So those are the biggest things to solve for. As we’ve been calling out, it is a multi-quarter implementation with the most meaningful revenue to hit early in ’24. So that’s with regard to your question on Amazon. The other part of your question around potential new partners. We’ve commented on this from early on that part of what made us a multi-quarter implementation is that we were fitting the platform for third-party demand generally, not just one partner. So that sets us up well to expand to other partners as we go into ’24. And we’ve been really focused on making sure we get this first one really right, but we have built it in such a way that we think it sets us up well to expand to additional partners as we go into ’24. And then on the generative AI front, it’s — AI generally has been a really great source of strength for us.
And the places that we’ve talked about where our use of large language models, 100 times larger models, GPU serving all these great AI technologies, giving us significant increases in relevance, approximately 10 percentage point increase in relevance we talked about on last quarter’s call and at Investor Day. So that’s been going really well. And the newer products like our creative studio that is creating dynamic content for ads, for advertisers to make those ads more compelling and engaging for users. We’re early in our testing there, but we’ve seen just a palpable response from advertisers in terms of their excitement about engaging with that product and the things that we can do there. And then finally, other elements of the core experience.
There’s just really nice improvements happening throughout the core user experience that maybe more subtle as a user, but actually driving really great results like guided search experiences, getting you to more related items. These are things where we’re seeing just areas where you would normally see progress on basis point. I mentioned that approximately 10% increase in relevancy. That’s a place you normally see improvements in basis points to say, 10% — approximately 10 percentage points improvement in relevancy there. Those kinds of gains are just unheard of. And so we feel really good about the progress that we’re making with AI and large language models and the core technologies there.
Operator: Thank you. Our next question is from Rich Greenfield with LightShed Partners. Please proceed.
Richard Greenfield: Hi. Thanks, Bill and Julia for taking the question. I got a couple. One, just — I want to follow up, Bill, you were talking about sort of the, I guess, you called it the compounding benefit to both Pinterest and your content and advertising partners on ad load and how it’s not impacting engagement. What does that tell you in terms of like, does that imply that there’s still lots of room to run in terms of what the ad load could be? I mean you sort of framed it as I think you were about to say multiples and then you sort of said it was the percentage increase. But I’m just wondering, is there any way to frame how much is ad load up? And as you sort of see the engagement benefits from better targeting of ads and content do you think that there is multiples to go in terms of ad load over the course of the next several years?
And then two, just because I’ve been using Pinterest a lot over the last few months and one of the things I certainly noticed is the presence of Timo (ph), wondering if there’s any way you can sort of quantify, has Timo been a major advertiser year-over-year? How much of an impact has it had positively on your business year-over-year, would be helpful just to put it into context.
William Ready: Great. Thanks for the question, Rich. So stated very simply, there’s a lot of room to run on ad load, again, both through the unlock of whole paid optimization and our ability to look at the ads of relevant content as well as being able to ingest more relevant third-party demand as well as getting more first-party demand as our advertisers, particularly retailers are seeing more performance from us and wanting to bring more relevant inventory onto the platform. We think there’s a lot of room to run there. Over a multiyear period, the analogs that I would have in my mind would say to me that when the users in a commercial context on our platform, which again more than half users are here to shop, but when they use it in a commercial context I think the analogs I would look to would say that we could support a multiple of the ad load that we currently have.
That will be a multiyear journey to get to that. But I think that means there’s just a lot of runway in front of us on where ad load can go. And the key proof points that I would have had in my mind coming in five quarters ago, were really around whole page optimization, proving out that the users wanted to shop and would engage all the way through the lower funnel and the ability to bring in more demand rapidly and all those major unlocks, we have progressed quite nicely on all those, but there’s a lot more runway to go to get to get all the goodness out of that. That will be a multiyear journey, which is why we feel confident in our prospects as we look forward. And then on the other part of your question on Timo (ph), our strength in retail has been broad-based.
So if you looked at our strength in retail, the largest advertisers, the larger, more sophisticated advertisers they’ve been seeing really great results from our latest product developments and they have been leading into the platform, each of our largest, most sophisticated advertisers across retail have been growing faster than the growth rate of our revenue overall. So they’re leaning in and contributing to growth. And then Timo has been a nice — I will not comment on any one specific, but just sort of Asian e-commerce exporters generally have been a nice contributor to that. But we don’t feel particularly concentrated or exposed given that our strength in retail has been very broad-based.
Neil Doshi: Operator next question.
Operator: Thank you. Next question is from Brian Nowak with Morgan Stanley. Please proceed.
Brian Nowak: Great. Thanks for taking my question. I have two, excuse me. The first one, at the Analyst Day, you gave that great chart where you sort of showed the growth in engagement versus that basket of saves and impressions versus the DAU versus the mile growth. Can you just update us for how fast that engagement grew in the quarter? I think it was up 14% in 2Q just for context. How quickly was it up in 3Q? And then the second one, you have a lot of different monetization sort of irons in the fire. Can you sort of walk through some of the key technological hurdles you still have to overcome for direct links and deep links as we think about those fully rolling out throughout 2024?
