How to spot patterns in shopping?

Question

Algorithms

How to understand the frequency with which they buy on a more or less voluminous base of user purchases and, most importantly, to distinguish groups by the frequency of purchases?
There is sales data for 2 years with the following fields: user id, purchase id, purchase date, cost.
I would like to single out user groups, for example, those who buy every month, those who buy 3 times in a row and then do not buy for 3 months, etc.
What algorithms are there for this and what tools are preferable?

Answer 1

In general, all these trends can be tracked and built in Excel. This is a very serious tool for analysis.
But I don’t use it, we build algorithms ourselves, identify metrics, write scripts for them to determine the trend.
You will have several dozen metrics, offhand:
1. Seasonality (in which season which gender, age, region bring in more revenue)
2. The same for months
3. Groups of goods (what gender, age, region bring more revenue in groups: equipment, clothes, perfumes)
4. What purchases do customers make from organic, context, from partners.
If you can figure out the trend, you're on the right track.

Answer 2

There are such marketing tools: abc- and xyz-analysis. In the original, they go as assortment analysis algorithms, but if you replace the analyzed characteristics in a certain way, you can get an answer to your question. For example, here is a link to xyz-analysis of clients. On the net, you can also find examples of the implementation of these analyzes (hee-hee) in Excel.