Ask what is being counted
A website responding, a customer signing in and a programme playing successfully are different success conditions. A service can define a measured indicator around one of them, but it should not silently stand for every other stage of the viewing chain.
Google's SRE guidance discusses service-level indicators and objectives as explicit measures and targets. Citing that engineering framework does not establish that a particular supplier has an objective, meets it or offers a contractual remedy.
Work the time arithmetic transparently
For an illustrative thirty-day window, total time is 30 × 24 × 60 = 43,200 minutes. A time-based 99.9% availability figure allows 0.1% of that window outside the defined successful state, which is 43.2 minutes under that simple calculation.
Those numbers are a mathematical example, not an IPTV Canada commitment or outage record. Request-based metrics, exclusions, maintenance treatment and other definitions can produce a different interpretation. Do not import the calculation into an unknown contract as a compensation promise.
Keep the observation coverage visible
A customer checking three evenings has observed those samples, not every intervening minute or every programme. An outage log should retain actual failure and recovery observations with uncertainty between them. Do not invent continuous monitoring or fill unseen periods with assumed success.
An illustrative successful homepage check during a media failure does not prove the viewing task was available. Keep the resource, device, player and stage with the incident. A single endpoint can be a useful indicator only within its stated scope.
Separate a target from a guarantee
Ask for the supplier's actual published or agreed conditions if a service commitment matters. A marketing percentage, technical objective and enforceable contract term are different claims. The approved subscription totals do not establish an uptime promise or remedy.
After a correction, retest the original task and report the actual window observed. The useful conclusion is a defined metric supported by real evidence, with any contractual question handled separately. A neat percentage should make the claim easier to assess, not conceal what was never measured.
Related questions
Can a short trial prove 99.9% annual availability?
No. It provides limited observations under its actual conditions, not continuous annual evidence.
Sources and scope
Scope notes distinguish published documentation, supplied information and editorial methods. They do not report a hands-on service test.
- Google SRE: Service-level objectives — Metric and objective definitions; no evidence that IPTV Canada adopted or met an objective.
- CIRA: How the Internet Performance Test works — What the test measures; examples are not measured service performance.
