The Challenge of API Quality
There are many metrics you can apply to the performance of an API, including latency, availability and pass rates. However, how do you know they are measured in the same way? Are they applied in the same way?
APImetrics has specialized only in API performance, and in how it applies to run-time governance and life-cycle management, since 2016. It is a leading provider of data-centric performance analysis for the ecosystem.
What Is CASC?
CASC stands for Cloud API Service Consistency. APImetrics realized early on that there are many ways to measure API performance, including availability, uptime, pass rate and latency. All of these feed into the quality of your run-time solution, including how it works for the end user from where the application is based.
APImetrics took these measures and used its database, the largest in the world of real API call performance, to identify the issues that really affect API users. Key to the experience was consistency. An API can be slow, but you can build for that. You cannot build for inconsistent service, and that is what the score detects and warns you about.
How Does It Work?
APImetrics analyzes performance by location. It looks at whether pass conditions are met and at how consistent performance is over time, from the locations the calls are made from. It then blends these into a score out of 10. The closer to 10, the closer to perfect.
What does the score mean?
| CASC score | What it means |
|---|---|
| 8.00+ | The API is performing consistently with no performance issues. |
| 6.00 to 7.99 | The API is generally performing adequately, but some performance issues need attention. |
| Below 6.00 | The API is performing poorly. There are serious performance issues in need of urgent remedial attention. |
Example of How It Works
A major IT solution provider noticed that the CASC score for one of its APIs had dropped relative to other APIs. The API still had a 100% pass rate and an acceptable latency. When APImetrics drilled into the performance, the latency had changed relative to previous performance. This indicated that something was not working correctly in the load balancer infrastructure. The client was able to identify the issue quickly and implement a fix.
The Cost of Poor Performance
Through deep analysis of client performance, APImetrics has identified the cost of poorly performing APIs to an enterprise.
| CASC score | Monthly incidents per endpoint | Overhead annually (average) |
|---|---|---|
| 9+ | 1 to 4, typically minor | Up to 400 hours a year |
| 8+ | 2 to 8, minor but occasionally serious | Up to 800 hours a year |
| 7+ | 4 to 12, needs attention | Up to 1,200 hours a year |
| Below 6 | 12 or more, unacceptable performance | 2,000 or more hours a year |
APImetrics' approach to logging and data storage reduces the time taken to identify performance changes, to understand the impact on end users, and to fix them.
Learn more about CASC scoring at docs.apimetrics.io.
Features
- Security: Integrated OAuth handler with JWT support, FIPS 140 compliant HSM and MTLS support, and automated token refresh.
- Call types: All REST and SOAP APIs, and individual or multi-step API calls with conditions and assertions set at a call or project level.
- Scheduler: Run API calls at frequencies from one per minute, from 85 data centers across AWS, Azure, Google and IBM Clouds.
- Network breakdown: All TCP call steps logged and stored.
- Importer: Postman Newman runner, OAI importer and cURL importer.
- Data storage: All calls, metadata and returned content stored for as long as you are a customer.
- Quality scoring: Cloud API Service Consistency scoring (CASC).
- Insights: AI-generated performance insights, including call outlier detection.
- Reporting: Custom report builder with groups and public dashboard options, graphing, heatmaps and regional performance data sets.
- API access: Full API access to all results and data sets.
- Alerting: Webhook integration with Teams, Slack and other DevOps workflow tools, and email alerts.
- Organization and project management: Collections arranged by project with role-based management, and SSO/SAML support.
APImetrics Has You Covered
APImetrics' API-native runtime monitoring platform integrates with many leading developer operations suites to provide API-centric views of:
- Real-time API performance from more than 80 locations on six continents and four clouds
- Fully integrated security monitoring designed and built for the financial services industry
- Machine learning-based analysis driven by a database of more than a billion real API calls
- Integrated reporting, analysis and alerting
- 360-degree visibility and competitor comparisons with CASC (Cloud API Service Consistency) scoring
Many of the world's largest banking groups and financial services providers use APImetrics to ensure their Open Banking solutions are functional and compliant with local regulatory frameworks. APImetrics' patented technology allows clients to produce regulator-ready reports for financial services providers in the UK, in Europe under PSD2, and around the world.
Contact us for a consultation today at www.apimetrics.io, [email protected] or +1 833 274 6389.
Glossary
The term used on this one-pager, defined for teams that compare API quality.
- CASC (Cloud API Service Consistency)
- The APImetrics scoring method that produces a score out of 10 for how consistently a cloud API performs. It blends pass conditions, consistency over time and performance by location.
About APIContext
APIContext eliminates blind spots for enterprises across the digital delivery chain with proactive synthetic monitoring, performance analytics, and automated conformance validation. Our platform delivers actionable insights so connected systems perform and conform—ensuring every interaction is trusted, secure, and compliant.
What are your APIs saying to AI? Contact us to find out.
Download the PDF
The one-pager is also available as a PDF. Complete this form and we will send it to you.
- How the CASC score is calculated from pass conditions, consistency and location
- What each CASC score band means for an API
- How a poorly performing API adds incidents and overhead for an enterprise
- What a change in latency can reveal about your infrastructure