Etsy, Inc. (ETSY) Earnings Call Transcript & Summary

Well, hi, everyone, and welcome to the Loop Capital equity conference. We're so glad you can join us on our first big firm-wide conference even on this virtual platform. And we're thrilled to have Etsy's CFO, Rachel Glaser, join us for the next hour to chat about Etsy. If you have questions that I am not asking, I believe there is a chat function where you can toss in a question, and I will see it. Otherwise, this will be a 2-way dialogue. Rachel, thank you so much for joining us this morning.

Thank you so much for having me, Laura. It's great to be here.

Let's start on the search function, which is a place where, when I first picked up Etsy, you had plenty of room for improvement. But it seems like there's always room for improvement given the way you don't really sell from a catalog. Please update us on the progress you've made in search. What are your latest wins? And how would you like us to measure and gauge your success there?

Great question. So for starters, Etsy now has 80 million different items on the site. So as sellers have grown so has the large breadth and depth of product and content that we offer, and so search just becomes even more important as a product development area where we can help people find what they're looking for and help them convert at a higher rate and come back more often. We have made great strides. One of the metrics that we showed in our earnings call was the metric that shows the percentage of time that somebody finds something on their -- on the first page of search results, and that metric has gone up and up and up. And then sort of the inverse of that is the metric that shows the number of times somebody has to reformulate their search query to try again and again, to find the item that they're actually looking and that reformulization rate has actually decreased quite rapidly. And those things are the results of improved search efforts. And so one of the things that has been a really big focus is personalization, so really knowing, Laura, if you're looking for table lamp and I'm looking for table lamp, that you and I might see completely different search results. And we're super excited that that's now live and part of the experience. Because we know things about what you last looked at and what your prior purchases are, so we know a bit about your -- the average order value a little bit, inferring what your demographic might be, something about your taste. You might be like mid-century modern and my taste might be more Boho, so we're going to try to serve you up a result that is much more likely to convert you. Another thing that we've done is we have more of an inference in the search. So we've used these examples before. But if you were to type in gown, previously, you might have seen bathrobes, you might have seen wedding gowns, you might have seen wedding gown hangers. And now we're going to infer from the other things that you're doing on the site a bit more about what you meant by gown. And so those things are really starting to have very real effects in the searchability of Etsy site. And I guess the last thing I'll mention is we're doing a lot more with curation. So you can go in and see -- not have to look through the 80 million items but to start to be served up things that are recommended for you or might be curated for you based on -- an editor's picks page, for instance, might contain things for you that are different than another person. And soon, you'll also see us extend that to -- we have a lot of influencers now on Etsy that you can follow specific influencers so that you're looking at a more narrowed-down set of results rather than the entire corpus of 80 million items. So we're really excited. There's so much more to do there with machine learning and even AR so that we can show you items within a setting or within wall art on your wall that will start to make the -- not only the search but the decision-making process a lot more engaging for a consumer.

Yes. In the Champine household that recommended for you section has been a hot part of our process to turn us into power users, and it seems to be getting better and better as we shop more.

Agreed.

What's next on your road map to make search easier for Etsy's customer?

So when I gave you -- I talked about curation, and collections would be another area where we're investing time. One of the things that we've done is we've introduced video into our capabilities. And so we've given sellers the ability to upload video, which is sort of an augmentation to what you see when you search. So like we have 3 million videos now. And that gives a -- first of all, it sort of enhances the effect of human connection, which is a differentiator for Etsy, that you're able to see the seller's creation process. You're able to get much more of a dimensional view of a product, the look of it, the feel of it, the -- its origins, how did this wool actually get created. And so the video -- it's 3 million videos, but that's just a sort of a drop in the bucket for what it can be. And so the ability to start to incorporate that into search results is one of the things that we're focusing on now.

Got it. I wanted to talk a little bit about browse time because it's kind of a funky metric because we want it to be low to the extent that people can find what they're searching for but we also want people to explore the platform and get to know all the different products that you offer. What's happening with browse time during this -- while we're -- a lot of us are still kind of stuck in the house more than usual? And how did you measure -- where do you want browse time? And how is that metric trending?

So we haven't used -- some people call it browse time, and we -- internally, we call it dwell time. And we haven't used that metric as a sort of external reporting metric. I think it's a great question and something that we can talk about. But you're right that right now, we're more focused on helping people get to their ultimate purchase decision faster, so that metric I talked about in the beginning of finding your -- finding something on the first page of search results rather than having to reformulate, reformulate or dig down into page #32 until you actually find what you want. And so we're measuring success right now by faster time from entry to purchase. The -- but it is interesting because we've always also talked about our buyers as some people love the adventure of, "I just entered a treasure trove of, fill in the blank, a flea market, an antique store a shopping mall," and they just love to hang out and keep looking and keep looking. And other people are more pragmatic, and they find that adventure to be annoying or a challenge. And we have a blend of those experiences on our site. And if I gave you a number, I'd be averaging. You would -- it would end up not actually telling you that much. So I'm going to sort of punt on the -- what is the dwell time and what's been happening with it to say that we focus on the 3 metrics that sit behind GMS, so that's conversion rate, visits and average order value. And obviously, GMS has grown up -- has grown quite significantly, and the 3 metrics behind it are driving that. A lot of it coming from conversion rate, meaning people are finding things effectively with our improvements in search.

What's happening with AOV? I think your visits are probably up huge, too. But what's going on, on the AOV side?

Yes. We've made -- and we've talked about this a fair, but we've made -- and we call it average converting visit value internally both -- while companies call it average order value. We've made smaller increases in AOV, but there have been some last quarter. I think it was the fourth quarter that we introduced installment payments using Klarna as a partner, which effectively does help move the needle on AOV because people can find things more -- they can find higher price point items and pay them in a way that they can afford to. So that's one of the wins that has been an AOV driver. Reverb, which is now blended in with our results, has a much higher AOV than Etsy. Now they're much, much smaller than we are so it doesn't -- effectively that doesn't move the needle that much on AOV. And then masks have driven AOV down. So because we had $740 million of masks sales at some average price point of $12 or something, that drives things down the other way. I've always thought like in my career, in other lives that AOV is a less important -- many people might not agree with me, but it's an output metric. So you could have a very low AOV but huge volume, and it shouldn't really matter as much what the AOV is as long as you're having GMS growth. And I think the things that we found we've had much more control on through our product and marketing efforts have been visits and conversion rate.

Got you. I wanted to talk a little bit about shipping speed, which I know when I chat with Josh, that's not what Etsy is about and I get it. Etsy is about special and custom and personalization. But if people didn't care about shipping speed, Amazon would be a much smaller business right now. So I just wanted to talk about -- because that is a pain point for some customers. What is within your power to get shipping speed accelerated?

So we've focused on expected delivery date. So you're right, Amazon stands for value and convenience. If you want something at the lowest possible cost and you want it at the soonest possible moment, even an hour or 2 later, you're probably not going to think of Etsy first for those things. And so we've -- actually, I'm going to zoom out a little bit. We've -- internally, we have -- a lot of our product squad's focus is on the entire post-purchase experience; so once you purchase, what is that experience like. And 1 year, maybe 1.5 years ago, we really started focusing on shipping cost. So what we really were trying to do was drive down the percentage of items that had an egregious shipping costs, so something -- shipping costs very high relative to the item price; and to make a significant portion of all the listings be free or free-eligible. So we did a lot of work to say if items are -- if your total order value from a single shop is $35 or more, we had a lot of work to have that be free shipping and really educate our sellers to think about shipping as a cost of goods sold. So it's just -- you don't need to break it out separately. Think about your pricing to be inclusive of shipping so that on the presentment of your item, you can say shipping is free. And we got pretty far with that. So over 77% our listing views are free-shipping eligible. And we said that's good. That's good enough really. There's reasons why a large percentage of our listings might be under $35, so they're not free shipping-eligible or they might be cross-border where it's not possible for the seller to make the item free. And so we sort of did pause our efforts on shipping, but we've turned our focus next to what's the experience like for a buyer. If I'm buying something in time to get you, Laura, a gift by your birthday, how do I know it's going to arrive there on time? Like it's almost like you don't want it at all if it's not going to be able to arrive by the time that you want it. And so we've done a lot to be able to improve the accuracy of an expected delivery date and to increase the number of listings that show an expected delivery date, which not all of them did because not all of our sellers were using tracking so we couldn't estimate the expected delivery date; and then to change some of our search functionality so that you're able to search on only show me things that are going to arrive by X date or are ready to ship within 1 day or are ready to ship within 3 days. So you can filter down from the 80 million items just to the ones that are in the relevant consideration set for me to buy a gift for you. And so we've made a lot of progress there. I'd say it's much better. I don't think -- we're not done yet, but that's been one of the main focus areas is to make it -- we believe special takes time. We're not competing on value and convenience. We're competing on uniqueness and quality of the item. And I think people -- most people know with Etsy, you need to plan a little bit ahead. But when you do decide to buy from Etsy, we want it to be a great experience, and so that's where our efforts have been.

Those tools that offer your buyers visibility to delivery times, what percentage of sellers are using those? And is that something you would ever mandate?

We can't mandate it in countries where we don't have an integrated shipping partner. So in the U.S., we have partnerships with USPS. In Canada, we have Canada Post. We have Royal Mail in the U.K. And in Australia, we have Australia Post. And they -- so at this point in time, we wouldn't want to mandate something that's not the world-class solution for sellers or inconveniences them more than it helps them. So I'd say that at this point, we're more in the realm of educating sellers on why it's important to be able to have a tracking on their package. And to have tracking, you need to use a postal service like FedEx or USPS that enables them to have tracking. I wouldn't say never. Like I think there might be a time where we can get all of the quality of our shipping partnerships up to a level where it becomes mandated that to ship with Etsy, you need to be -- we require that you use a tracking number, but we're not there yet.

Got it. I got a follow-up question on video from an investor that I thought was worth stating. So on video, can you comment on whether you're working with Vimeo right now as your partner? And if so, is that long-term thing or just a patch while you work on something internally? And then secondarily, if 3 million videos is just a drop in the bucket, what's the aspiration? Like what's -- what numbers -- how big do you think that video count could go on those 80 million SKUs?

Yes. We haven't commented on our partner externally. It isn't Vimeo though, so I can -- I guess I feel comfortable saying that. And I think those are great questions as far as thinking about our own internal road map. When we first came to Etsy, Josh and I and other members of the management team, the company previously had been very much "If we can build it, we will." And we shifted the thinking to just because you can build it, it doesn't mean you actually should. And so an example of that is going to Google Cloud rather than hosting ourselves and all of our data centers. So at the moment, we're using a third-party provider to help us with the video capabilities. I could see some analysis that might say it's better for us to develop our own tool and host it ourselves and provide much more proprietary or unique capabilities than the third-party provider can satisfy. But right now, it would have to pass that threshold of "Just because we can build it, maybe we shouldn't." We're really loving our business model where we're not capital-intensive, and we're not -- we rely on the capabilities of third parties that that's their core competency. Another example would be using Zendesk for all of our CRM, customer management front-end servicing. So again, I don't want to -- never say never, but at the moment, we're using a third party. And I'm not -- I'm uncertain how far that can take us and what the limitations might be, but we're looking at it carefully. I say 3 million, a drop in the bucket. We have 4 million sellers. So first of all, that's less than 1 video per seller. It's what you -- ideally, you'd want to have. And we have 80 million items so we don't have a video per item. So I can see it getting to be a much more robust experience where sellers are showing more of their listings with an attached video or at least a video per shop or something like that, that shows the making of. So you can see that there's a long way to go to make it a much more predominant feature on Etsy's experience.

Got it. I wanted to dig into TAM a little bit, just given the massive growth that you've had. I think Josh has commented that you're selling some bigger-ticket furniture items, which you've always had but I'm not sure the velocity has been great. As you look at your TAM today, how large do you think it is realistically? And what's your penetration at this point?

So we've had a little journey with our TAM. We said at our Investor Day in March of 2019, we were trying to make the TAM small enough that it was absolutely believable by all of our investor communities. Because people tend to think of Etsy as a niche site, so when you throw out a big number, they say -- we get a lot of but, but, but, people are -- most people aren't shopping for handcrafted items and whatever. So we took a lot of slices off of a very big TAM. And we said, look, we're -- if we only take the top 6 countries that we're in, we only take the top 6 categories that we're in and we only take the online portion of those people, and then you slice this back further by only looking at consumers that are interested in self-expression and uniqueness, which is by -- some of our market research says about 40% of our buyers are -- 40% of buyers out there care about personal expression, and we came to a number that was a couple of hundred billion dollars in size. And we said we were less than 5% of that at the time. Well, one of the experiences we've had with pandemic is that people ubiquitously shop -- jump easily between offline and online. So why are we artificially taking that slice. People -- and we believe that's permanent, and it probably has always been that way. When I shop for a sofa, I'm looking online at westelm.com but I'm also going to Crate and Barrel, the store, to sit on the sofa. And wherever I find the best sofa is where I'm going to buy it. So that online/offline thing is sort of an artificial or made-up delineator. Same thing is we're in a lot more than 6 categories and we're in a lot more than 6 countries. And so when you take up all those limiters, what we said more recently is we think the TAM is -- begins with T, not a B, and we actually think it's in the trillions of dollars. And you can do it another way, which we actually showed at the same Investor Day presentation. If you just take the TAMs for each of the big categories that we're in, so home furnishings alone, if you use the TAM that Wayfair has put out there, I think it's at $400 billion, and you can do that with fashion and clothing, also hundreds of millions, and add them all up and you get into the trillion. So a lot of different ways you can say we're in the trillions. We just did $10 billion of GMS. So we are something sub 1%. So there's a lot more room for us to grow and take our fair share of that very large e-commerce marketplace. We also naturally think we have -- people ask all the time, "Who are your competitors?" It's hard to think of any one competitor that is -- does exactly what Etsy does. Maybe eBay comes close, but they're really not -- they're in such a broad range of categories, everything from automotive all the way to stamp collecting, and don't focus on sort of the uniqueness and the specialness of an individual seller as a creator. So that -- I guess it's hard to think of anybody that is uniquely like Etsy. And so I think we've got a big differentiator in all of that sort of high -- the TAM that has a T attached to it.

Got it. Let's skip ahead to take rate because I'm getting investor questions on that already. Would you still agree that your mandated take rate remains below other marketplaces? And do you feel that there is room for further fee increases as a result?

So we gave our take rate at about 17%. We said high watermark would be probably Q1 of 2021, but then that would be approximately the take rate you can expect for the full year as well. And that's up probably about 100 basis points from a year ago. We got some take rate improvement not by doing a pricing change but by launching Offsite Ads. So there's multiple ways to grow the take rate. One is by adding new services for which we can charge a fee. Another is by changing that mandatory transaction fee, which we did 2 years ago. Our philosophy the entire time though has been what is a fair exchange of value. So even when we did just the wholesale price change from 3.5% to 5%, we did it with the philosophy that we're going to create ourselves more margin so that we can invest more in you, and we feel that it did create significant incremental revenue when we did that take rate increase. We reinvested 80% of that incremental revenue back into marketing in a way that sellers can really see and experience and feel the results of. So we did it in a very virtuous flywheel where we could reinvest in marketing, which drove more GMS to them, and it was a win-win for both Etsy and for our sellers. And so we very much live and believe that fair exchange of value. So I wouldn't necessarily say the best way to do it is where we are in the ranking of highest take rate by one -- competitor A and lowest take right by competitor B. It's what can we provide our sellers and what's the right fee to charge for that in a way that they're going to say, "Yes, that's a fair exchange." And I think we have line of sight to further investment in services that we can offer our sellers for which we may be able to grow the take rate a bit. Nothing planned. We haven't guided towards that in 2021. So nothing that I can speak to specifically, but that's the philosophy.

So to follow up on that, I would think that given the growth in your platform, you are offering more value to sellers today than you did a year ago. And I know that investors have been asking you, since you made the last wholesale price increase, when is the next one. And I also know that Josh said that he's been at a firm that did it every year and that, that was wrong. But what -- how should we think of the timing of a potential wholesale take rate increase, mandated take...

We've done something that's been effectively a take rate increase to our sellers every year for the past 2, maybe 3. So giving them a little bit of time to grow into the rates we're charging now makes sense to us. We haven't given or changed our long-term guidance at all since we gave it in March of 2019. So I don't want to do that here. But I do think that -- I mentioned Offsite Ads but another thing that we did in 2020 was we expanded Etsy Payments to 9 additional countries. So there's things that we're doing that wouldn't be a headline but that are gradually helping to increase our take rate. We also have continued to grow Etsy Ads as the on-site ad product and see a lot of ways to -- and that doesn't have to take -- that doesn't change pricing for anybody, but getting more sellers in and adopting that product and benefiting from that product also helps to grow take rates. So I think there's a lot of things that we can do more of a gradual turn of the dial than a one shot sort of change to pricing.

Let's dig into those other things a little bit. Because -- I hear your comments. You've been very clear that you believe the take rate is kind of peak in Q1, but I don't really get why that would be because I see the opportunity for Etsy Ads. It used to be 50% utilization of seller budgets. I don't know how much that has grown. But also, on Offsite Ads, why would that not grow as a percentage of GMS and sort of organically improve your take...

Yes. It will grow, we believe, as a dollar percentage of GMS. So we've talked a lot about what's happening with GMS in 2020, lapping some peaks in 2020 that are going to go away. And then what we did say was that we introduced Etsy Ads in Q2 of 2020. And so right now, what you're seeing is take rate relative to a lower take rate from last year. We also launched Etsy Payments in 9 new countries later on in the year. So we're going to lap that as well. And then Etsy Ads as well has had significant growth, and we're going to be lapping that. So it's more of a period-over-period constraint. And as GMS -- we didn't give GMS guidance for this -- for the full year. But as we normalize and get into 2022, you might start to see some of the benefit of that -- not having to lap those peaks show up in the numbers.

Okay. So on Offsite Ads, as it represents a materially higher percentage of GMS over time, where should it go from the 9% of GMS today? I mean it's lovely that you have organic traffic, but this is potentially more profitable for you. Where would you like to manage it to? And how much of a boost can it make over time to your overall take rate?

Well -- so I'll answer, it's a great question, in 2 ways. One is we've already -- since we gave that 9% number, we've already launched a new channel for Offsite Ads. So previously, we had been PLAs for Facebook and Google primarily with a little bit of Bing and a little bit of Pinterest in there. But we've now launched our affiliate channel, which gives sellers much broader distribution and opportunity to showcase their listings. And potentially, some of the Google Display Network enters into the opportunity set for Offsite Ads as well. So that's one way to grow it. But you rightly point out we do care about our organic traffic, too. So when I tell you, as a percentage of our total spend, the percentage itself varies. So we say that about 20% of our GMS comes from paid marketing, but the way to make that number higher or lower is more of your GMS coming from organic means. So I -- the denominator might change as we have really successful efforts with our CRM tools, with our SEO efforts, with our above-the-line brand marketing on television and digital video that can influence what that percentage is. So we haven't put out a specific number for where the chargeability rate should go. I'd also point out that we haven't -- right now, we're strictly PLAs. So we also spend performance marketing on SEM. And that number -- we will keep spending there as long as it's ROI positive. That today is not a channel for Offsite Ads, could be in the future. So those are the variables that could influence that number and make it -- continue to make it expand the sort of cooperative marketing efforts that we have with our sellers but without giving you a specific chargeability target.

Got it. On your overall marketing budget, what's your outlook for the marketing budget growth and -- just given what we know about price increases on digital advertising and its effectiveness today?

I love how we manage marketing at Etsy where we really manage it with an ROI target rather than an absolute dollar budget. An exception to that might be the spend on our funnel marketing, which, particularly on the brand spend, part of the value we're getting is not quantifiably measured because we're looking at the impact through market, via our brand research, how much did we move the needle on top-of-mind awareness. But on the -- all of the performance marketing, it's done with a focus on the ROI threshold. That's also dynamic. So as lifetime value increases -- because of take rate changes or product changes that we make, lifetime value can go up and we can spend deeper. So -- but even -- so it happens organically. If CPMs and CPCs go up, it natural -- there's a natural pullback because we have trained our algorithms to not cross over that ROI threshold, and we train them on every marginal dollar of spend. So it's not just for the month it has to be ROI positive. It's on every dollar of spend, it has to be ROI positive and there's a diminishing -- point of diminishing returns on that spend. So I think we've done a really good job with that even though there's this dynamic -- LTV is looked at daily and fed back into that algorithm daily so that it's constantly reeducating. One of the things that we've started to evolve more, and I would still call it sort of not fully mature process, is to make those algorithms like a dynamic pricing algorithm where it knows a bit more about the buyer that we're acquiring and how much we value that buyer. If it's a repeat buyer, maybe even habitual buyer, maybe we don't want -- maybe we don't value them as highly because they're going to come to us anyway and we're spending money where we don't have to. And then maybe it's a buyer that's shopping for furniture and we have more margin to spend versus a buyer that's shopping for slime, where there's very little margin to spend. And so we're starting to make it even smarter in terms of how we target and what it is we're acquiring and how much we're willing to spend there.

What's your payback period when you look at spend based on ROI?

It's -- we -- the payback period is very fast. It's usually within the first 90 days or even within the first 30 days because most of the value in a lifetime value curve comes very, very early in the life. Again, those things are also dynamic because we've seen, as we've increased frequency, that we have more tail on the lifetime -- the user lifetime value curve. We would -- we're not throttling it. We're not requiring payback within 90 days. It just happens to pay back within 90 days. And I think because we haven't given you LTV-to-CAC ratios really clearly, people are trying to use the blunt instrument of total spend divided by total new buyers. And it's -- you lose something in how much spend is to reactivate a buyer, how much spend is to move the needle up on frequency and how much of that value is coming in the tail of the lifetime value curve.

We've talked in the past about Etsy historically did not do a lot of personalized marketing. If you got an e-mail from Etsy, everybody else got the same e-mail. How far have you come on that? How much personalization are you using? And what do you know about how effective that is?

Yes. So we've totally changed our marketing technology stack, and we used 2 vendors called Braze and Segment to really have a lot better information about the database of customers that we have and what message goes to that part of that helps us with that dynamic bidding that I was talking about before. But it also helps us with customization of messaging through every touch point that we have with the consumer. So we talked earlier on in our history where Josh used the example of he bought a watch stand. I don't even -- didn't even know what a watch stand was until I heard the story from Josh. But I guess if your watch -- if you wear a lot of watches and you're into watches, you want a place to display them and have them out on your bureau. And he started getting tons of emails, "Hey, we see you like watch stands. Would you like to buy another one?" Well, actually, once you have one, you don't need any more. And so that isn't happening anymore. We're now saying, "Hey, here's a man that likes watches. Maybe he likes other leather haberdashery items. And let's send him things that are more to his taste." So the e-mails are getting a lot more customized than they ever were before. And the other thing to work on is how often you're getting an e-mail and what's -- what are we prompting you with. Is it just to buy more leather haberdashery items? Or should we say that vendor you like is now having a sale? So we're able to send promotions. We're able to remind you about birthdays. We're able to give you more customized messages and measure the frequency, what's the right number of touch point to have with the consumer before you've been tuned out by them. So we're getting better and better at it. But I would call that still lots of opportunity ahead of us. I wouldn't call us world-class in that area yet, but we've laid down the right tools to become world-class.

Got it. So bearing in mind your ROI basis for spending based on what you're learning recently from advertising channels, where do you expect to lean in and lean out?

So one thing we're doing is spending more geographically. So we have seen big opportunity in both the U.K. and Germany, but there is the whole rest of the world. And we did spend more in the fourth quarter in a way that we believe was ROI positive, not only in television, which we did in the U.K., but in the mid-funnel with more paid social in ways that are customized to that market. So the products that are listed, the words that are used are much more customized, and we were able to get a better yield on the spend there. So we're now developing creative that is specific for Germany, not just language but also the content of the message, and we'll roll that out in Q2 or Q3 of -- lost track of which quarter, but we're investing internationally. We'll also develop new creative. We ran 2 campaigns on television in the fourth quarter in the U.S. One was focused on what we call DR, direct response, much more -- even though it wasn't like call this 800-number right now, but it was much more focused on go shopping on etsy.com with the etsy.com prominently listed in the 30- and 15-second ads and really showing you examples of products. Like these are some of the things you can go buy on Etsy. And then we also ran a brand-focused campaign that was much more emotional, like why would you ever come to Etsy, when should you think of Etsy what is Etsy. And those things ran in parallel and worked really, really well. They were holiday-themed. So now we're going to do creative that is more evergreen and do some more of that sort of brand in parallel with DR and keep putting the dollars to work in the upper funnel. And then on the mid- and lower funnel, as we've talked about, we'll keep spending to our ROI threshold breakpoint. And because LTV is going up through product enhancements, we can keep spending deeper. But as you pointed out, the headwind the other way might be higher CPCs and CPMs as the world reopens and some of those advertisers that have been sitting on the sidelines for the past year come back into the competitive bidding process. So it will be balanced with that constraint as well.

Got it. Understood. I'll never forget how to pronounce Shiori, so I saw plenty of those ads. On the international expansion, how are the economics for you in those different geographies, sort of non-English-speaking geographies? Or even if you want to talk about U.K., Australia. And do you see yourself spending a lot more on advertising as a percentage of revenues in those markets as they scale up?

Yes. So we've been -- this was the first quarter we even broke out the fact that U.K. became 10% of our revenue. So we're required to say that and break it out. But we haven't given sort of P&L by market. So I can say that our take rates do vary by market because you have different adoption of seller services in different markets, particularly things like shipping that have -- we don't have a robust product in Germany for shipping, for instance, and there might be more domestic shipping that's going on. So the labels that get printed are -- the margin on that varies in those markets, and we make less money on some of the seller services there. Adoption of Etsy Ads in those markets has also been varied and not as robust as the U.S. So that is some consideration in the algorithm of how much we can spend and what's the ROI to spend in those markets. At the same time, they're growing at very high rates. And so we are seeing -- we've bumped ourselves up to being a top 5 site in the U.K., and we really have moved the needle on top-of-mind awareness, which means that's an indicator of conversion rate is up, which allows us to spend more in a profitable way. So there is a virtuous flywheel that goes on there with marketing spend and product improvements that we've been able to make to conversion and frequency working the way we want it and enable us to spend more.

I wanted to ask one -- a question we got from an investor directly on Pinterest because it's a really interesting cross-section with what would make sense for Etsy, but I'm not sure how actionable it is. So here's the question. If you look at Pinterest and platforms like that, do you consider those middle of the funnel? And can you help us kind of quantify Pinterest as a percentage of your marketing mix? Or do you think that that's a channel where you might spend more there in the future? Or how good is the ROI on Pinterest?

So we haven't broken out -- all of the sort of social networks and search engines have different ad products, and some of them are more effective than others. So that's the first point I'll make. And then there's some -- so for instance, Facebook has an enormous audience, but they have a lot of products to be able to give us more precise targeting to the audience we are trying to reach. And so we -- even though it's an enormous audience and people might be going to Facebook to find their friends and aren't necessarily looking to be shopping for unique handcrafted items, we're able to find both Etsy customers within Facebook or Etsy look-alike customers that we're prospecting for in an effective way. So it's a lot to do with what products those companies offer us. Today, the products that Pinterest have offered us do not produce as high of an ROI as some of the others. And like I've said, I think probably at the risk of repeating myself like a fourth time, we will keep spending as long as it's ROI positive. So we're kind of agnostic to where we're spending as long as we can get the yield. And so right now, Pinterest is sort of -- we introduced PLAs, so product listing ads, included Pinterest in that sense. But it's a minority of our spend because we max out where we can get the ROI on that particular channel.

On the recent earnings call, Josh called out frequency as a key metric. I mean that's something I've been obsessed with for years. But it seems to be working out. But how do you drive that higher? It's unclear to me. Is that something you drive? Or do you drive it sort of sideways through like making search more effective and so forth?

Well, I think, yes, frequency is the right place to focus. And right now -- we talked about our habitual buyers. Those that come to us 6 or more times a year and spend $200 or more is the fastest growing segment. So it grew over 157% in the quarter, up from 100% in Q3 and mid-60% in Q2. And the next fastest segment are repeat buyers, those coming 2 times a year. And the conversion of a repeat buyer to a habitual buyer is accelerating. So 11% of repeat buyers in Q3 became habitual buyers in Q4, up from 7% conversion in the prior quarter. So they're moving fast. And the other -- just one more data point in there is that for a long, long time, we've been saying about 60% of Etsy's buyers come one time a year and 40% come 2 or more times a year. And in this last quarter, we reported that, that 40% jumped to 48%. So it's almost now 50-50 that half our buyers come 2 or more times a year, and the average of that 48% is 5x. So we're making real progress on frequency. And we attribute that both to marketing efforts and product efforts, so they work really nicely in tandem with each other. The marketing, of course, is to remind you why Etsy and when Etsy and to help lock it in to your top-of-mind awareness that you think of Etsy first as a place to go to find items that you're looking for not just once a year for the Mother's Day gift but anytime, whether you need it for something for yourself or you need it for somebody that you care about for a gifting occasion. But when you come to Etsy, you want to be able to find what you're looking for amongst 80 million items, and that's where product comes in. So we've talked about search. We've talked about the post-purchase experience. But there's a lot of things that we're doing specifically for trying to inflect that frequency, some of the marketing and some of the product initiatives. I'll insert for a moment the fact that it's one of the top priority and focus areas for Etsy in that we've given significant amount of resource to initiatives that are intended to have an impact on the frequency metric. And we've put -- Ryan Scott, who's our Chief Marketing Officer, he's been sort of tagged to be the executive sponsor over all frequency initiatives, even those that are products. So we have what we call a cross-functional initiative that is looking at all of the things that can inflect frequency and put one executive sponsor to manage all of those work streams. So some of the things that improve frequency are discounted offers that we want present to certain categories of buyers to get them to come back in. Another example might be how we handle -- from our member support organization, how we handle a return or a claim on -- I didn't -- the item was not as I thought it was going to be, how we proactively refund or gives them a concession of some sort. Search initiatives are in there, where -- what you last searched for or what we're recommending for you so that people can pick back up their shopping journey without having to start all over again. So there's a lot of things. And I think when you think about -- that a habitual buyer is somebody who comes only 6 times -- buy 6 things a year or spends $200, that's not a very high benchmark for when you think about all the shopping occasions that you have in your life. It seems completely reasonable to say that should be at least once a month. So just doubling that seems not an unreasonable sort of aspiration to have, which would be an enormous improvement to GMS. So we're just very early days in the frequency endeavors.

To pursue that a little further, I think one of the drivers will be if you can get consumers to buy across categories, buy across sellers. Do you measure the number of sellers that a buyer will -- and do you track the number of sellers that a buyer will purchase from or the number of categories a buyer will purchase from? And how is -- that consumer breadth of interest, how is that accelerating?

Yes. I mean, I think that's a great place to focus on. We definitely look at -- rather than looking at seller experiences per buyer, we look at categories per buyer and the propensity to buy from a second category. One of the metrics I love, just an easy one to remember, is that 50% of the people that came in the third quarter only for a mask came back for a second purchase in the fourth quarter that was not a mask. So they -- that entry point of mask they immediately said, "Oh, I remember Etsy" or "I didn't know Etsy sells sofas," and came back for a non-mask purchase. And so we've been looking a lot at the number of buyers that come across category and seeing some acceleration in not only the number of categories they come to but within a certain -- we look at how fast they do that to say is there like an inflection point, that if they come back for a second category within the first 90 days, what's the lifetime value of that buyer versus the one that doesn't come back for a second category until some time later. We can report more on that as we make progress. But that cross-category dynamic is an important one. And I think more and more, we're seeing traction there.

Got it. To switch gears a little bit, I think you're making progress with Reverb. I think that's obvious. I don't think that's all COVID-driven. But what does your experience with Reverb mean for potential M&A going forward for Etsy?

So let me talk about the balance sheet for a minute. We had $1.7 billion of cash on our balance sheet when we closed Q4. We actually converted more EBITDA into free cash flow than -- the conversion was over 100%, having to do with some of the working capital dynamics on our balance sheet. And we also announced another $250 million share repurchase authorization from the Board. And previously, we've laid out 3 uses of capital. We're not a capital-intensive company. We don't have warehouses. We don't have physical retail. And so we have very light capital needs. So we anticipate that we have high cash flow generation. We love our business model for that reason, high cash flow generation from the operating income that we make. Where we use our capital though are organic investments, and we've signaled that we're going to keep spending on marketing. That's a place we're leaning in, and we're going to keep spending on product development because we think there's so much opportunity ahead of us. And we actually thought margins were higher than we wanted them to be. We love the proof point that the model works. But we see so much opportunity that we want to accelerate our hiring ramp, so we're organically investing in people and marketing. We've done that in the past with other organic investments, like invested in Google Cloud. And thank goodness we did invest in Google Cloud because when we had a huge surge in demand in 2020, we would not have been able to serve that flawlessly to customers without having laid down that infrastructure. So bucket 1 is organic investment. Bucket 2 is M&A. So we acquired Reverb, and that was what I would call an investment in a new category. So we now have the Etsy of musical instruments in our portfolio. And you can think of other categories that we are really not in today or that we're in lately compared to competitors. So jewelry or fashion might be an example. We're not really in collectibles, stamps and coins. So there's -- you can think of category add-ons or adjacent marketplaces that make sense. The year before we did Reverb, we did a geographic transaction that we bought -- or we made a referral agreement with a large marketplace in Germany called DaWanda. And so we expanded geographically through that transaction. And then you can also think of IP or talent acquisitions like we did with Blackbird, which was a year or 2 before DaWanda, where we expanded our machine learning capabilities significantly through that to that agreement. So that's sort of the M&A bucket, and we're constantly looking and evaluating proactively and reactively to opportunities that are out there. And then the third bucket is return of capital scenarios. We've done sort of continuous share repurchase authorized at certain levels by our Board. That isn't -- we're not day trading. What we're doing is we're offsetting dilution that's created by giving equity as a form of compensation to our employees. That's dilutive to shareholders. It's not free. So we repurchase to make shareholders whole. It's a good return of capital housekeeping, we consider it. It did happen to be a great investment to the average price on the left of the total. Prior authorization was $65 per share. So that was a nice result as well. So those are the outline of how we use our -- how we think about using our capital.

Got it. I know we've just got a couple of minutes left. One of the things I appreciate about your senior management team is that you are reflective. So I remember during the -- that on Etsy Ads, you tried something that didn't work. You reframed it and moved differently sort of right away. But as we look back on the free shipping initiative, how important do you think that's been? How do you reflect on the success of that initiative? And also on Etsy Ads, at the time, you were at 50% utilization of budgets -- of seller budgets. Where are you now? What's your progress? And how do you look at the payback of those 2 initiatives?

Yes. So I think first of all, I'd love that you pointed out sort of I love that Etsy is very agile and not afraid to call it if we've made a mistake. So we went down one path. We said this is going to take us a long, long time to scale, it's not a great result for our sellers, it's not a great result for Etsy and we pivoted. I think similar agility was demonstrated when we had a huge surge in demand and we had to quickly get in there and help sellers that had been previously selling a couple of dozen items per week and now they're selling several thousand items per week, how to help them throttle that demand and serve the customer without losing -- creating a bad experience for customers. And we were very agile right then, too. That was -- demonstrated agility not only on Etsy but on the part of our sellers. So really good things to point out about how Etsy operates. Etsy Ads, I think, has proven to be very high ROI for sellers, a good revenue source for Etsy. We continue to make that a good customer experience, too. So we're not just saying we need 3 rows of ads per page and we want X dollars of yield -- of revenue per page. We're saying we only want to show things that are relevant to what the consumer asks for. And so we've continued to invest in the search capabilities behind serving an Etsy ad, an on-site ad so that your -- you don't -- your query is for leather phone cases and you also happen to see some blue lamp shades because that happened to be an ad that someone placed. That's not a good experience for the customer or for the seller because as they become more and more relevant, the conversion rate on the ads goes up, which means the CPCs can also go up. And so it dynamically starts to create a higher revenue opportunity. And so I think that growth -- you'll continue to see that growth. We have places where sellers really aren't using Etsy Ads at all in other parts of the world, and our -- the work is on us to demonstrate that it's a positive use of their budgets. And part of that comes from the work that we do to have a dashboard where they can actually see it clearly and understand it clearly and simply and make that an easy process for them.

Got it. Well, I think we're at time. So thank you so much, Rachel. Always a pleasure to spend an hour with you. And thanks for everyone who's dialed in, and I hope you guys continue to have a great conference.

Laura, thank you for all your great questions.

See you, Rachel.

Take care.