Good day, ladies and gentlemen, and welcome to the Q4 2018 Etsy, Inc. earnings conference call. At this time, all participants are in a listen-only mode. Later, we will conduct a question and answer session, and instructions will be given at that time. If anyone should require operator assistance during the conference, please press star then zero on your telephone keypad. As a reminder, today's conference call is being recorded. I would now like to turn the call over to Deb Wasser, Vice President, Investor Relations. Ma'am, please begin.
Thank you. Good afternoon, and welcome to Etsy's fourth quarter and full year 2018 earnings conference call. Joining me today are Josh Silverman, CEO, and Rachel Glaser, our CFO. On today's call, we'll be providing an abbreviated summary of our Q4 results as we look forward to providing a deeper dive on our business at Investor Day on March 7th. Before we get started, just a reminder that our remarks today include forward-looking statements relating to our ability to bring buyers back to etsy.com and our financial guidance and key drivers thereof, including marketing spend and anticipated timing, anticipated product launches, investments in headcount, and timing of our migration to the cloud.
Forward-looking statements involve risks and uncertainties, which are described in our press release, our 10-Q filed with the SEC on November 7th, 2018, and our 2018 10-K that we expect to file with the SEC in the coming days. Any forward-looking statements that we make on this call are based on our beliefs and assumptions today, and we don't have any obligation to update them. Also during the call, we'll present both GAAP and non-GAAP financial measures. Reconciliation of non-GAAP to GAAP measures is included in today's earnings press release, which you can find on our IR website. A link to the replay of this call will also be available there, and if you prefer to access a replay via phone, you can find that information in the press release as well. We've created a short slide presentation to accompany today's opening remarks and recommend that you follow along.
With that, I'll turn the call over to Josh.
Thanks, Deb. Good afternoon, everyone. I'll start by touching on Etsy's Q4 results, which you can see on slide four. Currency-neutral GMS growth was 23%, accelerating for the fifth consecutive quarter and representing our highest Q4 GMS growth rate since we've been public. Revenue was up nearly 47%. We delivered adjusted EBITDA margins of approximately 26%. I'll talk more about key drivers of this performance, specifically our strong holiday results, in a moment. Slide five shows a recap of our 2018 performance. These results are evidence that in a sea of sameness, Etsy offers something different. GMS reached nearly $4 billion, revenue was slightly over $600 million. We significantly improved adjusted EBITDA. With a strong economy at our back, we believe Etsy's growth was higher than the overall e-commerce market, indicating that we are gaining share in a very large and expanding addressable market.
Our continued focus on improving the Etsy platform for sellers and buyers is paying off. Throughout 2018, we focused relentlessly on our four key initiatives, invested in the product experience for both sellers and buyers, improved customer support and infrastructure. We tested new marketing channels. We moved the needle on growth with context-specific search ranking, signals and nudges, personalized recommendations. A host of other product launches. In total, we drove more new buyers to the site and gave existing buyers reasons to come back more often. Digging a little deeper into the fourth quarter performance, our strong results were driven by our focus to make Etsy a great place to shop for the holiday season with gift-buying insights, a focus on personalization and customization, improved landing pages. Most importantly, unique inventory you can't find anywhere else.
We worked with sellers to create a sense of urgency for buyers, made promotions more prominent. We helped them offer more competitive shipping prices. Improvements such as estimated delivery date helped drive buyer confidence that their special items would arrive in time. In fact, turning to slide seven, on Cyber Monday, we transacted nearly $19,000 in GMS per minute, our highest-ever single-day performance. Also, approximately one-third of items on Etsy were available to ship for free domestically during the holidays, which drove meaningful improvements in conversion rate, evidence that our product initiatives, seller education, and incentives are paying off. All in all, it was a great holiday season for Etsy and our sellers, with holiday GMS, the five days from Thanksgiving through Cyber Monday, up 30% versus the prior year.
We're confident that the positive experience many buyers had will bring them back to us for that something special again in the future. As you know, one of our goals following our price increase was to test various new marketing channels to determine which of them could potentially increase awareness, frequency. We helped make Etsy more top of mind among both active and new buyers. We elevated our marketing spend in Q4 to find out two things. First, we wanted to test new channels such as TV and paid social to find if we could make them ROI positive. Second, we wanted to see how elevated spend in Google products would impact our returns. In other words, how much can we step on the gas in each channel before the tachometer reaches yellow or red?
Overall, we were pleased with the preliminary results of our testing as brand awareness and visits showed signs of improvement. In Q1, we're taking a step back to recalibrate and reassess our spend to ensure we optimize investments in the highest potential channels and strategies. Rachel will explain how you should think about our marketing spend for 2019, you can be sure we will continue to be disciplined in our approach, always guided by our return on investment. To conclude, we feel great about the progress made in 2018, and believe we are well-positioned for continued growth in 2019. I want to thank our sellers for partnering with us, our buyers for shopping with us, and our employees for all their hard work in helping to support our vibrant community and delivering a fantastic year. With that, I'll turn the call over to Rachel.
Thanks, Josh. Etsy ended the year on a strong note, as we exceeded our guidance on every metric. Turning to Slide 10, 2018 full year currency-neutral GMS, and as-reported revenue growth accelerated to 20% and 37% respectively, and adjusted EBITDA margin expanded to 23%. 2018 revenue growth was driven primarily by GMS growth, changes to our pricing model, and growth in Promoted Listings. Our improved adjusted EBITDA margin was the result of revenue growth and cost discipline driven throughout the year. These results demonstrate our ability to drive top-line growth and deliver healthy margins, the beauty of a well-executed marketplace business. In addition, we successfully improved many important metrics. Most notably, in 2018, we made progress increasing frequency, which we believe was a result of our product development and marketing efforts.
GMS per active buyer on a trailing 12-month basis was up 2.2% year-over-year and accelerated for the fifth consecutive quarter. We have been particularly focused on increasing our number of habitual buyers, defined as buyers who spent $200 or more and made purchases on six or more days in the previous 12 months. As of December 31st, 2018, Etsy had 2 million habitual buyers, up nearly 22% compared to 2017, which grew faster than overall active buyer growth, indicating that our efforts to convert buyers into more loyal shoppers is showing signs of success. Overall, active buyers increased approximately 18% to 39 million, and active sellers grew approximately 9% to 2 million. As we've been saying, our goal is to continue to improve our buyer to seller ratio.
While we are not walking through the P&L on this call, shown on Slide 13, I do want to highlight two points that impact the year-over-year comparison of net income. First, in 2017, our net income saw a significant benefit from U.S. tax reform legislation, which we anniversaried in the fourth quarter of 2018. In Q4 of 2018, we released a valuation allowance on our foreign jurisdictions, which led to a sizable positive impact on net income. Second, our expenses in Q4 include $6 million of stock-based compensation expense related to certain employee departures. Overall net income for the fourth quarter was $41 million with diluted earnings per share of $0.32. We were really pleased with Etsy's ability to generate cash in 2018. Net cash provided by operating activities was nearly $199 million, compared with $69 million in the prior year, as shown on Slide 14.
We ended the year with $624 million in cash equivalents, and short-term investments. During the year, we repurchased approximately $135 million of common stock, including $45 million during Q4 as part of the new $200 million authorized by our board in November. Our 2018 share repurchases were made at an average price of $30.28 per share. We also recently entered into a $200 million revolving credit facility. Our strong cash position sets us up for growth in 2019 and beyond. Turning to our 2019 outlook, there is some uncertainty surrounding the macro environment, specifically global consumer spending, and we continue to closely monitor trends for potential impact to our business. We're currently forecasting sustained growth and expanding margins for 2019, with GMS growth in the range of 17%-20%, revenue growth of 29%-32%, and adjusted EBITDA margins of 23%-25%.
I want to provide some thoughts on how you should think about your 2019 models. First, in terms of the cadence of GMS growth for the year, we had a soft start to Q1 as we pulled back on some marketing investments and launched new product development initiatives for the year following our December slush period. FX became a headwind in the back half of 2018 and continues to be a headwind to our GMS growth. Second, marketing will continue to be one of the areas of investment for us in 2019. In Q1, we have pulled back on spend somewhat while we digest what we have learned from heightened spend levels in Q4 and expect marketing expense in the first quarter of 2019 to decrease as a percent of revenue compared to the fourth quarter of 2018, which could positively impact Q1 adjusted EBITDA.
Going forward, we will continue to test less mature channels such as TV and digital video, and you should anticipate that marketing expense will be back half loaded in 2019. Third, another area of investment is in people, especially product and engineering resources to improve the core product experience. Please note that a portion of these efforts are capitalized. In 2018, total CapEx was approximately $21 million, which included $20 million for capitalized web development and $1 million for property and equipment. Capitalized web development in 2018 was elevated due to capitalized expenses related to our cloud migration.
Fourth, on our cloud migration, we now expect to be fully migrated in the beginning of 2020 and anticipate spend on capacity of approximately $25 million for 2019, most of which will be expensed through cost of revenue. Fifth, we redesigned sellers' payment accounts to simplify their finances and began disbursing seller funds net of fees in a single transaction. This reduces credit card fees we incur for processing seller bill payments, and we expect this to expand gross margins by at least 100 basis points in 2019. Lastly, as usual, please also keep in mind that historically, our Q4 margins have been seasonally higher than the rest of the year. Please refer to slide 19 in our earnings presentation for specific factors that we have incorporated into our guidance. Note that we plan to file our 10-Q shortly.
I'm pleased to let you know that we have integrated our social, environmental, and economic impact metrics into our 10-K for the first time. To sum up, we're building on our momentum following a great quarter and a great year. We expect to continue to capitalize on our large market opportunity and deliver sustained growth and expanding margins. We hope you'll all tune in to our Investor Day next week, which will be available via webcast on our investor relations website. We plan to review our longer-term growth strategy, multi-year product and marketing vision, and our view of Etsy's growth opportunity ahead. Thank you all for your time today. We will now take your questions.
Ladies and gentlemen, if you have a question at this time, please press the star and then the number one key on your telephone keypad. If your question has been answered or you wish to remove yourself from the queue, please press the pound key. To prevent any background noise, we ask that you please place your line on mute once your question has been stated. Our first question comes from the line of Heath Terry of Goldman Sachs. Your line is now open.
Hi. Thanks. This is Daniel on for Heath. Just a couple quick questions from us. One around the marketing, talking about pulling back a little bit in Q1 and reassessing some of the spend there. Are you guys still considering over the balance of the year being able to reinvest the 80% of incremental take rate revenue that you had quoted a couple of quarters ago? The channels that you're looking to spend that into, what did you see specifically around those in the fourth quarter? On the take rate in the quarter, it came in a little bit lighter than we had been modeling. Just curious if there's anything from a contra revenue perspective, particularly within marketplace revenue or anything that you're seeing from how sellers are reacting to the inclusion of shipping fees and the take rate that might have impacted that. Thanks.
Great. I'll start on the marketing pieces. First I would say that if you look forward into 2019, the guidance we've provided of EBITDA margins of 23%-25% are inclusive of marketing. That gives you a sense of how we're thinking. While we haven't committed to a specific what percentage of the take rate increase goes into marketing. We will continue to see EBITDA margins expand, but you can expect that we're continuing to invest meaningfully in marketing in 2019. We also said we expect that to be more back-half loaded as we digest and absorb and consolidate the learnings from Q4, incorporate that into our investments going forward in 2019. TV is something that I know is of interest to many people.
We did spend and invest in TV in the second half of 2018. Our learnings there are, in general, we're pleased with what we saw. We're tracking two things. One is, what is it doing to our brand health metrics? The second is, how is that translating into spend in the near term? On the brand health metrics, we're seeing that TV does appear to be positively impacting our brand health metrics, particularly around things like intent to purchase. We think that that's a great leading indicator that is encouraging. In terms of translating into purchase activity, it is translating into purchase activity, not enough to fully fund the TV campaign in period, but we think that this is an investment we're making for the medium term.
We're seeing some near-term results and encouraging signs for the future that make us want to continue to invest in it in 2019 and beyond. Maybe now I'll turn over to Rachel to sort of take.
On the take rate. If you take all of our revenue and divide by all of our GMS, that is the total take rate, but it really breaks down into the elements of the take rate that are mandatory to participate as a seller on the site. There's a transaction fee rate and a listing fee rate. Nothing's changed with that. The additional services are somewhat variable, so I'm not sure exactly how you modeled it, but there's nothing from contra revenue that would be impacting those numbers specifically. If you think about it in those two components, I think you maybe have a model issue that we can look at with you at another time.
Yeah. No, I think we had just been making some assumptions around the amount of shipping that would be included in the GMS that a new take rate was being applied to. Didn't know if there was anything different than you might have expected initially with how sellers were responding with their pricing there.
Yeah, I don't think the price change necessarily. We haven't seen any deleterious effect of the price change really on seller behavior in any metric, be it shipping attach rate for our shipping label products or other forms of behavior. Shipping label attach rate is really affected by what other alternatives they have in the market, how well we've marketed that product to them, et cetera. What I am pleased with shipping is that we saw a meaningful improvement in the percentage of items that were listed with free shipping in the fourth quarter, and that translates into higher conversion rates. We think that's helpful and we're going to continue to be pushing hard on that.
Great. Thanks, guys.
Thanks.
Thank you. Our next question comes from the line of Kunal Madhukar of Deutsche Bank. Your line is now open.
Hi. Thanks for taking the question. Wanted to focus on the slide 11 that you had on moving the needle on frequency. It's good to talk about the habitual buyers. What about the ones that have multiple purchase days? How much do they spend, and how has that frequency, or has that frequency changed at all in the fourth quarter? Thank you.
Yeah. We're talking about a specific segment here that we haven't talked about before, that we've defined as habitual buyers, and that's a group of people that have bought on six or more purchase days and spent $200 or more. That's a new segment we're just starting to talk about now. I guess what I would say is that, this idea of looking at our very best buyers and using that to think about what we can learn about the rest of the buyer ecosystem is helpful and is something that we're excited to just begin on that path. A couple of things we observe is that there's a meaningful number of people out there, about 2 million, who already shop regularly on Etsy and have figured out that Etsy is good for many different purchase occasions.
When we look at that segment and compare it to other buyers, demographically and psychographically, they're similar to many other buyers. The difference is they've sort of figured out some hacks to unlock the value of Etsy. It's encouraging to us to then think about how we can make other buyers have an experience that looks more like what the habitual buyers experience. We can both continue to grow share of wallet with habitual buyers, where I do think there's a meaningful opportunity to grow share of wallet, but also to take other buyer segments and have them become more habitual. Do lookalike analysis and other things to find people on the web who maybe haven't shopped Etsy before but look like habitual buyers, and disproportionately invest in those segments.
On the slide, the multiple buyers is the next tier, so it's two or more purchases. That's the next tier down from habitual versus all the others that are the once a year people. The once a year people, we've said for quite some time, are 60% of our overall active buyer base. Moving the needle on those top two tiers is helpful to accelerate GMS growth.
Totally. Thanks, Josh. Thanks, Rachel. A quick follow-on on the marketing side and the marketing pullback, the pullback in the marketing spend for 1Q and possibly in the first half. How is the decision process kind of shaping up in terms of when do you think you'll get to a point where you know that TV maybe works or social maybe works better, and you start putting more dollars against each of those mediums?
Yeah, great question. Broadly speaking, we think about marketing in two buckets. There's a proven set of investments we make, where we're always testing and optimizing to see how we can shift the ROI curve up. Can we invest more at the same return or get better returns at the existing spend? Google right now, there's several Google channels that really are where we see proven ROI, particularly in Google Shopping, so product listing ads. As we took up our take rate, and as we invest in more capabilities, we're leaning into when we spend more, what happens, and sort of how do we follow the marginal return curve down. The only way to know that is to increase your spend and test it.
You saw us increase our spend to find out what the limits of, and what the shape of that marginal return curve are. We're going to be challenging our team to continuously improve our capabilities so that we can move that curve up over time, and in a competitive marketplace where other people are bidding against us and where the dynamics change. We're going to need to continue to test. The second bucket is new channels, and we mentioned paid social and TV. In those, we've got to test and learn to, A, find out whether we can make them ROI positive, and then, B, once we do, figure out the return curve of those channels as well and make investments to make them more effective.
On TV, for example, we were encouraged enough by what we saw in the fourth quarter to make us say that we're going to continue to invest in 2019. We're developing new creative right now that incorporates the learnings from Q4. We also learned things, for example, on in front of what audiences, at what time of day do we seem to be getting better performance. By taking a minute now in the first quarter to consolidate those learnings and translate that into fresh creative and a media strategy that can incorporate those learnings, we think we can put that to work more effectively. You'll continue to see us test and learn, and I would expect that we'll be elevating our investment, our marketing investment in some quarters, and then I think it's healthy every now and then to pull back a little bit, too.
The other thing we learned when we pulled back is incrementality. How much of this would you have gotten anyway had you not invested is one of the things that we as marketers are always asking ourselves. A good way to test that is to pull back and find out what happens to organic traffic. We're going to continue to test and learn, but again, our margins, the EBITDA margins that we've guided to for 2019 do suggest that we're going to continue to have meaningful investments in marketing through the year.
Thank you.
Thank you.
Our next question comes from the line of Nick Jones of Citi. Your line is now open
Hi, thanks for taking the question. After testing some of these new marketing channels, are you finding buyers buying across different categories on the site, or are people still kind of focused on one category and still need to learn about the different offerings Etsy has?
It's a great question. There's, broadly speaking, two kinds of engagement we can be driving. One is to bring new buyers to the site, the second is to get existing buyers to engage again or more often. Candidly, when we talk about investing in marketing capabilities, we're also relatively early at focusing our spend against those segments differently, and I'm very excited about the roadmap over the next few years to get better and better at that. What is the value of getting someone who's only engaged in one category to now engage in a second category? When you're acquiring a buyer, buyers coming in in certain categories are likely to be higher value than buyers coming in in another category. Candidly, we incorporate some of that into our analytics, but there's an opportunity for us to get more robust in doing that.
When we talk about investing in marketing capabilities, that's part of what we're talking about. We're going to be talking about a lot of these topics at Investor Day on March 7th. We'll go deeper on them.
Great. Just one follow-up kind of on the same line of thought. With kind of focus on paid marketing channels, is there an opportunity to bolster organic search and maybe get some free channels by focusing internally as you kind of develop search and discovery on platform? Can that kind of parlay to off platform?
We're proud of the fact that the substantial majority of our traffic today does come to us for free. Either people coming to their web browser and typing in Etsy, launching the app, finding us on SEO. Quite a lot of it comes direct to Etsy. We also have a significant social following, people who are posting about Etsy on Instagram and on Twitter and other places. I think the opportunity to leverage that advocacy off of our platform is an opportunity that we can be, and plan to be, investing more in and pushing harder on.
Definitely have product people focused on SEO as a channel to improve that kind of free source of traffic as well.
Yeah.
Great. Thanks for taking my questions.
Thank you.
Our next question comes from the line of Edward Yruma of KeyBanc Capital Markets. Your line is now open.
Hey, good afternoon, guys. Two quick ones from me. First, on the weakness exhibited quarter to date, I know you cited marketing or the pullback of marketing spend as a potential cause. I guess, are you seeing any other consumer behaviors that have changed? Second, very nice holiday results, particularly the GMS growth. I know that you've been attempting to try to extend the selling period for holiday, and I know you were not successful last year. Were you able to actually move the center of gravity during that key holiday season? Thanks again.
I can start, and Josh will jump in. The answer to the second question first is, not really. I think we saw more growth during the periods of time that are important to us, but we didn't really extend the number of days. We did have substantially more estimated expected delivery date transparency on all of our shipping, and we'll continue to work on shipping, and we'll talk more about that at our Investor Day. On the Q1, I think we saw macro softness as I think many companies did. There was polar vortex, there was government shutdown. I mean, who really knows, to be honest? I think maybe an overall macro haze in addition to our own decision to sort of put some marketing channels on hiatus so that we can better evaluate and test. I call it test and rest as we move forward.
I think the combination of those things are what we saw in the beginning of Q1.
Great. Thanks so much.
Our next question comes from the line of Mark Mahaney of RBC Capital Markets. Your line is now open.
Great, thanks. This is Shweta for Mark. Two questions, please. One on shipping. You said a third of listings are now shipped for free. That's up from 20%, I believe, in the quarter before. Do you guys have a goal for this year? Do you think it can reach 50%, or how is just the education of sellers going so far? Clearly good. Any update on that in terms of this year's goal? Then the second, any update on international post integration and some of the key focus areas now, in Europe or any other markets? Thanks.
Yeah, thanks, Shweta, for the questions. On free shipping, we haven't published a goal for 2019, but this is definitely something we're going to be talking about a lot at Investor Day, and we think that more is better. We know that buyers prefer free shipping and that sellers win when they offer it in most cases. We're going to be working hard to educate them about that and provide the right incentives to make that happen.
It's in our release also, but we also said that 80% of our shipping was available at competitive prices. We've talked about this in the past, that we're really focused on what is reasonable shipping. Free is great, but also what's reasonable. We don't want to be excessive shipping price. That was another focus in the quarter that we feel we've succeeded at.
Yep. International GMS grew at a rate faster than the core business. Again, we feel good. It was robust. One highlight I'd point out is even with all of the Brexit mania that's happening, we had a robust Q4 in the U.K. In both the U.K. and Germany, those businesses are now more than 50% domestic businesses, meaning the buyer and the seller is in the U.K. or Germany respectively. We think that's terrific. Getting those businesses to continue to grow and getting the other core markets to be increasingly domestic trade markets is really our focus at the moment.
Okay. Thank you, Josh. Thanks, Rachel.
Thank you.
Our next question comes from the line of Thomas Forte of D.A. Davidson. Your line is now open.
Great. Thanks for taking my question. I had two things I want to talk about. One is, when you raised price last year in the second half of the year, you talked about how you were going to invest all the incremental dollars. Should we think about those investments as having an impact only on 2019 revenue growth? Or is the impact expected to be more further out? Second, there's a lot of talk recently about changes to the regulatory environment in India. I know India is not a big area of focus for you right now, but have those changes in regulation at all affected your view of India longer term?
Yeah. I mean, no. Of course, our ROI, the calculation is based on a lifetime value. Lifetime would be beyond this fiscal year. We're still evaluating LTV channel by channel. TV is a very new channel for us, then we're also exploring the right mix of TV and digital video. I don't know the slope of the tail specifically yet for those new channels, and other less mature. That's exactly what we are testing, is how much of the yield comes in a future year. We've also coupled those upper funnel types of marketing spend with consumer sentiment research. How much are we impacting intent to purchase and sentiment around the brand?
How much reach and frequency, how much staying power you get with the more reach and frequency you're getting with that sort of spend. I think there's a thesis there that the LTV of upper funnel market, I mean, it's not my thesis, it's the marketing thesis, that upper funnel marketing is intended to have more of a stickiness to it and a longer lifetime value, thus we're trying to test for ourselves what we believe that number to be. Right now, our current guidance for both GMS and EBITDA margins would indicate what we're expecting to get in the year.
Now turning to India, we're making some relatively small investments in India with a hugely talented team of folks on the ground to really drive some export activity there. The insight is that India has a long history of wonderful craftsmanship and artisanal products, there's big demand for them in other markets that we serve. For example, we find that when a buyer in, for example, the U.S. or the U.K. buys from someone in India, the lifetime value of that buyer goes up. We thought that was very interestingly encouraging. Due to the nature of the Indian market, hiring people who can go village to village, actually help artisans to get online and sell on Etsy is actually cost-effective in a way that it wouldn't be in many other markets.
It also allows us to get to know those sellers and personally vet them and make sure that their practices and procedures are things that we're comfortable with and fit within Etsy. We're not currently focused on building a domestic business in India. We're definitely always paying attention to the regulations and laws of each market in which we operate in. Should we want to make investments to build a domestic market in India, there's a bunch of regulations that, of course, we are always paying close attention to now and would have to be part of our consideration about that going into the future.
Great. To quickly recap, the investment spending as a follow-up to the price increase is intended to drive long-term performance, not just 2019, and it doesn't sound like there's any change in regulatory environment that'll change how you're pursuing India.
Both correct.
Thank you very much.
Thanks, Tom.
Our next question comes from the line of Darren Aftahi of Roth Capital Partners. Your line is now open.
Yeah, good afternoon. Thanks for taking my questions, and congratulations on the quarter. Just two, if I may. I think you called out in the release, your Promoted Listings business, you expanded that. I'm just kind of curious, can you maybe quantify how much demand is outstripping supply? If not, kind of perhaps what inning we are in that expansion. Then second question, on those that are doing free shipping over, say, the second half of 2018, what kind of relative lift are you seeing in perhaps like GMS per seller relative to those who are maybe offering competitive or maybe uncompetitive shipping levels? Thanks.
Okay, let me try on both those. Promoted Listings first in terms of supply and demand. Sellers give us more funds than we are currently able to spend by a fairly meaningful amount. It's about having the right inventory on the site, which is the right number of visits and how much real estate per visit. Then as importantly, picking the right listing to show for any given visitor. Then optimizing the bidding algorithms for how much people are willing to bid for each of those. Those are kind of the three components, right? Views times click-through rate, times cost per click. On those, I think we've talked about in the past how we've expanded the amount of inventory fairly significantly. There can be probably some more inventory that's given over to Promoted Listings in time.
We think we've probably mined that reasonably well. We are at the early stages of having really good relevancy algorithms to pick exactly the right listing per visitor, that's a very similar journey to search in general. There's a lot more we think we can do around personalization and understanding item quality and all kinds of things to pick the right listing for the right visitor at the right time. As we do that, we'll get more ability to raise CPCs. It's all really about delivering quality traffic that converts to sales for buyers. I think we're still in relatively early innings on Promoted Listings, but it's going to be more and more about how good are our algorithms. The opportunity to help our sellers invest in their growth, not just on Etsy, but off Etsy.
We know that there's a significant appetite from sellers to invest, to buy traffic from other sources, driving it to their Etsy shops. Thinking about ways to help them with that, we think, is also a rich opportunity going forward. The second question-
Was about shipping.
Shipping. Remind me again. I'm sorry.
The impact. We haven't disclosed The short answer is we haven't disclosed what impact there is to conversion rate. I think it's incorporated into the successful GMS growth we had during the holiday period, that we were competitive with other merchants out there and offering free shipping. I think, we believe, and we'll talk more about this at our investor day, we believe that shipping continues to be one of the areas of friction that we have in our shopping experience that we need to solve for. Without giving you a direct answer to correlate exactly how much increase in shipping has how much increase in GMS, we're continuing to work very hard on that area. It's not just shipping free or not free, it's shipping is a reasonable price relative to the competition.
It's what do you do about return shipping and the transparency and communication with which our buyers are able to have confidence that they're going to get their item in time, or when are they going to get their item, and we're working on all of those factors. One of the things we've done in the last quarter was to bring online not only USPS, but we also have Canada Post, Royal Mail. We've brought on Australia Post as well. Through those deals, we're able to offer better and better shipping rates to our sellers, as well as give them the ability to calculate shipping, which is the first step in being able to also track packages and other things, and being able to get more consistency around what is the price that a buyer's going to receive.
We're working on a number of levers relative to shipping, which we think over time is going to impact GMS growth.
Thank you.
All right.
Thank you.
Our next question comes from the line of Ygal Arounian of Wedbush Securities. Your line is now open.
Hey, good afternoon, everyone. Thanks for taking my question. I have a couple questions. Maybe I'll just ask the first one and then ask a follow-up. Obviously strong GMS growth in fourth quarter, especially when you think about the tough comps from last year. Were there maybe one or two drivers that drove the strength more than the others? I know you called out a number of things, but was there one or two that kind of had a outsized impact relative to everything else?
Broadly speaking, there's two big levers. There's product and marketing, and both of them played a meaningful role. It wasn't really one or the other. Talking about product for a second, first I would say that the work we did throughout 2018 paid off in the fourth quarter. Continuing in 2018, we continued to improve the quality of search and discovery in a bunch of ways, and that was very helpful. We talked about shipping, but that mattered. Having more items that are free shipping in the fourth quarter mattered. Shining a brighter light on things that can be personalized or customized was very helpful. Making items that were on sale a little bit more visible was very helpful.
I'd point to a number of things that we did in the product, that really made the buying experience more competitive and more exciting in the fourth quarter. That made a big difference. One other thing I'd say about product is we really focused this quarter on only doing things that were not just helpful in the holiday season, but that could help us all year long. Something that was going to be really great for six days and then not matter for the rest of the year, we decided the opportunity cost was just too high. We did things that we think were really beneficial in the fourth quarter and repeatable. Then as we discussed, we leaned harder into marketing. That was helpful. A good amount of that was things that were ROI positive and sort of paid for themselves right away.
Others were things where we're testing the limits of marginal return and we found those limits. You see us pulling back a little bit in Q1 and reconsolidating to be prepared to continue to invest in the rest of the year. Broadly speaking, it really wasn't one or two things. There were a number of things in product that really made a difference. Our marketing investments overall, we feel good about in the fourth quarter.
Great. That's helpful. I guess I wanted to dig in a little bit deeper on marketing. Maybe so partially tie it into the paid GMS, which is obviously up really strong, up 55% year-over-year. It's now 20% of the total. I guess first, are things like TV and radio brand advertising, is that built into the paid GMS number, or is that kind of strictly performance marketing types of things? How do you see the long-term balance of what's an appropriate level of paid GMS versus organic? Second, as you think about the marketing, it feels like a lot of it worked really well and it contributed well to the fourth quarter GMS growth. Now you're kind of stepping back. There's some things you're reassessing.
You're talking about re-accelerating again in the second half, some quarters where you step back and some quarters where you push again. On some of the things that you've been testing, is there a chance that you kind of reach the maximum point as you talked about the marginal return, it doesn't make sense to re-accelerate, there's a chance that we don't get back to the levels that we saw in Q4? Thanks.
Yeah. We absolutely are very, very conscious of those very things. What we're trying to do is grow profitably, I think we've been doing that. We're not afraid to experiment and to learn from those experiments. The performance marketing dollars is what drives our percentage of paid GMS. We said that there's about $112 million of performance marketing dollars in 2018, that's driving the 20% of GMS coming from paid. There's an amount that you'll see in our 10-Q that we say we spent $129 million total in marketing, the balance of it would be inclusive of what we've spent on brand marketing and sort of upper funnel initiatives. We talked about doing a test in the September-October timeframe to learn that we knew wasn't ROI positive.
We turned that into a national campaign that we ran in November, December that we got closer to being ROI positive. We want to continue to optimize. For instance, one of the ways we optimized there was we initially had two different creative campaigns. When we ran it nationally, we reduced it to one creative campaign. Even that one creative campaign had two different versions, we learned one of the versions was performing better than the other. We took one of them down to 25% of the rotation, the higher performing one stayed up in 75% of the rotation. We learned that certain networks perform better than other networks. As we optimize, we get closer and closer and closer to being ROI positive just on its own merits.
What we want to do now is test new creative, we want to also learn from the consumer sentiment, the intent to purchase, the top of mind awareness. We couple that with the actual quantitative results. If we don't feel comfortable that it's actually yielding for us, we won't spend it anymore. You sort of see us doing that here in Q1, where we said we pulled back a bit because in Q4 we elevated our spend, we might have pushed it past a line that was ROI positive, we're pulling back to really assess what's truly incremental in our spend and what isn't. You sort of are getting the commitment from this management team that we're going to spend if we believe it's ROI positive, it's not, we won't do it.
We've given you the guidance that we believe would allow us to test some less mature markets and continue to invest as aggressively as we can in mature markets that we believe are ROI positive. We'll continue to update you on that every quarter. I'm sorry for the long-winded answer, I think there's some blend in there where there's some things that are not ROI positive yet, and we will stop spending them if we don't get conviction that they are ROI positive.
Right. Long term, Google is the one channel right now that we feel we've got a high degree of conviction that at certain spend levels, we get strong ROI. It stands to reason that we ought to be able to make others of these channels work. It really seems that we ought to be able to. We're testing and learning our way in. If your question is it possible that they don't work? Yeah, it's possible, we're going to test and learn, and I believe that we can make several more channels that are scalable, potentially substantial channels work in time, and that these will have been great R&D investments that pay off as we scale these new channels.
Thanks so much. I appreciate the long-winded answers. Thanks, guys.
Our next question comes from the line of Laura Champine of Loop Capital Markets. Your line is now open.
Thanks. First, a quick housekeeping one. Can you give me headcount at the end of the year? Secondly, can you talk a little bit more about your cloud conversion, how that is proceeding at or off pace, and what you expect to learn and be able to gain as that conversion progresses this year?
Yeah. Headcount 874.
874 headcount. On cloud, it's going apace. We're pleased with the progress. It's going largely on track. Substantial parts of Etsy are now being served from the cloud. And substantial parts of our machine learning are happening in the cloud. Specifically, the training of the models is now happening in the cloud, but the real-time execution of the models is still happening on-premise as of this moment, to give you a little peek under the covers. What does that really mean? The cloud is not an end to itself. It only exists as a means to the end of having a better experience on Etsy. For example, by having machine learning model training in the cloud, we can now start to run more sophisticated models. We're not as capacity constrained as we'd been when we were in our own data centers.
That starts to translate into more robust algorithms to power better search and to do it in a more personalized way. We're excited about that, and we're also working on how do we have the right cost discipline so that we don't use more compute than we need, but we also don't unnecessarily constrain that compute so that we can really take advantage of a lot of these new models and technologies.
Great. Thank you.
Thanks, Laura.
Our next question comes from the line of Marvin Fong of BTIG Capital Markets. Your line is now open.
Great. Thank you for taking my questions, and congratulations on a very nice quarter. Just a couple quick ones. First, you mentioned in the press release that the conversion rates for desktop, mobile web, and mobile app were all up, but that mobile web still seems to be about half at desktop, and I guess desktop is now tied with mobile app for the best conversion rate. Could you just talk a little bit more about, should we think that mobile web will kind of stay at half the conversion rate of desktop, or is there room for that relative percentage to go up?
Yeah, good question. I think where we're evolving to right now, largely the desktop, the mobile web, and the Buy on Etsy app are largely the same thing presented on three different platforms. We're moving to a place where each of those platforms serves a different job, sometimes with a different audience, and we're learning to tailor the platform to the audience and the job. For example, mobile web is disproportionately newer users and wants people become really habitual buyers. For example, habitual buyers disproportionately use the mobile app. Starting to think about what is the purpose of each of these platforms, and therefore, how do you tailor it to the right audience, I think is appropriate. Mobile web is a lighter experience, and it may have a less convicted audience, it may have a lower conversion rate than other things.
That doesn't mean that we're satisfied with the conversion rate, I believe it can and should go up over time as we execute well. I continue to think that should be true of the Buy on Etsy app and the desktop as well. How they will perform relative to each other is hard for me to predict right now, but I see opportunities across all three.
Great. Thank you. Just one other quick housekeeping. If you're able to speak to these statistics that you've given in the past, what was the % of GMS from repeat buyers? I apologize if I missed that. Also, what % of sellers are using Promoted Listings? Thank you.
First, let me start by defining what we call new versus repeat. We call a new buyer somebody who's never ever used Etsy before, and a repeat buyer is somebody who has ever used Etsy, even if it was five years ago. I think last quarter, we said it was 83% of our GMS was from repeat buyers. This quarter, it's about a little over 80%. This time, we started talking about the other split is the habitual versus multiple versus active, just a regular active buyer. We're really focused on those segments of buyers that because some of our marketing investment is actually intended to lure back in buyers that might not have been on Etsy site for a number of years, that other companies might characterize those as new buyers. I just wanted to be clear about that.
I'm sorry, was there one more part to your question because?
Yeah, just the percentage of sellers who are now using Promoted Listings, if you're able to speak to that.
Yeah. It's 16% of sellers have adopted Promoted Listings. I would call out that not all sellers are equal in terms of their sales on Etsy. One thing we've talked about in the past is that some of our largest sellers are disproportionately represented on Promoted Listings. I just want to caution us away from thinking that the number one way to grow Promoted Listings is to grow the percentage of sellers that are currently adopting it. We've shared that the amount of investment capacity they've given us is well in excess of our ability to spend it. That gives you some sense of adding more sellers may not directly translate into driving more Promoted Listings sales. It's more around improving the relevance of our algorithms and bidding engines and things like that.
Okay. All right, I understand. It's actually up from, I think it ended 2017 at 15%, it did improve, that's-
Yeah. We're very focused internally on thinking about stages of growth. Someone who's brand new to Etsy, who's listing their first listing, Promoted Listings isn't necessarily the first thing that we would encourage them to do. You want to see them get some success, really learn how Etsy works, have high-quality listings, high-quality photos, get some feedback, and once they've really got a bit of a track record, and understand how it works, then they might be ready to go to Promoted Listings. Thinking about speaking to the right audience with different tools, including Promoted Listings, is something we're very much moving in that direction.
One other point I'll make is that because active sellers grew by 9%, the denominator got bigger. The actual count of sellers that took Promoted Listings grew by more than 16%.
Right.
Got you. Okay, great. Thank you. Thank you very much, and congratulations again.
Thank you.
As a reminder, ladies and gentlemen, if you have a question at this time, please press the star and then the number 1 key on your telephone keypad. If your question has been answered or you wish to remove yourself from the queue, please press the pound key. To prevent any background noise, we ask that you please place your line on mute once your question has been stated. Our next question comes from the line of Ronald Bookbinder of IFS Securities. Your line is now open.
Good evening, yes, congratulations on an excellent execution this past year. I have a question about your use of cash. You have very strong cash flow, and it seems to be even getting better. You've done a great job of buying back shares at advantageous prices. Are there opportunities to use the cash for other things, possibly helping to finance the merchants as they build inventory towards Holiday? What do you look for as uses of cash?
High return on invested capital is, I guess, the guiding principle there. We really look pretty open-mindedly at all the things we could build, buy, or partner with. We look at the net present value of all of those opportunities. Last year, for instance, in addition to share repurchase, we entered into a referral agreement with DaWanda in Germany, which allowed us to not only migrate many of their sellers over to Etsy's platform, but we redirected all of their traffic to Etsy, and that was an investment of about $35 million in cash at the time we did that transaction. That's one example of putting cash to work to grow a geography faster. We've looked at other things that, opportunistically, that we've determined weren't going to be a positive NPV for us, and we've decided not to pursue.
We're constantly evaluating those sorts of strategies. Of course, putting our dollars to work on internal investments. We're building up our product development engineering bench so that we can, quote unquote, "build things" faster that we think will have a positive yield as well.
Given the strong cash position that you ended the year at, why do you feel the need to take out a $200 million revolver?
I would just characterize that as good corporate housekeeping. We have access to capital at attractive rates. There's very low unused fees on that balance, that gives us access to capital markets at a time when we can, there may be a recession in the future when the credit markets would seize up, we wouldn't have access to that capital. It's just, for us, a good policy to have that kind of flexibility and access to cash, so we have it if and when we would need it.
There's not some sort of major acquisition on the horizon?
Well, if there were, you know I couldn't answer that question, really.
All right. Well, thank you for taking my questions. Once again, congratulations on a nice finish to the year.
Thank you.
Thank you.
I'm showing no further questions at this time. Ladies and gentlemen, thank you all for participating in today's conference. This does conclude the program, and you may all disconnect. Everyone, have a great day.