The issues concerning timeliness are really good ones. In many ways, if we take a traditional survey-based approach, we're really constrained by the ability of our respondents to provide us with the information in a timely manner. For instance, if I take a look at our sub-annual statistics on energy, I see that we obtained information from provincial authorities in order to reduce response burden. In some cases, provincial authorities won't allow us to release data until they've released data on their jurisdiction themselves. In other cases, we're waiting some time in order to have data released.
I think we have to continually take a look at the different ways we approach data. In the agriculture space, we're currently working on a pilot project with Agriculture and Agri-Food, the Alberta Ministry of Agriculture and Forestry, and the provincial body that deals with crop insurance. In that project, we're starting small. We're attempting to take a look at whether or not we can combine data from all of these sources and model crop yield data on a weekly basis that would be available a day or two after the end of the reference week. I think that's the type of environment that we need to go into, and for sure, we're not there yet.