Software
Performance
Testing
Solutions for Technology Companies
PFLB runs SaaS performance testing for software companies whose customers share one platform, where a slow release or an onboarding peak reaches every tenant at once.
Below: what PFLB tests, how an engagement runs, and published results from SaaS, database and mobile clients.
Unmatched Expertise in
Software Load Testing
PFLB load tests SaaS platforms, APIs and microservices, web applications and proprietary IT systems. The work covers the paths users click and the background jobs and queues they never see, and measures application performance, reliability and scalability under the varying loads your customers actually create, so you can ship features without compromising performance.
How to Get Started
Case Study
800 Registrations in 5 Minutes
on an E-Learning Platform
Every September, FolderWave expects at least 1,000 students to register simultaneously on MEFA Pathway after major online events. PFLB emulated both peak scenarios: 1,000 questionnaire submissions per hour on College Board, and 800 users registering within 5 minutes on MEFA Pathway. Under stress, the basic configuration of one service showed high response times, root cause analysis aimed the fix at it, and FolderWave's team reran the tests themselves while tuning. The peak load was successfully managed.
This case study is for SaaS teams whose onboarding peak is already on the calendar.
PFLB Client’s Review
Ready to
Get
Started
with Load Testing?
Dive in Our Knowledge Base.
Software engagements with numbers.
Six published case studies - two of them commissioned by vendors who needed an unaffiliated party to measure their own product.
Frequently Asked
Questions
What is SaaS performance testing, and which testing types does it include?
SaaS performance testing, as PFLB runs it, measures how a multi-tenant platform behaves when many customers use it at once: response time, error rates and resource usage under realistic user loads. Performance testing for SaaS applications combines several testing strategies. Load testing checks the expected peak. Stress testing finds the breaking point. Spike testing reproduces a sudden surge, such as an onboarding wave after a launch. Soak testing, also called endurance testing, holds load for hours to detect memory leaks and slow performance degradation. Scalability tests show whether autoscaling and extra cloud services actually add capacity, and resilience testing shows how the platform recovers when a node or a dependency fails. Broader SaaS application testing also covers functional testing, integration testing, API testing, security testing to identify vulnerabilities, and manual testing or automation testing across multiple browsers; this page is about the performance part.
How does PFLB approach testing SaaS applications with many tenants?
Multi-tenancy is the hardest of the SaaS testing challenges, so PFLB models the tenant mix rather than one average user. Most SaaS platforms run multiple customers on shared infrastructure, so the test has to show whether one tenant's spike slows everyone else, whether tenant isolation holds under load, where shared resources run out first, from database capacity to connection pool exhaustion, and how the system behaves when external services slow down. For a membership SaaS moved to AWS, the load profile came from the platform's own IIS logs, and EBS burst credits had to be exhausted before every test, or the results measured a credit instead of the system.
Can SaaS performance testing run inside our CI/CD pipelines?
Yes, and SaaS apps that release often need it there. In a SaaS application testing strategy, PFLB adds automated testing of performance to the CI/CD pipeline, so each build runs the same test scenarios and performance regressions are caught when they are introduced, next to your unit tests, regression tests and functional testing. Scripts have to keep up with the product: on an ordering platform shipping 2–3 releases a day, PFLB updated load scripts continuously because any update could change the system's processes. In a bank's Jenkins pipeline, PFLB's quality gates either stopped a release or advanced it, and the time needed to install an update dropped 5×.
Which performance metrics should a SaaS company watch?
PFLB reports the key metrics your customers feel and your SLA names: response time at the 95th percentile rather than only average response time, error rates, and throughput. On the hosts it tracks CPU, memory usage and other resource usage, because memory leaks do not show in a twenty-minute run. Together these tie test results to user expectations, user satisfaction and the commitments you make to customers.
How do you identify performance bottlenecks rather than just report them?
PFLB finds potential performance bottlenecks by correlating load with what the system was doing at that moment, with monitoring next to every run, then reproduces the failure in isolation to prove the cause. A report that only says the system is slow is half the job. For FolderWave, high response times under stress traced back to the basic configuration of one service, and root cause analysis aimed the fix there rather than at symptoms.
Which performance testing tools do you use?
PFLB works with JMeter, k6, Gatling and LoadRunner, and with its own platform when you want scripts and test results to persist between engagements. The right tool is the one your team can keep running after the engagement ends, and protocol support matters: for NOV, PFLB added HTTP/2 testing to its platform so the product could be tested the way it actually communicates. Effective testing depends less on the tool than on the workload model behind it.
Why do SaaS companies hire an outside performance testing service provider?
Because the SLA you sign with customers is a number, and PFLB measures whether the platform holds it before a customer finds out. Independent results complement your own quality assurance and carry weight in a sale: two of the six engagements on this page were commissioned by vendors who needed an unaffiliated party to measure their own product. PFLB's full performance testing services cover the cycle from a single test to ongoing testing of every release.