HTTP Quick Check — Status Code and Response Time
This requests a URL and reports two things: the status code returned and how long it took. It is the least sophisticated check on this site and the one that should probably run most frequently, because its cost is close to zero and it catches the most common failure there is.
Why run it
Most outages are boring. The server is down, the process died, the certificate expired, a deploy returned 500 for everything. A simple request catches all of those immediately, and its simplicity is a virtue — there is very little to misconfigure and almost nothing to produce a false positive. The response time it returns is also a useful leading indicator, since a slow creep upward frequently precedes a hard failure by hours or days.
Why automate it
Automation is the entire point; running this by hand is just visiting the site. The value is in frequency and memory: checking every minute means the gap between failure and knowing is a minute, and storing every result means you can answer how long an outage lasted and whether it has happened before. Be clear about what it does not cover, though — a 200 response proves a server answered, not that anyone can log in or check out, which is a different check and a different page.
What it gives you
The cheapest possible answer to the most frequently asked question. It will not catch a broken checkout, but it catches the outage, and it catches it within a minute rather than whenever somebody complains.