{"id":10624,"date":"2020-08-12T17:33:03","date_gmt":"2020-08-12T17:33:03","guid":{"rendered":"https:\/\/www.genrocket.com\/blog\/?p=10624"},"modified":"2026-06-10T14:25:44","modified_gmt":"2026-06-10T14:25:44","slug":"how-to-use-synthetic-data-to-maximize-test-coverage","status":"publish","type":"post","link":"https:\/\/www.genrocket.com\/blog\/how-to-use-synthetic-data-to-maximize-test-coverage\/","title":{"rendered":"How to Use Synthetic Data to Maximize Test Coverage"},"content":{"rendered":"

There is growing interest among quality assurance professionals in the use of synthetic data generation<\/strong> for software testing<\/a>. Their interest in synthetic data is usually triggered by a requirement for data privacy or the need for accelerated data provisioning for Agile and DevOps. However, the most compelling reason to augment the use of production data, or manually created spreadsheet data, with synthetic data<\/a> is to have total control over the variety of data needed to maximize test coverage.<\/p>\n

<\/p>\n

 <\/p>\n

Properly designed synthetic test data can increase coverage from under 50% to over 90% when provisioned by an advanced Test Data Automation system.<\/em><\/h4>\n

 <\/p>\n


\n<\/p>\n

Using a platform like GenRocket\u2019s Test Data Automation<\/a> (TDA) solution, testers are able to design comprehensive synthetic test data<\/a> containing all of the possible variations, combinations and permutations needed for positive and negative test cases to fully exercise their code. With TDA, more exhaustive testing can be performed to uncover the defects missed by less sophisticated testing before they escape to a production environment.<\/p>\n

And because GenRocket\u2019s TDA is an automated self-service solution, it can deliver any volume or variety of test data at the speed of Agile. For more information about using GenRocket for Agile testing, read Test Data Design Should be Integral to Agile and DevOps.<\/u><\/a><\/p>\n

\"test<\/center>
\n<\/p>\n

 <\/p>\n

Synthetic<\/u> data should really be thought of as intelligent<\/u> test data.<\/em><\/h4>\n

 <\/p>\n


\n<\/p>\n

GenRocket\u2019s TDA technology can generate synthetic data that follows rules and conforms to business logic. It can generate real-time synthetic data in response to changes in state during complex workflows. It can compute results during program execution and generate test data used to verify the expected outcomes of a functional test.<\/p>\n

\n
\n
\"GenRocket<\/div>\n
\n

Using GenRocket\u2019s TDA platform, synthetic data can be blended with queried production data to accurately simulate real-world healthcare transactions or replicate financial data flows to assist in bank fraud detection. Synthetic data can be designed to train AI and machine learning systems or scaled to generate billions of rows of structured and unstructured data for testing Big Data<\/a> applications.<\/p>\n<\/div>\n<\/div>\n<\/div>\n


\n<\/p>\n

Following the GenRocket Methodology<\/strong><\/p>\n

GenRocket has developed a structured methodology to guide testers through a four-step process for designing and generating real-time synthetic test data. It\u2019s part of a flexible and adaptive self-service approach that supports functional, non-functional and regression testing at all levels of the Agile Software Test Pyramid.<\/em><\/p>\n

\n
\"Software\"GenRocket<\/div>\n<\/div>\n

Whether synthetic data is needed for testing at the Unit, Service or UI level, the GenRocket Methodology<\/em> helps testers follow a consistent process for generating the most effective test data in the most efficient manner.<\/p>\n

\"GenRocket<\/center>The GenRocket Methodology<\/em> starts with defining the data MODEL<\/strong> to be used for synthetic data generation. GenRocket self-service modules allow the tester to import a database schema and quickly validate the parent-child-sibling relationships for the data tables used by test cases. Alternatively, data can be modeled by uploading a DDL file to define the structure or stream of synthetic data to be generated.<\/p>\n

Once the data is modeled, the tester is ready to DESIGN<\/strong> the exact data needed for testing. This article provides an expanded explanation of the DESIGN stage of the GenRocket Methodology<\/em> in the paragraphs below.<\/p>\n

After the design stage, the tester is ready to DEPLOY<\/strong> the test data by creating Test Data Scenarios and Test Data Cases that contain instructions for the volume and variety of data to be generated by the GenRocket Engine.<\/em> GenRocket\u2019s 4th key component, the Receiver, is used to receive generated test data and deploy it into the test environment. More than 70 different Receivers are available to meet just about any deployment requirement.<\/p>\n

The fourth stage of the GenRocket Methodology<\/em> is when organization admins and testers MANAGE<\/strong> all of their test data projects. The G-Analytics dashboard provides deep insights into GenRocket use. Team Permissions provides granular control over access and permissions. And advanced GenRocket features intelligently automate common tasks so that testers can quickly organize, copy, and share their test data projects with other testers.<\/p>\n


\n<\/p>\n

Designing Test Data for Maximum Coverage<\/strong><\/p>\n

Now let\u2019s take a closer look at the DESIGN stage of the GenRocket Methodology. In this stage, testers determine what kind of test data they need to fully meet the objectives of their test plan. They create the ideal test data design that will maximize test coverage.<\/p>\n

\n
\n
\"GenRocket<\/div>\n
\n

For most QA professionals, this is a new and different concept. Conventional thinking has taught us that test data is either in the form of a production data subset or an Excel file with columns and rows.<\/p>\n

This paradigm can be hard to unlearn.<\/strong><\/p>\n<\/div>\n<\/div>\n<\/div>\n

For many, the realization that test data can be DESIGNED during an Agile sprint, integrated with a test case and then generated on-demand by a test script is an Aha Moment.<\/em><\/strong> However, once this light bulb is turned on, the paradigm begins to shift from traditional Test Data Management<\/a><\/em> to modern Test Data Automaton.<\/em><\/p>\n


\n<\/p>\n

Understanding the Characteristics of Test Data<\/strong><\/p>\n

As a foundation, let\u2019s begin our discussion of Test Data Design<\/em> with an understanding of potential test data qualities and characteristics. Ask yourself, what data characteristics do you need for a given test? Here are some items to consider.<\/p>\n