Considered for rating and invoicing but rejected. The subscription-centered model fit the required annual commit drawdown, per-meter dual rates, ramps, regional multipliers, and long-term line-by-line replay poorly, and it still would have needed a custom time-weighted gauge layer in front.
- What got in the way
- Did not cover gauge-to-usage integration, high-volume rollup grain, complex enterprise contract shapes, or replay from raw counters after long retention.
