The energy consumption of a company can vary even without modifications of the equipment. Timetables, processed quantities and environmental conditions affect the result. Comparing two bills only through the amount paid is likely to mix consumption and energy price.
I propose to distinguish the amount consumed and spent, indicating the period. If the activity changes, it may be useful to flank an indicator relative to opening hours or units produced, as long as that report really describes the process. Normalization does not eliminate all differences between periods.
A hypothetical example is a store that reduces consumption because it remains open for less hours. The overall figure falls, but does not show an improvement in plant efficiency. Similarly, more activity can increase total consumption even with an improvement per single operation.
The assessment of an intervention should start from a documented basis and comparable conditions. For technical choices we need competent checks on the specific case. At management level, consistent monitoring helps to recognize anomalies and ask precise questions. The useful result is to understand which part of the variation depends on the work done and which may require further study.