The problem shows up in a missed shipment
A regional distributor loses a pallet because the warehouse system does not flag a temperature excursion in time. The loss is $12,000 for that load. The same failure repeats three times in a quarter. The owner asks how to know the size of the problem without hiring a consultant.
The answer is three numbers you can track yourself: time to detect, time to recover, and repeat rate of the same failure. These three metrics form the core of ECO-249 operational resilience measurement for any small business.
Why these three numbers matter for a distributor
Time to detect tells you how long a problem lives before anyone sees it. Time to recover tells you how long the operation stays degraded after detection. Repeat rate tells you whether the same root cause comes back. Together they turn a vague feeling of “we have issues” into a concrete loss figure you can price.
If a temperature alarm is missed for four hours and the recovery takes another two hours, the window of exposure is six hours. At a throughput of $2,000 per hour that window costs $12,000 per incident. Three incidents a quarter cost $36,000. That arithmetic uses only your own throughput and incident count.
Instrumenting time to detect with tools you already own
Most warehouses run a warehouse management system that logs sensor readings. Export the log for the last 30 days. Identify the timestamp when the sensor first crossed the threshold. Identify the timestamp when the first human or automated alert was created. Subtract the first from the second. That difference is your detection time for that event.
Repeat the calculation for every excursion in the export. Average the results. If the average is 3.2 hours you now have a baseline. No new software is required. The same method works for any system that writes a timestamped log — ERP, CRM, or a simple spreadsheet.
Measuring time to recover using ticket timestamps
Your ticketing system (Jira, ServiceNow, a shared inbox) records when an incident is opened and when it is marked resolved. Pull the two fields for each incident tied to the failure mode you are tracking. Subtract open time from resolve time. That is the recovery time for that incident.
Average across the same 30‑day window. If the average recovery is 1.8 hours you can add it to the detection average. The combined exposure per incident becomes 5 hours. At $2,000 per hour that is $10,000 per incident. The arithmetic is transparent and can be verified by any team member.
Calculating repeat rate from incident history
Tag each ticket with a root‑cause code. Count how many tickets share the same code in the period. Divide that count by the total number of tickets for the period. The result is the repeat rate for that cause.
Example: 40 tickets in a month, 12 carry the code “temp‑sensor‑drift”. Repeat rate = 12 ÷ 40 = 0.30. That means three out of ten incidents are the same failure returning. A repeat rate above 0.20 signals a systemic fix is needed, not just a one‑off patch.
Putting the numbers into a simple dashboard
Create a three‑column sheet: detection average, recovery average, repeat rate. Update it weekly. Plot the three lines on a single chart. When detection drifts up, investigate sensor coverage. When recovery drifts up, review escalation paths. When repeat rate climbs, launch a root‑cause project.
This dashboard costs nothing beyond the time to pull the exports. It gives the owner a hit‑rate view: each week you see whether the coverage against recurring loss is improving. The goal is not replacing staff; it is giving staff a clear target so they can act before the loss repeats.
When to bring in outside help
If detection stays above four hours after two improvement cycles, or if repeat rate stays above 0.25 despite a root‑cause fix, the problem likely spans multiple systems. That is the point where a specialist can map cross‑system dependencies faster than trial and error.
Alpha Edge offers AI consulting for your operation to design automated alerts that shrink detection to minutes. We also run a program for qualifying businesses that funds the first instrumentation sprint. Automation cannot fix a broken process on its own; it only accelerates the feedback loop you already measure.
Track the three numbers this month. Price the exposure with your own throughput. Decide whether the hit‑rate improvement justifies the next investment. The data you collect today is the only baseline that matters.