Creación de gráficos de relación: doble eje y dispersión

How does the relationship between metrics influence our decisions?

When we are faced with data and charts, it is crucial to understand how metrics relate to each other. While we might sometimes assume connections that don't exist, when we explore data correctly we can discover correlations that help us make more informed decisions. This article discusses how to create and understand these relationships by using scatterplots and combo charts to make crucial business and data analysis decisions.

How do I set up a scatterplot to understand relationships?

A scatter plot is a powerful visual tool that allows us to understand the relationship between two variables, in this case, metrics. Here, I explain how an analysis can be performed using a scatter plot to identify patterns and inverse relationships.

To illustrate how one metric can affect another, consider the following example: is it true that more expensive products have fewer transactions?

1. Selecting the right metrics: In this case, to measure product price, we need to select therevenue metric. Then, we look for the number oftransactions.

2. Build the scatter plot:

   # Example of data representationX-axis: Product price (revenue per product)Y-axis: Number of transactions.

3. Interpretation of the graph: We can observe an inverse relationship, i.e. the more expensive a product is, the fewer transactions it has. On the left side of the graph, we find cheaper products and a higher number of transactions. As we move to the right, more expensive products show fewer transactions.

How useful is a combo chart in advertising management?

For day-to-day analysis of ad spend and associated conversions, combo charts are extremely effective. They allow you to compare metrics that would otherwise not be easily comparable due to their different value scales.

Consider the following scenario: advertising spending is suspected to be related to conversions throughout the day.

1. Defining the dimension of interest: For this analysis, the key dimension is "time", specifically "hours of the day".

2. Use of appropriate data: A data source is selected that allows mixing multiple metrics. In this case, the source that provides both spend(cost) and conversions is required.

3. Data visualization: Create a chart in which the hour dimensions organize the data to range from 0 to 24 hours.

4. Customize the chart: Adjust the style of the chart to ensure comparability, adjusting the conversions axis so that it is represented on the right side of the chart.

5. Interpret results: Identify inefficiencies, such as times when ad spend does not produce corresponding conversions and observe patterns useful when planning marketing strategies, such as the case observed at 8 p.m. where less is spent, but more conversions are generated.

What types of charts are suitable for different types of analysis?

There are several types of charts that can be used to analyze relationships between variables, from scatter charts to combo charts. Each type offers specific attributes that should be considered depending on the data analysis questions you are seeking to answer.

Scatter plots are ideal for:

• Identifying direct or inverse correlations.
• Reviewing the behavior of various metrics as a variable changes.

While the combined graphs are useful for:

• Comparing metrics with different scales.
• Review trends on a daily or time period basis.

Although illustrated charts are particularly effective for certain analyses, there are multiple types of charts that can fulfill different visual and analytical needs in your data analysis journey.

As you become familiar with how to interpret different charts and identify relationships, you will develop invaluable analytical skills that will allow you to make more informed, efficient and effective decisions. Continue to explore, learn and be curious to master data analysis and metric relationships!