Priceline relies on innovative proprietary architecture that combines internal and third-party partner components to offer high-performing websites and services to millions of customers. Speed, scalability, and consistency are the keys to Priceline.com's continued success.
Priceline.com partnered with Catchpoint to:
- Rapidly identify performance bottlenecks, problems, and degradations.
- Simulate user experiences from within their own network to minimize Internet noise.
- Inform and drive infrastructure changes through "What-If" scenarios.
- Establish baselines and benchmarks to track long term KPIs and set performance thresholds.
- Evaluate third-party vendors and ensure their performance via SLA monitoring and reporting.
- Create custom metrics to monitor individual servers that are behind load balancers.
Ensuring constant customer and partner satisfaction is vital for Priceline. Catchpoint, originally purchased as our main alerting platform, has become our trusted go-to-strategic partner for continuous web performance monitoring, analysis, and optimization. Not only can we more quickly identify and resolve problems, but Catchpoint’s analytics have helped us uncover many new opportunities for system and business improvement. With Catchpoint, we can do significantly more observing for the same cost.
Priceline.com – a ﬂagship site of Booking Holdings alongside sister companies such as Agoda, Booking.com, Kayak, and OpenTable – gives leisure travelers multiple ways to save on hotel rooms, rental cars, airline tickets, vacation packages, and cruises. In addition to getting compelling published prices, travelers can take advantage of Priceline.com’s famous Name Your Own Price® service, which can deliver the lowest prices available, or its Express Deals® service, where travelers can take advantage of hotel discounts without bidding.
Priceline.com's architecture requires constant active observability tests to help the Priceline operations team identify performance issues on their key pages and the infrastructure that powers them. This is fundamental to any company that seeks to provide the best possible digital customer experience, but the Priceline team has also expanded their usage of the Catchpoint platform to gain much more insight and business advantages than would be provided by basic active monitoring and alerting strategies.
In addition to using Catchpoint’s widespread active observability locations and ISPs to observe multistep transactions, DNS services, and API calls, Priceline continuously benchmarks performance with industry peers to deﬁne appropriate goals to maintain its leadership position.
Using the in-depth analytics at their disposal, Priceline can automatically correlate internal data with active observability metrics to diagnose problems and rapidly ﬁnd root causes across complex n-tier architectures, and to examine the impact of front-end code optimizations.
Additionally, the Hosts and Zones feature allows them to quickly troubleshoot underperforming components (third-party vendors, internal components, etc.), while the alerting platform makes them immediately aware of problems and bottlenecks.
On top of this straightforward approach to active observing and alerting, Priceline has greatly expanded its usage of the Catchpoint tool, taking advantage of advanced features that allow them to drive positive business outcomes across their digital platforms, delving into the areas below.
First-Mile Baseline Tests via Enterprise Nodes
To minimize the amount of “noise” from the Internet in their test results (i.e., issues with local ISPs, peering problems, etc.), Priceline deploys an Enterprise node in each of their data centers that sends tests out to the Internet, and then uses an override to force it back through their load balancers like any normal customer traﬃc.
However, because the two interfaces are just one hop apart from each other, this minimizes the amount of variability caused by the Internet, and gives them the closes view of what’s happening on their website.
Driving Infrastructure Changes via Request Overrides
When considering whether to deploy new servers, make changes to their cloud providers, or evaluate a new CDN, Priceline is able to use the request override feature within Catchpoint to create a “What if?” scenario and determine what eﬀect the proposed change will have on their end user experience.
One example of this would be by using DNS overrides to send the test to a speciﬁc CDN that’s been conﬁgured to support the new traﬃc, then running those tests in parallel with the ones that are set up with their normal conﬁguration and comparing the data directly in the Catchpoint portal to see the variant between the two scenarios.
The request override can also be used to determine the impact of a speciﬁc third-party element on the site. This is particularly useful for a company like Priceline, which features a wealth of third-party content from travel vendors. They also deploy the request delay feature, which allows them to specify a number of milliseconds to determine if there is asynchronous loading dependent upon the object, and whether its delay would aﬀect other elements on the page.
Due to SLAs that Priceline has with both external customers and vendors, they must provide regular reports for how their digital systems are performing.
Using Catchpoint’s SLA Observability tool, they are able to set tests with alerts tied to the speciﬁc thresholds determined by their diﬀerent SLAs. This allows them to track whether they’re within those service levels and receive alerts that ﬁre before the thresholds are breached.
This way, if things are headed in the wrong direction, they’re able to act on it before breaching the SLA.
Performance professionals can often lose sight of the long-term perspective in the course of their day-to-day ﬁreﬁghting operations. Priceline is able to mitigate this risk with the help of Catchpoint’s historical data storage, which they tie to hard KPIs as opposed to trend-related ones.
By setting alerts on those absolute values, they’re able to track long-term performance over time. They get alerted about gradual performance degradation even if it never triggered any signiﬁcant spikes.
With three years’ worth of stored data at their disposal, Priceline can compare long-term historical trends with short-term KPIs to ensure consistently excellent user experience.
Integrations with Key Tools
While Catchpoint has become the primary cog in Priceline’s digital experience observability eﬀorts, it’s not the only tool they use. That means it must be able to work in tandem with the others. Therefore, the integrations that Catchpoint has set up with other SaaS products have become vital to Priceline’s day-to-day operations:
- Splunk - Priceline uses a Catchpoint-provided add-on to push their data into Splunk, making it easy to overlay the synthetic test data from Catchpoint with the internal metrics available on Splunk.
- Slack - Using the Webhook API along with the Slack integration, Priceline pushes their alerts into a dedicated Slack channel, providing them with an easy way to view, share, and store reports internally without having to go into the Catchpoint portal.
- Okta - Like many enterprise companies, there are many individuals across several teams that must be able to access the Catchpoint portal. Using the Okta integration for single sign-on capability, Priceline’s administrators don’t have to worry about provisioning individual users or worrying about restricting access whenever someone leaves the company.
When Priceline.com ﬁrst implemented Catchpoint’s Digital Experience Observability solution in 2011, they initially saw a dramatic reduction in the time it takes to triage and troubleshoot problems, bringing the standard time down from hours or even days to mere minutes.
Since then, it has become as an integral part of their operations and development teams. Catchpoint has become a trusted partner, working with the Priceline team to adopt and develop new features that allow them to stay ahead of the latest innovations within the industry.