Accruals based on weighted average methodology

Currently, Envizi applies a 30% variance threshold when comparing the daily consumption rate from the prior month in the current year to the corresponding month in the previous year. For example, when calculating accruals for July 2026, Envizi compares the daily rate for June 2026 against June 2025. If the variance exceeds 30%, Envizi disregards the weighted average methodology and instead uses the June 2026 daily rate to calculate the July 2026 accrual.

This approach can undermine the purpose of using a weighted average methodology, as it effectively defaults to the most recent month's daily rate and does not adequately account for seasonal consumption patterns. Most of the time, succeeding periods are missing so Envizi will always default to using daily rate if the threshold is exceeded. For example, when accruing for July 2026, Aug 2026 bill is also missing.

This issue is particularly evident for natural gas consumption, which is highly seasonal and typically increases during winter months. In cases where May consumption is lower than the previous year, the resulting accruals for June may also be understated, despite June historically being a higher consumption period. As a result, accrued consumption may not reflect expected seasonal trends.

For clients such as Coles, this has led to the underreporting of gas usage during sustainability reporting periods. Since their sustainability report need to be finalised in early August, a significant portion of June data is still accrued at that time. Any understatement in June accruals therefore directly impacts reported emissions and energy consumption.

I propose removing the 30% variance threshold and consistently applying the weighted average methodology. In practice, missing invoices are common immediately following the accrual period, meaning the 30% threshold is frequently triggered and the system defaults to a simple daily rate calculation. This reduces the effectiveness of the weighted average approach and can result in less representative accrual estimates for seasonal commodities such as gas.