Counting losses tells you what already went wrong. These indicators show whether your cargo security program is actually reducing risk before it does.
By the Sentrick Cargo team · Published October 10, 2026
Many shippers and carriers judge their cargo security by one number: what was stolen last quarter. That figure matters, but it arrives late, swings with luck, and says little about why. A program can look successful in a quiet quarter while its controls quietly decay, or look like a failure after one unlucky incident. Better measurement combines outcome numbers with indicators that show whether the controls themselves are working.
Start with the basics, tracked consistently: number of incidents, value and share of loads affected, recovery outcomes, and claims results. Break them down by lane, facility, carrier, commodity and time of week so patterns are visible. The goal of this layer is honest history. It is also the evidence insurers and partners will ask for, as discussed in our piece on cargo theft insurance claims.
Leading indicators measure the behaviors that prevent loss, so they move before incidents do. Useful examples include:
Seal and handoff compliance. The share of loads where seal numbers were recorded and verified at each handoff. Gaps here point to weak points in seal verification and in chain of custody.
Pre-dispatch review rate. How many shipments received a risk review before leaving, using an approach like pre-dispatch risk scoring, and how many high-risk loads got extra controls.
Tracking coverage. The share of loads with active monitoring end to end, and how often data goes dark for longer than expected.
Dwell and stop compliance. How often loads stop in unplanned places or exceed planned dwell time, which links to geofence design.
Detection without response is only a record. Measure the time from alert to acknowledgment, from acknowledgment to action, and from action to resolution. Track how many alerts were confirmed as real, false, or unresolved. If response times lengthen or acknowledgments slip, the program is weakening even if no cargo has gone missing yet.
A monitoring system that cries wolf trains staff to ignore it. Track the false-alert rate and the share of alerts that led to a useful action. Falling quality is a prompt to retune rules, not to add more alerts. The same logic applies to the distinction between simple location data and context-aware analysis explored in GPS versus behavioral AI.
Compare carriers, drivers, yards and warehouses on the same measures, fairly and with context. Look for consistent differences rather than single events. Use the results to support conversations and improvement plans, and to reward partners who maintain strong controls. Avoid ranking on thin data; a handful of loads is not a pattern.
It is easy to collect dozens of metrics and act on none. Choose a short list, perhaps five to eight, covering outcomes, leading indicators, response and alert quality. Define each one precisely, name an owner, set a review rhythm, and write down what action a deteriorating number should trigger. If a metric never changes a decision, drop it.
Several mistakes recur. Rewarding low reported losses can encourage under-reporting. Comparing lanes with very different exposure can mislead. Treating a single quiet period as proof of success hides drift. And changing definitions midstream breaks trend lines. Document how each number is calculated, and review the data for quality as carefully as the results.
Pick one lane or facility, baseline four or five measures for a month, and review them with the people who run the process. Fix what the data shows, then widen the scope. Cargo security improves most when measurement feeds routine decisions, not only post-incident reports.
Loss figures tell you the score after the game. Pair them with indicators of control compliance, response speed and alert quality, keep the set small, and use it to make decisions. That is how a cargo security program shows it is working before the next incident tests it.