How can I tell if a hosting company’s uptime record is real?

Begin with the source. Was the record produced by the hosting company’s own tools, by a monitoring service outside its network, or typed up by the seller? A system that fails can also fail to record that it failed. Ask for the raw logs from an outside monitor and ask what counts as downtime in the seller’s definition.

Then cross-check. Pick a few dates when the record shows trouble, and look for the incident notes, customer messages and credits given. Pick a few dates when the record shows perfect service, and look for complaints anyway. Planned maintenance, partial outages and slow responses are often left out of a headline number, so ask how each was treated.

Finally, talk to the people involved. Ask the engineers what the worst incident was and what changed afterward. Ask for the largest customers’ views, with the seller’s consent. Check what the company promised in its service terms and whether it ever paid compensation. If the record is clean but the credits log is long, the story has a gap. Ask whether the record was ever adjusted after an incident, and who had permission to adjust it.

Key takeaways

• Find out who produced the uptime figure and how downtime was defined.
• Cross-check logs against tickets and customer messages.
• Ask whether partial outages and maintenance were counted.
• Compare promised service terms with credits actually paid.