Synthetic monitoring with browser and API checks from 50+ locations

What is synthetic monitoring?

Synthetic monitoring proactively tests applications and APIs by running scripted requests or browser interactions on a schedule, from defined locations, and checking the results against expectations. Because the traffic is generated by the monitoring system rather than by real users, it is called synthetic. A check might be as simple as "the health endpoint returns HTTP 200 within 300 ms" or as complete as "a user can log in, add an item to the cart and reach the payment page".

Synthetic monitoring is active monitoring. It produces consistent, comparable measurements because the same steps run in the same way every time. That makes it ideal for availability tracking, SLA reporting and regression detection.

Why synthetic monitoring matters

Real user monitoring only sees problems when users are affected. Synthetic checks find them first, including outside business hours, in regions with little traffic and on flows that are used rarely but matter a lot, such as password reset or partner APIs.

• Early detection: a broken deployment or expired certificate is caught by the next check run, not by the first customer complaint.
• Regional visibility: running checks from many locations shows whether an issue is global or limited to one region, ISP or CDN edge.
• Third-party coverage: verify that payment providers, identity providers and other external dependencies respond as expected.
• Objective SLA data: consistent measurements are the basis for availability reporting and SLOs.

Browser checks vs. API checks

Aspect	API checks	Browser checks
What runs	HTTP requests, single or multi-step	A real browser executing a scripted journey
Best for	Endpoints, microservices, partner APIs, health checks	Login, search, checkout and other critical UI flows
Typical assertions	Status code, response time, headers, JSON body values	Element visible, text present, navigation completed, timing
Frequency	High, often every minute	Lower, for example every 5 to 15 minutes
Failure evidence	Request and response details, timing breakdown	Screenshots, console errors, network waterfall

Most teams combine a dense layer of API checks with a smaller number of browser checks for the journeys that generate revenue or sign-ups.

Synthetic monitoring with Obseria

Obseria runs API and browser checks from 50+ locations worldwide and from private locations inside your own network. Checks can be created in the UI or defined as code and deployed through your CI pipeline. Each test request carries trace context, so a failed or slow check links directly to the distributed trace in the backend, the logs of the services involved and the deployment that preceded the failure.

Results land in Obseria Lake together with RUM data, traces and metrics, and synthetic results are retained for 13 months, long enough for year-over-year availability reports. You can build SLOs on check success rates, feed status pages automatically and let Obseria Intelligence correlate failures with topology changes. API check runs are billed at $0.20 per thousand runs, with no per-location or per-user fees.

Because checks are defined as code, they can run as part of your deployment pipeline as well. Trigger a set of API and browser checks right after a release, compare the results with the previous run, and stop or roll back the rollout automatically when a critical journey fails.

Synthetic monitoring best practices

1. Cover critical journeys first: login, search, checkout, sign-up and the most important public APIs.
2. Test from where your users are and from at least two locations per region to separate local network issues from real outages.
3. Require multiple failures before alerting, for example two consecutive failures or failures from several locations, to avoid noise.
4. Use dedicated test accounts and data so synthetic traffic never touches real customer records.
5. Keep checks in version control and update them together with the application to prevent false alarms after UI changes.
6. Combine synthetics with RUM: synthetics tell you whether something works, RUM tells you how it works for everyone.