And what are the key kind of steps you’re going to take to get that done? Thank you.
Julia Donnelly: Okay. Maybe I’ll take the first part of your question, Ken. So it is not mix. It is more acceleration and ongoing acceleration from retail and an emerging contribution from 3P. So we do see continuing headwind from food and beverage, but it is more that it’s being even more offset by other positive growth factors that are driving our Q1 acceleration.
William Ready: And on your question on the value capture versus value create. There’s always a lag between these things. Advertisers need to see the performance sustain and you see it flow through into their models and their measurement. And so the thing we feel really great about is that, that value – the value create is the hardest part and has to happen first, and we’re seeing that as we drive more users with more intent to those advertisers. So that’s really fantastic raw material for us. And we’ve only really just started the value capture on that. And what comes ahead in that is some of that is just advertisers see that bake into their own models of performance. Some of that is about getting advertisers on to privacy safe measurement.
Julia talked about, and we started talking about this Q1 – late Q1, I believe, of last year around adopters of API for conversions as a cohort growing at 30%, approximately 30%, those that had not adopted being mid single-digit decliners. That trend continues as we drive that adoption. So as I mentioned in my comments, the performance doesn’t matter to the advertisers if they can’t see it through their measurement. So driving that adoption curve on measurement matters quite a lot as well. But we’re seeing – again, we’re seeing that budget shift toward us continue. We saw that through the year this past year, doubling our growth rate from the beginning of the year to the end of the year. We see further acceleration as we look into Q1 as reflected in our guide.
And we think there’s a lot more of that to go as we continue to drive through that adoption curve. I hope that helps.
Kenneth Gawrelski: Thank you, both.
Operator: Thank you. The next question will be from the line of Colin Sebastian with Baird. Your line is now open.
Colin Sebastian: Thanks, and good afternoon. A couple, I guess, for me as well. Bill, maybe a follow-up on the MAUs. Are you saying that there is a faster conversion of maybe those occasional users maybe quarterly or semiannual users to monthly users? Or are you seeing just users who are new to the platform overall. And then secondly, in highlighting the improvements you’ve made in relevancy and recommendations of ads and cons in the platform. I know there’s a lot of work with machine learning behind all of that and using signals from users. So I guess I’m just curious, how far along do you think you are in really optimizing that level of personalization and hitting those key advertiser objectives like conversion rates? Thank you.
William Ready: Yes. Thanks for the question, Colin. So on your first question around MAU, we are seeing progress both in bringing new users to the platform. We talked about Gen Z, where we are winning with Gen Z and they are finding Pinterest quite compelling and unique and different from the rest of social media. So we are bringing new users on the platform like our Gen Z examples. But it’s also the case that we are deeply engaged with existing users, giving them more reasons to come back to Pinterest more frequently. And I’d say there’s two large drivers in that. One is as we make better and better recommendations. That gives them more reasons to come back and giving them more suggested use cases, helping them to broaden out what they think they can do on Pinterest showing them adjacent use cases.
So that’s working quite well. And then it’s the actionability. As I’ve said many times before, Pinterest had previously solved digital window shopping, but all the stores were closed, actionability was hard. How good was the digital window shopping, so good that people would keep coming back even though all the stores were closed. But as we open those stores, people have good reason to come back more and more frequently. So the MAU acceleration and the growth there is broad-based and it is both bringing new users on the platform, winning with Gen Z, as well as driving depth of engagement with new adjacent use cases and better actionability that caused users to come back more frequently. On your second question, on relevancy. This is where we just get really amazing first-party signal that feeds AI.
AI is only as good as a signal upon which is acting. And we get really amazing signal on what users are interested in. I’ve talked about Pinterest predicts and our ability to predict there and see trends that are coming because users plan for the future on Pinterest. We see users in their shopping journeys long before that intent has been expressed on other platforms. And so that’s really rich signal. As we’re leveraging next-gen AI with models that are 100 times larger than they were before, we’re seeing really fantastic ability to take that completely unique signal to Pinterest and convert that into great relevant recommendations for users. I shared on our last call that we saw approximately 10 percentage point improvement in relevancy when we move to our large language models.
And that flywheel between the user coming and curating, giving a signal of what they’re interested in and then us being able to give better and better recommendations to users. We see that flywheel accelerating, particularly as we bring in more actionability. So again, AI is a core competency for us. And importantly, we have a completely unique signal that you wouldn’t find elsewhere around the user curation particularly around their commercial intent. And when we see that really just continuing to drive that flywheel as we go forward. Hopefully, that helps.
Operator: Thank you. That is all the time that we have for questions today. And with that, I will turn the call back over to Bill for some final closing remarks.
William Ready: Thanks again to all of you for joining the call and for your questions. We look forward to keeping this dialogue going. As always, we hope you enjoy the rest of your day.
Operator: That concludes today’s call. Thank you all for your participation and you may now disconnect your lines.
