So that’s how I think about the way that unfolds over time. And while we’ve had good early indicators, we are at the very beginning of the potential from that. And I think there’s — if you thought about our monetization on these commercial interactions, I think we’re at a fraction of the ad load that you would see in a lot of other places that have these highly commercial intents. So there’s a lot more we can do there. We’ve set the foundation for how we can dynamically take that ad load up in a way that’s good for the user, good for the advertiser. That’s a foundation that will allow us to grow quite a bit more. And actually tying back into the questions around third-party demand, one of the things you need to do first before you bring in more demand is make sure you’ve got the supply to be able to serve that demand.
With our supply growing — engagement is growing faster than users, supply is growing faster than engagement, we now very clearly have the supply and the ability to go serve that ad content in a way that’s relevant and helpful to the user that we think lets us unlock a lot more potential in the ad platform going forward.
Todd Morgenfeld: The only other thing I would add on that, so we — I’ve had an aspiration — over the last few years, you may think back to the IPO, we talked about bringing people back to Pinterest for more things in their life, because we know that that drives stickiness with our user base. We invested a lot in personalization and relevance last year because we wanted to address deepening engagement. You’ve seen the results of that this quarter with growing MAUs, our mobile application user growth at 14% globally, up 5% year-over-year in the US and Canada. Bill referenced the weekly to monthly active user ratio at an all-time high, sessions growing faster than all of the above. So, the deepening engagement story is working because we were investing heavily in personalization and relevance.
You saw that in the financials because our gross margin and cost of revenue climbed last year. Why did it climb? It climbed because we built 100 times the size of our machine learning models last year to power that experience based on unique first-party signal. We’re now seeing the results of that in the engagement figures, and that gives us a different foundation on which to deliver new use cases to our users going forward. Operator, next question.
Operator: Thank you, Mr. Greenfield. The next question is from Colin Sebastian with Baird. Please proceed.
Colin Sebastian: Great. Thanks and good afternoon everybody. Maybe first, just as a follow-up on the comments around the episodic users. I know this is in early stages, but what’s sort of the time frame you’d expect where we could see an acceleration in MAUs above sort of the seasonal trends? I think, Todd, you talked about that you saw in Q4. And then secondly, regarding features like Watch and Pinterest TV, which you’re gaining more visibility on the app, curious how these are impacting monetization or ARPU? Bill, I think you mentioned a stat around video and the portion of monetization growth. So, I didn’t quite catch exactly what that was, though. Thanks.
