P&T leadership is looking for additional tools to better understand expected changes in our existing metrics reporting. With enough metric history we should eventually begin develop more sophisticated forecasts but for now we should experiment with 'lower' effort forecasting methods.

For monthly reporting products use forecasting methods inspired by naïve forecasts add the ability visualize and present a projection of the upcoming period month of data in the time series.

One hypothetical example related to the readers metric:

There were 1.51B unique devices(UD) reported in July 2023 compared to 1.65B UD reported in July 2022 and 1.67B UD reported in Aug 2022. Based on these observed values there was roughly a 1% increase in this metric between July and August 2022. Using a our basic, naïve inspired projection methodology, we would estimate another 1% increase in UD between what we observed in July 2023 and what we would expect to observe in Aug 2023. This would result in a projection of 1.53B UD in Aug 2023.

Note this is just one example method of applying a low effort naïve forecast and we should consider other methods of arriving at our goal of better understanding expected changes in the metric.

Where possible, forecasts should be applied to all metrics currently presented in monthly reporting, inclusive of additional content metrics to be added in T346345

Forecasts should also be applied to Quarterly reporting products using a similar method applied quarterly.