An operator with 18 chargers across two sites found one site earning about 40% less revenue per connector per week despite similar session volumes. Session completion rate explained it: site A completed 97% of sessions normally, versus 81% at site B. Nearly one in five sessions at site B ended abnormally, leaving no payment collected and likely losing a driver. Similar total sessions had hidden the issue.
Summary-only dashboards hide the operational patterns behind revenue and driver experience. Review these four metrics weekly for each connector to spot those patterns before they become serious problems.
KPI 1: Session Completion Rate
Session completion rate is the share of initiated charging sessions that end normally, through vehicle completion, an app or RFID stop by the driver, or an operator's remote stop, rather than abnormally through hardware fault, connectivity loss, unexpected charger reset, or payment failure mid-session.
At network level, 95% or above is healthy. A connector below 90% for a full week has a recurring fault worth investigating. Timing matters: a drop only during 6 to 9 PM may point to load-related thermal cutout or voltage issues during high concurrent draw. A steady time-of-day pattern regardless of load may indicate a scheduled charger task or firmware process interrupting sessions.
Set a clear definition of "abnormal session end" first. A driver stopping through the app is normal; a charger going offline mid-session is abnormal. Remote stops for known reasons, such as maintenance or overstay enforcement, may be left out of the denominator based on your operational context.
KPI 2: Average Session Duration
Average session duration shows whether drivers leave before charging finishes and whether fully charged vehicles keep blocking connectors.
On a 7.4 kW AC charger, a typical two-wheeler with a 3 to 4 kWh battery reaches full charge in about 30 to 45 minutes. A mid-range four-wheeler EV with a 30 to 40 kWh battery and a compatible onboard charger needs 4 to 6 hours. If a corporate site's average stays above 8 hours, some drivers are likely parked through the workday after charging ends around hour 5. Their connectors remain blocked for the final 3 to 4 hours with no revenue.
Where a site serves both vehicle types, measure two-wheelers and four-wheelers separately; one combined average is meaningful for neither. Record the distribution too: 80% of sessions under 5 hours and 20% above 10 hours produces a 6 hours average that hides afternoon connector blocking.
Excessive duration is usually addressed with an overstay fee, following a notification when charging reaches 90% state of charge or a defined maximum session duration. Your charger must support mid-session notifications, and the billing platform must apply overstay charges automatically.
KPI 3: Revenue Per Connector Per Day
Revenue per connector per day is the unit economics measure. It shows whether a site covers variable costs, mainly electricity, and contributes toward fixed costs such as lease, platform fee, and maintenance.
For AC charging, a 7.4 kW charger at 50% utilization across 12 active hours delivers approximately 44 kWh per day. At Rs 14/kWh, that equals Rs 616 gross revenue per connector per day before electricity and other expenses. Actual revenue of Rs 180, suggests around 15% utilization, lower-than-expected pricing, or many sessions ending without payment.
Measure this per connector, not only per site. A site with 10 connectors and Rs 2,800 daily revenue could have 4 connectors earning Rs 600 each and 6 earning Rs 67 each. Those 6 may have a hardware fault, a poor location, such as the least visible or farthest from the entry, or an unsuitable connector type. A site total will not show this.
KPI 4: First-Authorization-Attempt Success Rate
This undertracked metric best reflects the driver experience. It measures the share of charging attempts, a QR scan or RFID presentation intended to start a session, that result in a session starting within 30 seconds.
Below 92%, roughly one in twelve drivers attempting to start a session encounters a failure: payment gateway timeout, OCPP processing delay, no response to a remote start command, or authorization failure caused by a billing issue. The driver may retry, contact support, or give up. At least some do.
Failures fall into three buckets. Payment gateway latency is the largest: a driver's UPI transaction takes more than 30 seconds to confirm, so authorization times out. Next is OCPP command latency: the CSMS sends the remote start command, but the charger takes 40 to 60 seconds to answer with StartTransaction, and the driver has walked away. Third is charger state mismatch: the CSMS shows Available while a local fault blocks session start.
Your platform must log every driver-initiated attempt, not only completed sessions, and pair each attempt with a timestamped outcome. If the CSMS logs completed sessions only, the denominator is missing and silently failed attempts remain invisible.
Weekly Review of These Metrics
A weekly review can be brief. For a network of 10 to 30 chargers, spend 15 minutes scanning a report with these four metrics per connector and flag any connector more than one standard deviation below the network average. This can catch issues weeks before complaints.
Suggested starting thresholds: completion below 90% over any 7-day period starts an investigation; revenue per connector per day below 40% of the site average starts a root-cause check; first-authorization-attempt success below 92% prompts payment gateway and OCPP log review; average duration above 8 hours at an office site prompts an overstay fee review.
Metrics need a response process to be useful. Their value lies in alert thresholds and the workflow around them. An alert nobody acts on will not improve the network.