Yes. Well, I'll go a step further. Not only has the payback period since the beginning of the pandemic not change, but frankly, if you zoom out even longer than that, we've kept the same high-level framework on paybacks for quite some time. And if anything, we've only tightened it a little over time as we've been able to get more nuanced in how we account for cost. So that's not a -- we continue to tighten that up, but basically that what you're saying is true, we've kept it -- since the pandemic started, yes, we've kept it very tight. The thing about LTV to CAC, I'm talking about the payback, right? So they have in the roughly a year. If you zoom out and you really try to locate a lifetime value, three years, five years, pick a period, you actually find that the LTV, the IRR, is increasing because what you're seeing is repeat gets stronger. If you'll come back more and more often, as you look at years two, three, four, five; if you would calculate that in, which, of course, you wouldn't know until you can look back on years two, three, four, five, you would actually see the IRR then is increasing because we don't give that credit in that first year. We only give credit in that first year is what actually happens in the first year. So year two of a cohort stronger than year two used to be. You're actually seeing an ever improving return on your ad spend, but we won't put that into the first year. And so we will, therefore, not put that into the ad spend.