The Importance of Metrics and Their Usage in GoodData.CN

Running a successful company and having satisfied customers requires analyzing your company’s results. How? By collecting data. However, without the necessary organization and analysis, your data will hold no meaningful value and, consequently, will fail to enable you to identify if you are performing well or whether something is holding you back. For that reason, the data you collect needs to be organized with a set of valuable metrics defined.

What are metrics, you ask? Metrics are essentially a set of measurements that help you track and calculate the performance of specific business processes. In general terms, metrics make reference to the large amount of data that has been obtained (by various methods) depending on their resources. Management then uses these metrics to generate relevant, in-depth reports giving them an objective overview of the company.

Why Do You Need Metrics?

Forming an essential part of analytical processes, metrics can be created according to general or individual criteria stipulated by the company. They can, for example, reflect the organization's performance as a whole or monitor specific activities or employees. A metric consists of different measures, according to the fact table of a logical data model; with the logical data model representing the relationships between different entities in your data. The fact table, the central table of the logical data model, contains at least one fact; a collection of numerical data stored to create metrics.

Well-defined metrics are able to answer your questions about your financial situation, human resources, or sales. They are also able to help you:

• Understand relationships between different events in your company
• Uncover strengths and weaknesses in your business
• Find ways by which to decrease or eliminate the impact said weaknesses have
• Change or modify the direction of your business strategy
• Improve your decision-making process

To make more efficient decisions based on your company processes, using only one metric is not recommended, as it often fails to give you the full picture — which I will highlight with an example. Imagine you are trying to track your net profit: Company successes and goal achievements are measured with a combination of metrics or measures, including several influences. Your metric can be defined in the following way:

Net Profit = (Total Revenues - Total Costs) * Tax rate

The mentioned metric consists of measurements as Total Revenues and Total Costs adjusted by the Tax rate. Similarly, various combinations of metrics could also be connected with different filter expressions to provide more detailed overviews of the company's performance.

Understanding the Basics in GoodData.CN

GoodData.CN is a cloud-native analytics platform that can be managed by users. It gives you a headless BI framework free of charge, deployed as a Docker Image — the Installation guide can be found here. You can connect your data to GoodData.CN in the same way as you would with other analytical engines.

As a self-service tool, GoodData.CN Community Edition allows you to manage your logical data model and dashboards by yourself. Before generating dashboards, you should get familiar with the basic terminology used in GoodData, including, metrics, measures, and MAQL query language.

Measures represent numeric aggregations of transactional records or events (during a specific period) stored in the fact table of your logical data model. Examples of measures include; total orders, total revenues, or extremes such as maximum orders in target regions. They do not necessarily have to carry valuable information and, for that reason, they are combined to form a given metric (which does).

In GoodData.CN, metrics can be presented as number aggregations of facts or counts of selected attribute values in final insights. They are customized by predefined formulas according to various requirements. To define them, you use MAQL, GoodData's query language. It has a similar syntax to SQL; its main advantage is the reusability and composability of built metrics. See this article for more information about MAQL or if you are ready to start creating your metrics, read the following documentation.

Creating Metrics in GoodData.CN

Besides the vital aspect of a self-service platform, the main advantage of GoodData.CN metrics are the possibility to use re-use them in different insights. There are 3 ways by which metrics and measures can be created in GoodData.CN:

1. You can create simple measures by dragging and dropping directly within the Analytical Designer and then save them as a new insight.

2. You can create more complex metrics via the Metric editor in GoodData.CN. To find out what your Order Amount is, for example, you can define it as a product of price and quantity using MAQL (as shown below).

SELECT SUM({fact/price}*{fact/quantity})

To save your metric as a new insight, repeat the previous step.

3. Lastly, metrics can be implemented using an API. The implementation can be carried out via Postman or via any other similar tool. First, however, you need to write down your cURL, as follows:

curl -H "Authorization: Bearer YWRtaW46Ym9vdHN0cmFwOmFkbWluMTIz" \
\-H "Content-Type: application/vnd.gooddata.api+json" \
\-H "Accept: application/vnd.gooddata.api+json" \
\-X POST \
http://localhost:3000/api/entities/workspaces/{here_enter_your_workspace_ID}/metrics \
\-d '{
"data": { "id": "order_amount", "type": "metric", "attributes": {
"title": "Order Amount", "description": "", "content": {
"format": "$#,##0", "maql": "SELECT SUM({fact/price}*{fact/quantity})"
}}}} -'

To find out more information about using APIs, read our public API documentation.

Try Creating Metrics Yourself

If you are interested in GoodData.CN, please contact us. Alternatively, sign up for a trial version of GoodData Cloud: https://www.gooddata.com/trial/