Organizing raw Data: Why not everything works for you, all the time
What is it about organising information? The importance of customer analytics lies not only in collecting data. Like all Big Data disciplines, the real importance is its analysis, and for this, we must create models that facilitate the task. Let’s see the steps that you need to take into account in this article about organising raw data.
Before you even deal with your customer’s data, you have to keep in mind the business objective that you are trying to achieve. What you seek through analysis is to prove or disprove a hypothesis. Data is your tool for this.
Therefore, before embarking on the adventure, you must define –
- What do you want to know?
- What data do you need to know?
- How that data is related to each other?
- How it proves or denies your initial hypothesis?
For example, in retail, companies are dedicated to observing the behaviour of consumers in stores to place products in the proper order to maximise sales. The Visual Merchandising department observes that customers analyse the shelves from top to bottom, so they conclude that the hot zone is the highest part of the shelves, so it is in this place that they place the promotional items.
In the same way, an e-commerce company analyzes the heat map of its sales pages and observes where visitors click and how far they scroll down, to make decisions regarding the design of its landing page. These examples are historical and right now they only serve to contextualise.
Today, it is possible to collect larger amounts of data like Netflix does, and even customise it for each client, establishing metrics such as –
- How much time each customer spends looking at a given shelf
- What products did the customer end up buying?
- Which sections has the customer visited?
- Which consumer profile a customer has (elderly, young, father/mother, single, …)
But, as we said above, all this is information that you can collect. Not information you need to collect. This will be determined by the objective with which you obtain it and the data model you use.
The reason for being of each analytical model is the business objective for which it works. This is why, when creating an analytical model one has to consider the different insights that are wanted to obtain for each business objective.
What do you think about it? How do you get organized with new projects?