Yes, thanks for the question. I'll focus my comments, on QxH customers just because it's the most valuable customers, and there's so many different stories across all of the different businesses we have. So, I want to just focus on that. Broad stroke demographics of the new customers look, literally identical to prior years, and our new customers tend to be six to seven years, on average younger than existing customers on average kind of you would expect, that isn't changing. So I always emphasize the fact that our businesses tend to be very stable around each cohorts and kind of life stage, both for existing customers and new that's remain true in the pandemic and it's remain true, as we’ve seen lots of new customers come in through streaming services or other forms of digital platforms. The overall profile of those customers, even though they're coming in at different platforms, is very like in prior year classes. And it's one of the reasons that this gives us confidence, all of the other, more quantitative metrics, that these customers will behave in similar vein that pass your classes because on both demographic elements and on behavioral elements, they are remarkably similar. They are just more OEM. The way I would think about cost of acquisition is, when you think about the 60% of new customers that come in organically, there's effectively zero marginal cost to bring in that 60%. And so the fact that we're bringing in a lot of them as zero marginal cost is actually a positive story. That 40% where we use paid marketing to bring them in, now take paid marketing is about stable, were less efficient in Q4 just because we grew that at a pretty rapid rate, and we're still ROI positive but obviously, when you grow faster, you tend to get a little bit less efficient. But if you wait all that out, combined, you're organically and the pay, I think, it's safe to say that customer acquisition costs have come down, just based on the volume of new customers that we're bringing in.