Down USOnly the measurements decide.
Nobody enters, pays or votes. The figures behind the monthly reliability report pick the winner, by the same rules, for every company. Sponsors and advertisers can’t change a result.
Highest measured uptime wins
In each sector, for the period, from the same figures as the monthly reliability report. Measured around the clock and confirmed from the US.
Measured on enough days
At least 80% of the days we measured anything in that period. So a service added late, or one we couldn’t reach, can’t win on a few lucky days.
Hard-to-measure sites are left out
If fewer than 90% of our checks passed but no outage was ever confirmed, the figure is our measurement problem, not their record.
Ties: fewer outages, then less downtime
Then more of our checks passed, then more days measured. Still exactly level on all five? Joint winners, named together.
The top ten are named
Each sector names ten: the winner, 2nd and 3rd place, and seven runners-up. Every one gets a certificate.
Published on the 2nd, frozen
A month’s awards on the 2nd of the next month; a quarter’s on the 2nd day of the next quarter; a half-year’s on the 2nd day of the next half; a year’s on 2 January. Never before. Once out, an award never changes.
What counts
The measured uptime of the service’s website or API, checked around the clock. Outages count only once a check from the US confirms them.
What doesn’t
Advertising, sponsorship, being listed by the business itself, customer reports on their own, international services, and Extrafemi Labs’ own products.
When each award is published
The 2nd of the next month. October’s winners: 2 November.
The 2nd day of the next quarter: Q1 on 2 April, Q2 on 2 July, Q3 on 2 October, Q4 on 2 January.
H1 on 2 July, the 2nd day of H2. H2 on 2 January, the 2nd day of the next H1.
2 January of the next year.
All at 09:00 US Eastern time, on the site, by email to the waitlist and on our Telegram channel at the same moment.