Manufacturing Example: Toyota uses "Overall Equipment Effectiveness $($OEE$)"$ as a key manufacturing KPI, measuring the percentage of planned production time that is truly productive. This differs from simply tracking "machine uptime hours" because OEE considers quality, performance, and availability together to assess strategic manufacturing effectiveness $($Nakajima$,1988).$

Software Development Example: Netflix measures "Time to Recovery" $($how quickly they restore service after an outage$)$ as a critical KPI rather than just counting "number of deployments per week." This KPI directly links to their strategic goal of maintaining $99\mathrm{.}9\%$ service availability and customer satisfaction $($Humble et al$.,2010).$

Follow$-$up questions:

In your vendor situation, how might you design a leading KPI $($predictive$)$ versus the lagging KPIs $($reactive$)$ you're currently using to potentially prevent performance issues before they impact your operations?

Given that your problematic vendor handles location$-$specific work that's hard to replace, what alternative KPIs could you implement to better manage this single$-$source supplier risk?