ChatGPT and STLC: A Synergistic Testing Approach

14 min readAug 24, 2023

The Software Testing Life Cycle (STLC) is a critical stage in the software development process that ensures a software product’s quality and stability. ChatGPT, a strong language model developed by OpenAI, is one of the developing technologies that has found use in STLC. ChatGPT can be useful at several stages of the STLC, from requirements analysis through post-release support. In this post, we will look at real-world examples of how ChatGPT may be implemented into the STLC.

1. Requirements Analysis:

The requirements analysis phase of the Software Testing Life Cycle (STLC) is crucial because it allows the project team to identify and record the needs and expectations of stakeholders for the software product. Integrating ChatGPT into this phase may considerably improve the process by offering human-like comments and insights that assist in requirement refinement. Let’s look at some examples of how ChatGPT may help with requirements analysis.

1. Establishing User Stories:

User stories are concise descriptions of a feature from an end-user’s perspective. ChatGPT can assist in generating user stories based on high-level discussions. For example:

Human Input: “We need a user story for the login process.”
ChatGPT Response: “As a user, I want to be able to log in using my email and password, so I can access my account and make purchases conveniently.”

2. Clarifying Ambiguities:

Requirements might have ambiguities that need clarification. ChatGPT can help by providing interpretations or suggestions for clarification. For instance:

Human Input: “The requirement states ‘fast response time.’ Can you suggest how to define ‘fast’?”
ChatGPT Response: “Certainly, ‘fast response time’ could mean the system should respond within 2 seconds for any user action.”

3. Identifying Edge Cases:

ChatGPT can help uncover potential edge cases that might not have been initially considered. For instance:

Human Input: “We need to define edge cases for the payment gateway.”
ChatGPT Response: “An edge case could be when the payment gateway encounters a transaction with an amount exceeding $10,000. How should the system handle such cases?”

4. Gathering User Insights:

By simulating user interactions, ChatGPT can provide insights into user expectations and requirements. For example:

Human Input: “What might users expect from the search feature?”
ChatGPT Response: “Users would likely want a search bar prominently displayed, with the ability to filter results by category and sort by relevance or price.”

5. Generating Use Cases:

Use cases describe interactions between actors and the system. ChatGPT can generate use cases based on the provided scenarios. For example:

Human Input: “We need a use case for the ‘Add to Cart’ functionality.”
ChatGPT Response: “Title: Add Item to Cart Actors: User, System Main Flow:

User searches for a product.
User selects a product.
System displays product details.
User clicks ‘Add to Cart.’
System updates the cart and confirms the addition.

Alternative Flow:

If the product is out of stock, the system displays an ‘Out of Stock’ message.”

6. Prioritizing Features:

ChatGPT can provide insights into the potential impact and importance of different features, aiding in prioritization discussions. For instance:

Human Input: “Help us prioritize the features for the MVP (Minimum Viable Product).”
ChatGPT Response: “For the MVP, focusing on user registration, product search, and cart functionality would provide the core user experience.”

7. Exploring Functionalities:

ChatGPT can assist in exploring new ideas and functionalities by suggesting potential features based on existing discussions. For example:

Human Input: “We’re considering additional features for the dashboard. Any suggestions?”
ChatGPT Response: “Integrating order history, personalized recommendations, and a quick reorder option could enhance the dashboard’s value to users.”

By Incorporating ChatGPT into requirements analysis, the team is better able to develop clear, comprehensive, and well-defined requirements. ChatGPT’s natural language features help to elicit user stories, explain misunderstandings, discover edge situations, and acquire useful insights. This eventually helps the software development process succeed by ensuring that the software output fits the needs and expectations of stakeholders.

2. Test Planning:

Test planning is an important stage in the Software Testing Life Cycle (STLC) since it defines the testing strategy, objectives, and scope. Recommend alternative testing methodologies, highlight crucial scenarios, and assist in the formulation of complete test plans, Including ChatGPT in this step may considerably improve the test planning process. Here are some instances of how ChatGPT might help with test planning.

1. Choosing Testing Strategies:

ChatGPT can provide insights into suitable testing strategies based on the project’s context. For instance:

Human Input: “What testing strategies should we consider for our mobile app?”
ChatGPT Response: “For a mobile app, consider a combination of functional testing, compatibility testing across devices and OS versions, and performance testing to ensure a smooth user experience.”

2. Identifying Critical Scenarios:

ChatGPT can help identify scenarios that are critical for testing, ensuring thorough coverage. For example:

Human Input: “What are some critical scenarios to test in our e-commerce application?”
ChatGPT Response: “Testing scenarios like successful order placement, payment failure handling, and accurate inventory update are critical to ensure seamless shopping.”

3. Creating Comprehensive Test Plans:

ChatGPT can assist in creating detailed test plans by suggesting key areas to cover. For instance:

Human Input: “We need to create a test plan for our online banking system. What should it include?”
ChatGPT Response: “Your test plan should cover areas like user authentication, fund transfers, balance updates, account statement generation, and security measures like encryption.”

4. Test Environment Recommendations:

ChatGPT can provide recommendations for setting up test environments effectively. For example:

Human Input: “What should our test environment for a real-time messaging app include?”
ChatGPT Response: “Your test environment should include multiple devices with different operating systems, network conditions mimicking real-world usage, and tools to simulate message delivery delays.”

5. Defining Test Data Strategies:

ChatGPT can suggest strategies for generating test data that effectively represent different scenarios. For instance:

Human Input: “How can we create test data for stress testing our weather app?”
ChatGPT Response: “Generate test data with extreme weather conditions, data from multiple geographical locations, and rapidly changing weather patterns to simulate stress scenarios.”

6. Integration of Automated Testing:

ChatGPT can advise on the integration of automated testing into the test plan. For example:

Human Input: “Should we automate regression testing for our content management system?”
ChatGPT Response: “Yes, consider automating regression tests for content updates, user authentication, and core functionalities to ensure rapid validation.”

7. Considering Security Testing:

ChatGPT can help in outlining security testing considerations. For example:

Human Input: “What security testing should be part of our online marketplace’s test plan?”
ChatGPT Response: “Include tests for data encryption, SQL injection, cross-site scripting (XSS), and user authorization vulnerabilities.”

8. Load and Performance Testing Strategies:

ChatGPT can offer strategies for load and performance testing. For instance:

Human Input: “How should we approach load testing for our video streaming platform?”
ChatGPT Response: “Simulate high user loads, measure server response times, and test video streaming under various network conditions to ensure optimal performance.”

Project teams may benefit from ChatGPT’s observations and suggestions to produce well-structured, comprehensive test plans by including them in the test preparation process. ChatGPT may help improve the entire quality of the testing process by selecting appropriate testing techniques, identifying essential situations, creating test data strategies, and addressing security and performance testing.

3. Test Case Design:

Test case design is an important stage in the product Testing Life Cycle (STLC) in which thorough test cases are produced to confirm the functionality of the product. Integrating ChatGPT into this step may improve the test case design process greatly by providing complete test cases, spanning diverse circumstances, and even recommending innovative edge cases. Here are some instances of how ChatGPT may help with test case design.

1. Generating Test Cases:

ChatGPT can assist in generating detailed test cases based on the specified functionality. For example:

Human Input: “We need test cases for the ‘Login’ functionality.”
ChatGPT Response:
Test Case 1: Verify that a registered user can successfully log in using valid credentials.
Input: Valid username and password
Expected Outcome: The user is logged in and redirected to the dashboard.
Test Case 2: Verify that an invalid login attempt with incorrect credentials results in an error message.
Input: Invalid username and password
Expected Outcome: An error message is displayed, and the user remains on the login page.

2. Boundary and Edge Cases:

ChatGPT can suggest boundary and edge cases to ensure thorough coverage. For example:

Human Input: “What are some boundary test cases for an ‘Age Verification’ feature?”
ChatGPT Response:
Test Case 1: Verify that a user with the minimum valid age (18) is granted access.
Input: Age = 18
Expected Outcome: Access granted.
Test Case 2: Verify that a user with an age below the minimum valid age is denied access.
Input: Age = 17
Expected Outcome: Access denied.

3. Combination of Inputs:

ChatGPT can suggest test cases that involve a combination of inputs. For instance:

Human Input: “Generate test cases for a ‘Search and Filter’ feature.”
ChatGPT Response:
Test Case 1: Verify that search results are displayed correctly when filtering by category and keyword.
Input: Category = Electronics, Keyword = “Smartphone”
Expected Outcome: Relevant smartphone products from the Electronics category are displayed.
Test Case 2: Verify that applying multiple filters results in accurate search results.
Input: Category = Clothing, Size = Medium, Color = Blue
Expected Outcome: Clothing products of medium size and blue colour are displayed.

4. Negative Testing:

ChatGPT can suggest negative test cases to validate error handling. For example:

Human Input: “Provide negative test cases for a ‘Credit Card Payment’ feature.”
ChatGPT Response:
Test Case 1: Verify that attempting to process a payment with an expired credit card displays an appropriate error message.
Input: Expired credit card details
Expected Outcome: Error message indicating card expiration.
Test Case 2: Verify that a payment with an incorrect CVV results in a transaction failure.
Input: Incorrect CVV
Expected Outcome: Payment failure message.

5. Exploring Unexpected Scenarios:

ChatGPT can assist in exploring unexpected scenarios and suggesting corresponding test cases. For example:

Human Input: “Can you suggest some test cases to ensure data integrity in a ‘File Upload’ feature?”
ChatGPT Response:
Test Case 1: Verify that uploading a file with a non-supported format results in an error and no data corruption.
Input: Upload a file with an unsupported format
Expected Outcome: Error message, no data corruption.
Test Case 2: Verify that uploading a large file does not cause data loss or system crashes.
Input: Upload a file exceeding 1GB
Expected Outcome: Successful upload, and system stability.

Testing teams may benefit from ChatGPT’s capacity to produce precise, complete, and creatively varied test cases by incorporating them into the test case design phase. ChatGPT may help improve the overall quality and coverage of the testing process by creating boundary cases, evaluating input combinations, investigating negative possibilities, and verifying unexpected occurrences.

4. Test Data Preparation:

Test data preparation is an important stage in the Software Testing Life Cycle (STLC) in which a variety of test data is created to facilitate complete software testing. By recommending different data sources, simulating real-world circumstances, and creating data that covers several use cases, including ChatGPT this step may considerably improve the test data preparation process. Here are some instances of how ChatGPT may help with test data preparation.

1. Creating Test User Profiles:

ChatGPT can assist in generating different user profiles for testing purposes. For example:

Human Input: “We need diverse user profiles for our social networking app tests.”
ChatGPT Response:

User Profile 1: Name: John Doe, Age: 28, Location: New York, Interests: Travel, Photography
User Profile 2: Name: Emily Smith, Age: 22, Location: Los Angeles, Interests: Fashion, Fitness

2. Generating Realistic Transaction Data:

ChatGPT can help in generating transactional data for financial software testing. For instance:

Human Input: “Generate sample transaction data for testing our banking application.”
ChatGPT Response:

Transaction 1: Account: Savings, Amount: $500, Type: Deposit
Transaction 2: Account: Checking, Amount: $200, Type: Withdrawal

3. Simulating Product Inventory:

ChatGPT can assist in creating realistic product inventory for an e-commerce application. For example:

Human Input: “Provide test data for products in our online store.”
ChatGPT Response:

Product 1: Name: Smartphone X, Price: $699, Stock: 50 units
Product 2: Name: Laptop Y, Price: $999, Stock: 20 units

4. Generating Sensor Data:

ChatGPT can suggest sensor data for testing IoT devices. For instance:

Human Input: “Generate sensor data for testing our smart thermostat.”
ChatGPT Response:

Sensor Data 1: Temperature: 72°F, Humidity: 45%, Mode: Cooling
Sensor Data 2: Temperature: 68°F, Humidity: 60%, Mode: Heating

5. Diverse Geographical Locations:

ChatGPT can help in simulating data from different geographical locations for location-based services. For example:

Human Input: “Provide test data for our weather app from various cities.”
ChatGPT Response:

City Data 1: City: New York, Temperature: 75°F, Weather: Clear
City Data 2: City: London, Temperature: 60°F, Weather: Cloudy

6. Simulating User Interactions:

ChatGPT can generate data simulating user interactions for chatbot testing. For instance:

Human Input: “Generate chat interactions for testing our customer support chatbot.”
ChatGPT Response:

Chat Interaction 1: User: “How do I return a product?” Bot: “To initiate a return, visit our Returns page and follow the steps.”
Chat Interaction 2: User: “Can I change my delivery address?” Bot: “Certainly, please provide your order number and the new address.”

7. Complex Input Scenarios:

ChatGPT can help in generating complex inputs for testing software behaviour under challenging conditions. For example:

Human Input: “Generate complex input data for testing our scientific calculator.”
ChatGPT Response:

Input Data 1: Equation: (3.5 * 10⁴) / (2.1–1.8)
Input Data 2: Expression: √(9.6) + sin(60°) — ln(2.5)

Testing teams may benefit from ChatGPT’s capacity to create a wide range of useful and diverse test data by including it in the test data preparation process. ChatGPT may help improve the overall quality and coverage of the testing process by developing test user profiles, replicating product data, producing sensor readings, or designing complicated input scenarios.

5. Test Execution:

Test execution is a critical stage in the product Testing Life Cycle (STLC) in which test cases are performed and the functioning of the product is confirmed. By mimicking user interactions, delivering expected outcomes, and even aiding in troubleshooting, including ChatGPT in this phase may dramatically improve the test execution process. Here are some instances of how ChatGPT may help with test execution.

1. Simulating User Interactions:

ChatGPT can simulate user interactions as if it were a user, allowing testers to execute test cases more effectively. For example:

Human Input: “Simulate a user interaction for the ‘Add to Cart’ functionality.”
ChatGPT Response:
User Input: “Search for a blue t-shirt.”
System Response: “Search results: Blue T-Shirt, $20.”
User Input: “Add the blue t-shirt to the cart.”
System Response: “Blue T-shirt added to cart.”

2. Providing Expected Outcomes:

ChatGPT can assist in providing the expected outcomes for specific test cases. For instance:

Human Input: “What should be the expected outcome for testing the ‘Checkout’ process?”
ChatGPT Response:
Expected Outcome: “After successful checkout, the user should receive an order confirmation email with the order details.”

3. Handling Exception Scenarios:

ChatGPT can suggest responses for handling exception scenarios encountered during test execution. For example:

Human Input: “What error message should be displayed for an invalid credit card during payment?”
ChatGPT Response:
Error Message: “Invalid credit card details. Please check your card information and try again.”

4. Assisting in Troubleshooting:

ChatGPT can guide how to troubleshoot unexpected behaviour during test execution. For instance:

Human Input: “The system is not responding as expected during the payment process. What steps should I take?”
ChatGPT Response:
Troubleshooting Steps: “Check the network connection, verify if the payment gateway is reachable, and review the payment API logs for any errors.”

5. Verifying Successful Transactions:

ChatGPT can assist in verifying the success of transactions during test execution. For example:

Human Input: “How can I confirm that a successful transaction has occurred?”
ChatGPT Response:
Verification Steps: “After a successful transaction, the user should receive a confirmation message on the screen, and the order should be visible in the user’s order history.”

6. Validation of User Inputs:

ChatGPT can help in validating user inputs during test execution. For instance:

Human Input: “How should I validate user input for the ‘Search’ feature?”
ChatGPT Response:
Validation Approach: “Ensure that the search results match the user’s input keywords or filters and that the displayed products are relevant to the search query.”

7. Testing Various Scenarios:

ChatGPT can assist in executing various scenarios to ensure thorough testing coverage. For example:

Human Input: “Execute a test case where the user attempts to log in with incorrect credentials.”
ChatGPT Response:
User Input: “Username: exampleuser, Password: incorrect123”
System Response: “Invalid username or password. Please try again.”

Testers may successfully imitate user interactions, acquire intended results, troubleshoot errors, and assure thorough testing coverage by including ChatGPT in the test execution process. ChatGPT adds to a more efficient and complete testing process by managing error scenarios, aiding in debugging and confirming successful transactions.

6. Defect Reporting:

Defect reporting is an important step in the Software Testing Life Cycle (STLC) in which any abnormalities or errors discovered during testing are documented and submitted for resolution. Aiding in the creation of clear, thorough, and actionable bug reports, including ChatGPT in this phase may dramatically improve the defect-reporting process. Here are some instances of how ChatGPT might help with defect reporting.

1. Describing Reproduction Steps:

ChatGPT can help in describing the steps to reproduce a defect accurately. For example:

Human Input: “Describe the steps to reproduce the issue with the ‘Add to Cart’ functionality.”
ChatGPT Response:
Reproduction Steps:

Log in to the system with valid credentials.
Search for the product “Blue T-Shirt.”
Click on the product to view its details.
Click the ‘Add to Cart’ button.
Observe that the cart icon does not update, and the product is not added to the cart.

2. Including Screenshots or Visual Aids:

ChatGPT can suggest including screenshots or visual aids to enhance the bug report. For instance:

Human Input: “How can I improve the defect report for the UI alignment issue?”
ChatGPT Response:
Recommendation: “Attach a screenshot of the affected screen with annotations indicating the misaligned elements to provide a clear visual representation of the issue.”

3. Providing Expected vs. Actual Behavior:

ChatGPT can assist in comparing the expected behaviour with the actual behaviour observed in the defect. For example:

Human Input: “Explain the inconsistency in the search results.”
ChatGPT Response:
Expected Behavior: “Search results should display products matching the user’s keyword.”
Actual Behavior: “Search results are displaying unrelated products, including those not matching the keyword.”

4. Including Environment and Configuration Details:

ChatGPT can suggest including information about the testing environment and configuration to aid in bug reproduction. For instance:

Human Input: “What details should I include to report the issue with the payment gateway?”
ChatGPT Response:
Inclusion: “Specify the browser and version used, the payment method selected, and any relevant settings or parameters used during the test.”

5. Describing Impact and Severity:

ChatGPT can help in assessing the impact and severity of the reported defect. For example:

Human Input: “How severe is the issue with the login process?”
ChatGPT Response:
Severity Assessment: “The login issue prevents users from accessing their accounts, impacting the user experience and hindering core functionality.”

6. Suggesting Additional Information:

ChatGPT can assist in suggesting additional information that could be included in the defect report. For instance:

Human Input: “What other information should I provide in the bug report?”
ChatGPT Response:
Recommendations: “Include the browser console logs, any error messages displayed, and the timestamp of when the issue occurred for more context.”

7. Clear Summary and Title:

ChatGPT can help in crafting clear and concise defect summaries and titles. For example:

Human Input: “What should be the title of the defect related to the ‘Checkout’ process?”
ChatGPT Response:
Defect Title: “Checkout Process Fails to Proceed After Payment Confirmation.”

Testers may benefit from ChatGPT’s capacity to assist in the creation of thorough, concise, and actionable bug reports by incorporating them into the defect-reporting process. ChatGPT helps testers and developers communicate more effectively by outlining replication procedures, offering visual assistance, comparing expected and actual behaviour, and recommending more information. This leads to faster issue resolution.

Finally, ChatGPT can be useful at several stages of the Software Testing Life Cycle. Because of its natural language features, it is a useful tool for creating test scenarios, test cases, test data, and even mimicking user interactions. Teams may improve their testing process, increase test coverage, and ensure the delivery of high-quality software products by including ChatGPT in the STLC.