Service Performance

We publish these figures because we would rather earn your trust with evidence than adjectives. Every number here comes straight from our own systems and refreshes daily. Here is exactly where each one comes from and how it is worked out.

Where the data comes from

• Helpdesk tickets and resolution times – our service-management system (GLPI), where every support request is logged and timestamped from the moment it is raised to the moment it is fully resolved.
• Call answer rate and speed – our 3CX telephone system's own call records.
• Security updates – the count of security patches successfully applied across the machines we manage, from our patch-management tooling.
• Google rating – pulled live from our Google Business Profile through Google's own API, so it always matches what you see on Google.

How each figure is worked out

• Support calls answered – the percentage of inbound helpdesk calls answered live, with average pick-up time, across the rolling 30 days.
• Resolved without another customer reply – the share of requests resolved without the customer needing to reply again after we picked the request up. A request that is later reopened counts as not meeting this. One honest caveat: we do not yet detect closures where a customer simply stopped responding, so a small number of these may be non-responses rather than clean resolutions; we are building that detection before we claim otherwise.
• The within-window figures – the share of requests fully resolved within each window (one support hour, four working hours, one working day, and so on). These are measured in our support hours (Monday to Thursday 09:00 to 17:00, Friday 09:00 to 15:00, English bank holidays excluded), and time spent awaiting your reply is excluded, so the figures reflect the time genuinely in our hands. For each window, a request still open past it counts as not resolved within it.
• Migrated historical tickets are excluded – requests imported from our previous systems are not counted, because their timestamps do not reflect real resolution times.
• Security updates applied – the total number of security patches deployed to managed machines over the last 30 days.

What we include, and what we leave out

• Helpdesk only. These figures cover day-to-day support. We exclude engineering and project work, security-operations activity, automated system alerts and scheduled-maintenance notices, because folding those in would misrepresent the support experience an everyday user actually gets. Day-to-day requests remain included even when they are escalated to our Engineering team.
• Resolution, not response. Our headline times measure how long to fully fix an issue, not how quickly we send a first reply. First-response times look better and are easier to publish; we think full resolution is the number that matters.
• Rolling 90 days, refreshed daily. Always a current, moving picture, never a hand-picked good month.

What we are not showing yet, and why

We only publish numbers we can stand behind. We are deliberately holding back a few (such as first-response times and customer-satisfaction scores) until we are measuring them cleanly and have enough data to be honest about them, rather than guess. We would rather show you less and have it be true.