Uptime gets quoted as a headline number - “99% availability” - and then rarely translated into what an outage actually costs. That's the more useful question, because unlike most of the revenue side of connector economics, downtime cost is something a platform's monitoring and alerting directly controls.
Downtime costs more than the outage window
- Direct lost revenue. Every session that would have happened during the outage, gone - not deferred, just gone.
- Compounding utilization loss. A driver who arrives at a broken charger twice doesn't file a complaint - they route to a competing charger going forward, which shows up as a permanently lower utilization rate, not a one-time hit.
- Site-host relationship cost. If the site host is paid on a revenue share, an outage costs them too - and repeated outages are the fastest way to lose a hosting agreement at renewal.
- Support and dispatch cost. A truck roll to fix something that could have been resolved remotely adds a real, avoidable cost on top of the lost sessions.
A framework for pricing an hour of downtime
At minimum, the direct cost of an outage is:
Lost revenue = Sessions per hour at that connector × Average revenue per session × Hours down
That's the floor, not the full cost - it doesn't include the utilization drag from drivers who don't come back, which is harder to measure but often larger than the direct number over time.
A worked example
Using the same illustrative site as the connector revenue worked example - 4 sessions a day, 20 kWh average, $0.40/kWh - to show how the numbers move. These are illustrative figures, not a benchmark or a projection for any specific site.
| Scenario | Illustrative impact |
|---|---|
| Normal day (baseline) | 4 sessions × $8.00 avg revenue = $32.00 gross revenue |
| Connector down for 8 hours during operating hours | ~2 of 4 sessions lost → $16.00 direct revenue lost that day |
| Connector down 3 days in one month | ~$48.00 direct revenue lost, before any utilization drag |
| 10% of returning drivers stop coming back after a repeat outage | Recurring loss that persists after the connector is fixed, not captured by the direct-cost number above |
The direct cost of an isolated outage is usually manageable. The risk is a connector - or a site - that goes down repeatedly, where the utilization drag compounds every month and never shows up as a single line item an operator notices.
What actually drives downtime cost up or down
- Detection speed. A fault caught by monitoring within minutes costs far less than one discovered because a driver complained - or worse, not discovered until the next site visit.
- Remote fix rate. A reset or configuration push resolved from the operator console takes minutes. A truck roll takes hours to days, during which the connector stays down.
- Which connector fails. A single connector out of twelve at a low-traffic site costs little. The same failure at your highest-utilization site or your only connector at a location costs disproportionately more - see why per-site uptime matters more than a network average.
Why network-wide uptime hides the real risk
A network reporting 98% uptime overall can still have all of its downtime concentrated at the two or three sites that generate most of its revenue - the aggregate number looks fine while the sites that matter most are the ones failing. Tracking uptime and outage cost per site and per connector, not just as a single network-wide percentage, is the only way to see that.
This is where the platform layer has real leverage: fast fault detection, remote fixes that avoid a truck roll, and alerts routed to operations before a driver ever notices are the mechanisms that keep downtime cost low. See how CarChargeOS handles monitoring, alerting and remote operations on the platform page.
Frequently asked questions
Is a truck roll always more expensive than a remote fix?
Almost always, once you count both the technician cost and the extra hours the connector stays down while waiting for a site visit. A meaningful share of common faults - stuck sessions, configuration drift, unresponsive firmware - can be resolved with a remote reset or command instead.
How do I know if downtime is costing me returning drivers, not just missed sessions?
Watch utilization at a specific connector or site in the weeks after an outage, not just during it. A connector that doesn't recover to its pre-outage utilization level, even once it's back online, is showing the compounding cost - lost repeat drivers, not just the sessions missed during the outage itself.
Does downtime cost the site host too?
Yes, if they're paid on a revenue share - an outage reduces their payout exactly as it reduces the operator's revenue, which makes reliability part of the host relationship, not just an internal operations metric.