It is important for any entrepreneur to understand his clients for effective marketing. That’s why we make a portrait of a client, determine his gender, average age and income, and so on.
We will talk now about even more subtle classification of customers, which is carried out on the basis of “historical” data and which gives a lot of useful information. To do this, it is sufficient to use an ordinary sales data.

Large, medium and small clients

Such analysis will be of great use for a company with variable checks, for example, from 100 to 1000 USD. It may be a store with a wide assortment or a restaurant. The purpose of the analysis is to distribute buyers in three groups depending on their revenue, profits, sales, or other criteria. It helps improve marketing activities, pricing, expand the assortment.

Without such analysis, you are probably wasting a lot of resources. Suppose you have a lot of small customers who generate revenue of 100-300 dollars each. Carry out the classification, and it may turn out that 20% of large customers bring 80% of revenue.
The definition of these groups is dynamic, it depends on the characteristics of the business. For the clusterization, data mining methods are used, namely clustering the entire amount of clients into three groups.
It should be determined, whether to divide clients by checks or by other characteristics.

Cluster analysis

If you analyze a sample of checks or customers, you can get such valuable conclusions for further processing:

— the amount of clients by checks, e.g.:

  • for the checks of 0-200 USD — 1000 clients,
  • 200-1000 USD — 400 clients,
  • 1000-10 000 USD — 150 clients.

— the share of each group:

  • 0-200 USD — 10%,
  • 200-1000 USD — 30%,
  • 1000-10 000 USD — 60%.

Having such a division, you may more effectively analyze promos, sales, loyalty programs for each group. For example, a certain product can attract only small buyers, and therefore bring a minimum profit comparing to the cost of resources it requires.

Leave a Reply

Your email address will not be published. Required fields are marked *