I'll be repetitive and say that the lesson is that the next iteration should build in an evaluation mechanism so that we can identify how to deliver the highest benefits possible to Canadian households from decarbonization and electrification.
Our findings that some insulation investments did not save any energy suggest that a targeting strategy can be adopted to give a targeted kind of retrofit pathway, just like the current program does. However, we can change that from the model-based approach to a hybrid approach where you're also using data from past participation. It has to be individual-level data and anonymized, with an anonymization procedure, which is very possible to put into place, so that there's no identifiable information. Build that into the program to learn how to get a targeted retrofit package that will maximize the savings to an individual house.