DentaQuest is the second-largest dental benefits administrator in the United States and the largest in the Medicaid space. Over 30 million members across the United States are taking advantage of its innovative solution offerings. The company is steadfast in its commitment to delivering better outcomes at lower costs. Rallying around an improved “oral health for all” charter, DentaQuest continually seeks to raise oral health awareness with an ardent promotion of oral solution effectiveness.
Applications under DentaQuest IT management represent three business units. Tests are conducted in concert with the cadence of each release, last up to four hours, and simulate up to more than half of today’s 800 total peak users. The performance engineers run tests end to end on claims intake, adjudication, and payout — from customer-facing .NET components on the frontend to the core backend systems of record.
Each test’s design requires a sign-off by the performance testing team, who compare results with the most recent run. Similar performance is considered testing success; inconsistency sends them back to the drawing board.
Production user load volume accounts for 800 peak users, most using concurrent sessions. Peak time is condensed into a narrow two-hour operating window — which includes application processing of nearly 30K claims/call tracking transactions.
Scott Schweitzer, SQA Supervisor at DentaQuest, sums up the application’s criticality to the business succinctly: “An issue with the application could have a potentially significant impact on our ability to process claims, answer customer calls, or otherwise provide optimal customer service promptly. As an organization committed to service excellence, this would be unacceptable. We needed a better way to ensure this would not happen.”
DentaQuest found that performance testing with Visual Studio was time-consuming and produced inconsistent results. “We tried to use [Visual Studio] for a while, but the results were unreliable,” says David Alves, DentaQuest’s performance and automation architect at the time.
As part of its evaluation to keep the current tool versus a fresh start with an industry leader, DentaQuest cites NeoLoad’s budget friendliness and ability to cover the full business stack with ease as key determinants. “We did a bake-off [against LoadRunner]. I was actually able to show how much slower things were with LoadRunner compared to NeoLoad. It’s inexpensive, it’s very capable — more capable than LoadRunner — it’ll get the job done reliably,” says Alves.
Working under a weekly release schedule, DentaQuest seamlessly aligned NeoLoad with its existing use of Dynatrace for APM (another critical solution requirement for DentaQuest). Performance and load testing focused on the vital claims processing/customer service call-handling application.
“With our old tool, releases were sometimes pushed as we could not complete performance testing on-time. Now, we have performance testing complete much earlier. We’re also not chasing the false positive test results we were used to seeing,” says Isha Satija, Lead Performance Engineer. The upshot to this, the DentaQuest team has the valuable ability to take on projects such as the Salesforce implementation without having to squeeze more hours out of the day.
According to Satija, other benefits include “faster scripting, maintenance with correlation and extraction; reports are comprehensive and pinpoint any issues.” This includes the comparison report view “allowing us to see builds side-by-side. We also have not seen any communication breakdown between controller/agent, which is a big win.”
Alves explains further: “One of the great things about NeoLoad is the fact that it’s scriptless. You can still be a power user and get the kind of logic and low-level work you need done when you need to, but it allows novices to get up and running a lot more quickly than actually writing code the way you have to with other tools. And obviously the reporting and the graphing — the analysis, which is the most important part — is top notch and makes it easy to determine pass/fail objectively. NeoLoad allows a team that’s transforming and learning to become performance engineers to do so without the tool ever becoming a hindrance — in fact, actually a thing of value.”
Regarding speed estimate, Satija estimates a “>50% faster runtime using NeoLoad versus the legacy solution.” The team also feels very comfortable with their ability to easily scale with the projected growth volumes upwards of 20%.
As the DentaQuest team considers the next chapter in their performance testing practices with NeoLoad, they feel confident that the increased productivity, testing speed, and quality will allow a more proactive role in SLA creation and attainment. Having NeoLoad as the core performance testing platform goes a long way in helping them realize the desired “shift further left” state. Alves adds, “When you’re shifting left, you’re not just testing GUIs and easy stuff like that. You’re testing low-level components and classes, and modules of functional areas — and you have to be able to do it quickly. So that’s where NeoLoad’s framework comes in. Whether it’s a .NET or Java or any other application programming language, it’s about how much NeoLoad — or any tool, really — does for you up-front. And NeoLoad has a whole lot of capabilities in that regard. They are ahead of the game there, simply because they’ve been doing it longer — and better — than anybody else.”