A for which nature of accumulation of qualitative criterion is homogeneous.
  4. Similarly we connect direct line point of border of category A and extreme point of curve.
  5. We find point of contact of the line of the parallel received straight line and we define category B borders.

As we the analysis for retail did ABC, or "without half of liter will not understand"


When carrying out ABC analysis the first that it is necessary to make, it will be defined

How, Why and For what we will use it?


It is important to answer such questions:
  1. What purpose of the analysis?
  2. What will be objects of the analysis?
  3. By what criteria?
  4. What percentage ratio will be optimum for ABC analysis?
  5. For what time period it is worth carrying out the analysis? and with what frequency?
  6. How to separate goods on And, In, From category?
  7. What will be interpretation and actions on the basis of results of the analysis?

Let's walk on points.

The purpose of the analysis depends on the existing problem or, and why we in general carry out it? Any analytics serves for achievement of some purpose, ABC analysis not exception at all. Accurate vision of the purpose already half of success of marketing activity.

The purpose predicts that we can reach by means of ABC analysis application therefore can will cause a stir even depending on that who the analysis carries out. Category managers most often analyze sales of the goods managing shops — turn, marketing specialists — entry of goods into checks of buyers.

The most popular purposes it:
  • to define the groups of goods making the greatest profit;
  • to optimize the range;
  • to select leading goods and outsiders;
  • to manage stocks and deliveries;
  • to compare indicators to the previous period, to analyze changes.

It is possible to achieve the goal using different Objects of the analysis. Can act as them — deliveries, warehouse stocks, the commodity range of separate shop or all distribution network, goods which are included into certain commodity group or category.

Here it is necessary to approach the analysis rather carefully. For example, for optimization of the range, the analysis on all range of shop or network will not give practically anything. After all we in shop only cannot leave bread, milk and alcohol though these groups and will be the most popular. And here in section of separate commodity group it is possible to trace easily goods of group C (outsiders on turn and number of sales) of which it is necessary to get rid.

Criteria. Are besides closely connected with object and the purpose of the analysis.
The most widespread:
  • turn;
  • revenue;
  • profitability;
  • number of sales;
  • number of checks, entry into checks — the frequency of purchases of goods.

The choice of only one criterion for the analysis significantly limits reliability of results. As a rule, use two-three criteria and carry out the cross-analysis about what we will tell below in more detail.

Percentage ratio. Unfortunately, the average values offered by Pareto principle not always are true. In reality the category manager or the managing director of shop when determining percentage ratio is guided, first of all, by the experience, the purposes and kritery the analysis, specifics of the range of commodity group, shop or network retail.
80-15-5,
70-20-10,
50-30-20,
and even 40-40-20, it is all possible options of percentage ratios of categories A, B and C.
The broad dispersion indicates variety of situations and impossibility of orientation to universal ratio of borders of categories. So, the kategoriyshchik of big distribution network is able to afford to bring significant amount of goods of category C out of the range, shelves of shop anyway will not be empty. Another matter the managing director of small distribution network from 2-3 shops where removal of 100-200 goods will perniciously affect the width of the provided range.

Time period. Often to carry out ABC analysis too costly on use of working hours of marketing specialists, kategoriyshchik or managing directors of shops and results of such analysis will to put it mildly not be obvious because of recurrence of sales of goods on days of the week or seasons.

For example, the analysis of all commodity range it is possible to carry out time to half a year to analyze what goods and groups of goods the most important and that has changed in comparison with the last period.

The analysis of goods in each commodity group is, as a rule, carried out time in 2 months, options of times in 3 months are possible. Everything depends on the value of the range and opportunities of analysts of network.

Division on And, In, From category.
Analyzing the trade range of shop the marketing specialist can use 1 criterion — for example, profitability of goods or commodity group, but data retrieveds are not always rather useful.

Therefore the cross the analysis by several criteria is applied at once. Yes, such approach is not simple, but use of bigger number of criteria allows to see the existing situation better. When carrying out some options of actions are possible:

1. Consecutive division into categories.

It is worth using if the range of commodity group too big. At first the range is analyzed by the first criterion (for example, to turn), further each received category are analyzed again already by the second criterion (number of sales), etc. As a result we receive subcategories with rather small inventory with which it is convenient to work.

2. Parallel division into categories.

We carry out ABC analysis at the same time (in parallel) by neskolny criteria creating categories of type of AA, AF and t.p …

Using 2 criteria, we will tell the Income and Number of sales, we receive already 9 categories:
AA AB BA
BB BC CB
CC AC CA

We use 3 criteria — 27 categories. For example:
Revenue Profitability Number of sales
Goods 1 And In And
Goods 2 And And And
Goods 3... With In With

Such approach is more difficult, gives categories, big by quantity of goods, but allows to receive extensive information on each category.

For example, using 3 criteria for the parallel analysis, goods the received AAA it is the most important goods for the retailer. They bring in the considerable income, are often bought, generate revenue. So have to be constant available, with uninterrupted deliveries and good stock.

Goods of categories ABA, BAA, AAB are also rather important and it is worth working with them actively. For example, the goods are included into category A on revenue and profitability, and into category B on sales. Once it finds the best place on the shelf, or to carry out promoaktivnost and shop will get considerable profit. Still option, commodity group with category A by number of sales and profitability, and with category B on revenue. For goods in this category review of price policy is possible, so insignificant increase in the price of goods will lead to increase in revenue of shop.

And here it is worth getting rid of type of goods of CCC precisely.

3. Use of synthetic approach to definition of categories.

For each criterion the weight factor (WF), depending on its importance for the analysis purpose is defined.
For example, the Turn is more important for the analysis than the Number of sales of goods, and Number of sales is more important than Entry into checks.
Criterion Weight factor
Turn 0,5
Number of sales 0,3
Entry into checks 0,2
In total 1

For each goods calculation of synthetic indicator is made.
Kumul. % according to Oborotu*vk Kumul. % by Quantity prodazh*vk Kumul. % on Occurrence in cheki*vk Synthetic indicator
Goods 1 10%*0,5 = 5 10%*0,3 = 3 11%*0,2 = 2,2 5+3+2,2 = 10,2
Goods 2 19,2%*0,5 = 9,6 20%*0,3 = 6 22%*0,2 = 4,4 9,6+6+4,4 = 20
Goods 3 27,2%*0,5 = 13,6 30%*0,3 = 9 28%*0,2 = 5,6 13,6+9+5,6 = 28,2

Further, it is necessary to carry out ranging of the received results.

This approach gives the chance one number to characterize each commodity position included in classification and to carry out ABC analysis as though only one criterion was used.

Interpretation. Results of ABC analysis have to be attentively studied, you should not accept hasty solutions.
The idea of classical ABC analysis anyway remains invariable — distribution of goods on categories for further work. The analysis allows to define the goods demanding the maximum attention of marketing specialists the kategoriyshchik managing on high-quality influence on activity of distribution network thus limiting area of management to the necessary minimum.

Number category A is always minimum, categories C — is maximum. In too time the category A is priority in respect of service and work with it. The category B has the standard level of service, category C — if goods are not brought out of the range, have naymenshy level of service and attention.

What it is worth remembering or exceptions to the rules


The goods of the main range and goods which have dropped out of it. In the main assortment goods are on sale at least 2 times a week for the period selected for the analysis. Goods which for any reasons began to be on sale less often 2 weekly drop out of the main range. It can be the elite, new, seasonal or absent in warehouse goods. It makes sense to carry out ABC analysis on the main range. And it is necessary to pay attention to the goods which have dropped out of the main range and to establish the reason of falling of sales.

Promotional goods. If for the period taken for ABC analysis from you passed actions in distribution network or separate shop, results of sales of promotional goods can affect reliability of the analysis considerably. Here it is important to marketing specialist to solve, whether to exclude the goods getting under action from data set for the analysis or to make for them certain amendment depending on action conditions.

Elite goods. Goods which do not enter into the main range of shop or network (are on sale less than 2 times a week and even are much more rare), but at sale can bring in the considerable income. They can be included in data array for ABC analysis where with considerable probability they will get to category C. But such goods are important for the range of shop, so it is impossible to display them. At the same time, because of the low frequency of sales it is inexpedient to select under elite goods place in store warehouse, it is simpler to organize their purchases upon sale.

Goods novelties. It is clear to any that what advertized was not new goods, at first its sales will be much lower than the checked brands. But, at the same time, new goods are absolutely necessary in any shop. There are candidate solutions.

New goods do not join in the analysis and the first some months of sales cannot be brought out of the range.

If to exclude new goods from data array too difficult technically, the label "New" is appropriated to them, and at interpretation of results of ABC analysis such goods do not fall under reduction.

One more option is inclusion of new goods in category A automatically. Than it is bad? That certain quantity of new goods in category A, displaces other goods in ranging below.

The absent goods. For various reasons goods can sometimes not be on shelves of shop or in warehouse. I.e., in principle, it could be on sale, and there was demand, but in data for the analysis of sales the goods are not present. Therefore it is useful to know date of the last arrival of goods to shop when carrying out interpretation of ABC analysis.

Let's give example on carrying out ABC analysis.


In network of supermarkets from 17 shops there were certain problems with commodity group "Alcoholic beverages". Goods of this group well were on sale and brought in the income, but occupied considerable half-internal space of shops. Also, it was required to define brands and separate goods for planning of autumn promotion actions. We have carried out ABC analysis by means of the BI Datawiz.io service.

So, the analysis purpose — choice of goods for promotion actions, reduction of the range of commodity group.

Object of the analysis — the main range of Alkogol group on all distribution network.

The time period — 2 months.
The analysis will be is carried out by means of parallel approach by 2 criteria: Turn and Number of sales. The choice of these criteria directly depends on the analysis purpose. Managing directors of distribution network needed to reduce quantity of the goods taking place on shelves and not considerably influencing turn of commodity group in general.

The analysis on the main range will allow us to obtain more exact data without the seasonal or absent on sale goods.

As we the analysis for retail did ABC, or "without half of liter will not understand"

Percentage ratio.
In this option the ratio 75-95-100 by the selected criteria because of specifics of commodity group will be optimum.
On screenshot lower we see quantity of commodity positions which enter each category A, B and C and percentage part of category from the general indicator.

As we the analysis for retail did ABC, or "without half of liter will not understand"

For descriptive reasons ratios of categories we will consider them on the chart.

As we the analysis for retail did ABC, or "without half of liter will not understand"

As we see, category A both on turn, and by number of sales the smallest, the category C the biggest — is from where to take goods for reduction.

Interpretation. Let's analyze the received results.
The analysis is possible both by means of tabular data, and by means of visualization.

The first purpose — the Choice of goods for carrying out promotion actions.
In category AA on the Turn and Number of sales 162 positions of goods, apparently on screen get below.

As we the analysis for retail did ABC, or "without half of liter will not understand"

We can visualize data on each category.

For example, now for visualization creation we used such indicators:
horizontal axis — number of sales for the selected period;
vertical axis — turn for the selected period;
diameter of circle — % of turn of the selected category. Also other options of creation of the diagram depending on the ABC analysis purposes are possible.

As we the analysis for retail did ABC, or "without half of liter will not understand"

As we see with considerable separation in category AA GreenDay Organic Life on sales in this distribution network is in the lead.

The most on sale brands it is GreenDay and MEDOFF. Work with suppliers of such goods has to be very well adjusted, they deliver us goods leaders. Creation of special best conditions for them, occasional seat on shelves, the organization of promoaktivnost, etc. is possible.

But, we consider inexpedient to carry out promotion actions for goods of category AA, these goods and without action perfectly are on sale.

In this case it is better to plan advance for category AV which considerably influences turn of shops, and the number of sales of goods of group will grow as a result of promoaktivnost.

Results of choice of goods of category AV it is visible in drawing lower.

As we the analysis for retail did ABC, or "without half of liter will not understand"

Apparently the most successful will be to carry out promotion actions for the Georgian cognacs and wines, and also cognacs of brand of "Blades".

Optimization of the range and disposal of not on sale goods was the second purpose of our analysis. Let's deal with category CC.
Here visualization even more simplifies the analysis. As we remember, elite goods also can get to this group. For example, on the picture below scotch whisky with the price higher than 800 UAH for bottle in 2 months were on sale only 2 times, but has made considerable profit.

As we the analysis for retail did ABC, or "without half of liter will not understand"

And here goods on 2 axes aiming at zero and with small diameter of circle, not influencing total turnover it is worth deleting from the range — they are not on sale and only take place on shelves. As example, in drawing The Sun in Glass wine — was on sale only 2 times in 2 months on 32 UAH for bottle so and does not influence turn in any way.

As we the analysis for retail did ABC, or "without half of liter will not understand"

Thus, ABC analysis has allowed us to distribute the goods entering into commodity group of Alcoholic beverages on 9 different categories and to develop recommendations for distribution network about optimization of the range:
  • category AA — leading goods, category of the highest precedence, goods constantly have to be available, careful control of stock rate is necessary;
  • category AV — goods which will bring maximum efficiency when carrying out promotion actions;
  • categories BA, BB, AF, SV — goods middlings, the average level of stockpile management and placement on shelves;
  • category CC — goods outsiders, the detailed analysis of category and removal from the range of the most low effective goods is necessary.

Work is performed, it is possible and to note! Especially we it is aware of hot trends now.

This article is a translation of the original post at habrahabr.ru/post/267175/
If you have any questions regarding the material covered in the article above, please, contact the original author of the post.
If you have any complaints about this article or you want this article to be deleted, please, drop an email here: sysmagazine.com@gmail.com.

We believe that the knowledge, which is available at the most popular Russian IT blog habrahabr.ru, should be accessed by everyone, even though it is poorly translated.
Shared knowledge makes the world better.
Best wishes.

comments powered by Disqus