Performance Testing of Salesforce Education Cloud for a Leading University

Challenge

Given the platform’s critical role in managing prospective student enquiries and internal staff processes, the university needed to ensure that Salesforce Education Cloud could reliably scale under predicted heavy usage while maintaining performance and user response times during peak periods.

The key objectives from performance testing were as follows:

• Simulate realistic and peak load conditions for CRM operations, while validating the system’s capacity to meet the needs of the university
• Identify system bottlenecks and determine performance thresholds
• Provide data-driven insights to support decision-making and risk mitigation

Key Scenarios Tested

Working closely with the university project team, our consultants identified the following critical scenarios as critical business processes that made up the majority of the anticipated load on the system.  The journeys were documented in our test plan, as well as test data, volumes and environment details for the tests.  These were agreed before proceeding to the detailed test design stage.  The scenarios under test were as follows:

1. Student Enquiry Submission
   Simulating high volumes of prospective student interactions via web forms.
2. Staff Enquiry Management
   Emulating administrative tasks such as viewing, updating, and processing enquiries.
3. Staff Processing of Enquiries
   End-to-end staff workflows for handling and resolving cases.
4. Email-to-Case Functionality (Simulating Emails via JMeter Tool)
   Emails sent by students or prospects automatically converted into cases in the CRM.
5. Public Contact Form
   Testing public-facing web forms that funnel data into the CRM system.

Approach

Once the user journeys were identified, our team used Apache JMeter to record performance scripts, which we edited to correlate dynamic values and parameterise key data items, so that each iteration of the scripts was treated as a genuine new request by the server and that a range of realistic data items were used for the tests.  We created a load profile for each user journey / script, which related to a number of expected concurrent users and transactions within the periods being considered for the tests – usually 1-2 hours.  Once the scripts were completed and the data provided, we could then proceed to the execution phase.  Note that when performance testing any Salesforce implementation, it is vital to inform Salesforce that performance testing is planned, at which point their team will need to agree date and time windows for the tests to be run, otherwise there is a real risk the tests will be blocked as a suspected DDoS attack.

We ran two types of test on the platform:

• Load Testing normal-to-peak concurrent users to gauge response times and system behaviour.
• Stress Testing to push the system beyond its typical capacity to identify breaking points, user number limits and degradation patterns.

Results

The performance testing exercise proved highly valuable in identifying both system limitations and configuration issues that could impact the university’s enquiry handling capabilities at scale. The Peak Load Test, which excluded public form and email scenarios, uncovered a critical bottleneck in staff enquiry processing when student and staff journeys were run concurrently at high load. In this combined scenario, only 67 out of 300 staff enquiries were processed successfully.  Performance degraded further at higher loads; at just 30% of peak, only 17 out of 900 enquiries were processed. However, when run in isolation, Staff Enquiries (2) and Staff Processing (2) completed without error, indicating that the issue lay in resource contention or concurrency handling under mixed loads.

In contrast, the Student Enquiry Peak Test confirmed that the solution was capable of supporting high-volume student interactions in line with business requirements. The system successfully processed 21,000 student enquiries within one hour in the UAT environment, with 90% of response times under 3 seconds and no signs of performance degradation, demonstrating strong scalability for student-facing scenarios, providing a high level of confidence in the solution to the university.

The testing also surfaced a configuration-related issue affecting variable handling between parallel thread groups in JMeter, which led to inconsistent behaviour during combined scenario execution. This was traced to shared variables across scenarios running on the same machine. The issue was resolved by segregating scenario execution across separate load generators, confirming that the root cause was not application-related but rather an environmental constraint. The resolution ensured greater reliability and consistency in following test cycles.

Overall, performance testing provided essential insights into system behaviour under stress, confirmed that core student enquiry processing is scalable, and pinpointed areas for optimisation and configuration refinement to improve staff-facing processes under load.  When considering the results, it was confirmed that the Salesforce platform was performing reliably at agreed service levels.  When these were exceeded, there was a noticeable increase in response times, but the tests were successful at identifying the upper limits of volume / users, while remaining within vendor SLA’s.

Conclusion

The testing exercise provided crucial insights into the capabilities and constraints of the Salesforce Education Cloud platform in the university’s specific context. Armed with this data, the university was able to:

• Confidently proceed with the full-scale rollout of Education Cloud
• Optimise configurations and plan for future enhancements
• Ensure a seamless experience for students, staff, and administrators during peak engagement periods

The testing conducted by Prolifics Testing was instrumental in validating the performance of Salesforce Education Cloud for their high-demand academic environment. The test not only ensured a successful deployment but also laid the foundation for scalable digital transformation in the university’s learner engagement strategy, with the advantage of being able to repeat these tests in the future, aligning with upgrades and infrastructure changes.