Should I Use Code or Low/No Code Tools For Salesforce Test Automation?
Our clients often have mature Salesforce orgs with intricate customizations, automations, and numerous integrations:
Any modification to these customizations, or the addition of new features can result in unintended consequences if comprehensive end-to-end testing of the Salesforce org is not performed.
Large organizations know that performing manual testing for each feature is a labor-intensive task. This process can also be prone to errors due to the large number of tests that QA needs to perform.
We’ve observed that the timeframe allotted to QA is often insufficient to conduct a complete regression test, due to the pressure from the business side to release new features and fix defects. It’s in exactly this kind of scenario that automated testing can be beneficial.
Before diving in, here is a quick overview of automated testing:
Automated testing has been in use for several decades. It has undergone significant advancements in that time. The earliest automated testing tools were designed to test mainframe applications in the 1970s. With the advent of graphical user interfaces (GUIs) in the 1980s and 1990s, automated testing tools started incorporating GUI testing capabilities for desktop applications.
Over the past two decades, web and mobile applications have become increasingly popular, and automated testing tools have adapted to support browser and mobile device testing. Salesforce falls under the category of web applications, which is an established area of automated testing.
While there are many sub-branches of automated testing, there are two types of testing that are particularly applicable to Salesforce:
UI Testing
UI (User Interface) testing involves verifying the functionality of the user interface in Salesforce applications. UI testing is typically done to ensure that the application behaves as expected from the user's perspective and that all user interface components are functioning properly.
A notable aspect of Salesforce testing is that the UI for Lightning Experience is standardized. There is a linear path for creating or updating records, unless custom Lightning Web Components (LWCs) have been developed. Even with LWCs, the UI components usually adhere to standard patterns, making it relatively simple to automate the testing of Lightning Experience UI.
API Testing
Automated testing for Salesforce Application Programming Interfaces (APIs) focuses on testing the integrations between Salesforce applications and 3rd party applications through APIs. The goal of API testing is to ensure that data is correctly sent and received between Salesforce and other systems, and that the APIs themselves are functioning as intended.
The standard API calls in Salesforce are analogous to users executing queries or Data Manipulation Language (DML) operations. For instance, an API call can be used to create a new lead record without utilizing the Salesforce Lightning UI. This type of testing mimics third-party applications that either retrieve or transmit data from or to Salesforce over secure request and response protocols.
Salesforce Testing Solutions
There are two categories of test automation for Salesforce: no/low-code testing & code-based testing. Let’s explore the pros and cons of each:
The No/Low-Code Testing Solution
Think of this as declarative test automation development. If you’re familiar with Salesforce process automations this looks similar to Flows. This option enables you to record your steps and provides you with the ability to rerun them. The recorded test scripts can be edited using visual tools.
Pros
No/low-code testing tools are user-friendly and require less technical expertise to use. Your organization can delegate the task of generating test scripts to non-coding resources. Your business analysts (BAs) or Salesforce Admins who are not programmers can participate in creating test scripts by using the test step recorder and visual editing tools.
Cons
With ease of use comes reduced flexibility and some limitations in terms of functionality. Generally, code-based solutions provide more flexibility and can handle the edge cases of test automation. The most significant downside of no/low-code is vendor lock-in.
No/low-code test automation tools are often provided by proprietary platforms that may require ongoing vendor support. If a feature is not supported, your team may need to wait for the feature to be released in order to resolve a particular test automation issues. Additionally, this may restrict your ability to switch to a different tool or platform if necessary.
The Code-Base Testing Solution
This involves a developer coding the precise test automation you require for your organization. It’s custom, flexible, and depending on your team, can be quick to execute.
Pros
Code-based testing offers greater flexibility by allowing test scripts to operate at the HTML Document Object Model (DOM) level. This means that test scripts can adapt quickly to changes in web standards or significant changes in Salesforce's UI structures, such as the transition from Salesforce Classic to Lightning Experience. With code-based testing, your team is not dependent on specific vendors to support new web standards, and can write and execute tests to meet your specific requirements.
Cons
This flexibility comes at a price. The code-base solution may require more technical expertise and resources to implement and maintain. Your team may also have a limited number of resources who can contribute to test script generations as they require proficiency in programming languages and frameworks.
Summary
Whether you choose a no/low-code or code-based solution for automated testing, it’s important to consider the skills and aptitudes of the people responsible for the testing process.
While no-code solutions may promise that anyone can contribute to automated testing, the reality is that QA is a specialized discipline that requires expertise. To ensure the success of your automated testing strategy, it’s essential to have QA experts on your team.
Our team of experts can assist you in devising test automation strategies, and help you effectively implement and maintain test automations using either no/low-code or code-based solutions.
Tim is a Salesforce developer, software engineer, and machine learning expert.