Ladies and gentlemen, thank you for standing by, and welcome to The RealReal's Third Quarter 2019 Financial Results Conference Call. At this time, all participants' lines are in a listen-only mode. After the speaker's presentation, there will be a question and answer session. To ask a question during the session, you will need to press star one on your telephone. Please be advised that today's conference is being recorded. If you require any further assistance, please press star zero. I would now like to hand the conference over to your speaker today, Mr. Paul Bieber, Head of Investor Relations. Thank you. Please go ahead, sir.
Thank you. Good afternoon, and welcome to The RealReal's earnings call for the quarter ended September 30th, 2019. I'm Paul Bieber, Head of Investor Relations at The RealReal. Joining me today to discuss The RealReal's results are Founder and CEO, Julie Wainwright, and Chief Financial Officer, Matt Gustke. Julie will provide an update on our business, including progress on a few key initiatives, and then Matt will review our Q3 financial results and provide a financial outlook. This conference call will be available via webcast on our investor relations website at investor.therealreal.com. I'd like to take this opportunity to remind you that during this call, we'll be making forward-looking statements, including statements relating to the expected performance of our business, future financial results, strategy, long-term growth, and the overall future prospects.
These statements are subject to known and unknown risks and uncertainties that could cause actual results to differ materially from those projected or implied during this call. In particular, those described in our risk factors included in our final prospectus for our initial public offering filed with the SEC on June 27th, 2019, and the risk factors included in our Form 10-Q that was filed with our second quarter results, and the Form 10-Q that will be filed with our third quarter results. You should not rely on our forward-looking statements as predictions of future events. All forward-looking statements that we make on this call are based on assumptions and beliefs as of today, and we undertake no obligation to update them except as required by applicable law. Our discussions today will include non-GAAP financial measures.
These non-GAAP measures should be considered in addition to and not as a substitute for or in isolation from our GAAP results. A reconciliation of GAAP to non-GAAP results may be found in our earnings release and supplemental materials, which is furnished with our Form 8-K filed today with SEC and may also be found on our investor relations website. I would now like to turn the conference call over to The RealReal's Founder and CEO, Julie Wainwright. Julie?
Thanks, Paul, and good afternoon, and thank you for joining us for our third quarter earnings call. We are once again excited to share our progress with you and provide you with details on our third quarter financial performance. Before I proceed, I'd like to take this opportunity to thank our consignors, our buyers, and our employees for helping make The RealReal into the world's largest online marketplace for consigned luxury goods. Q3 was a very strong quarter and speaks to the health and vibrancy of our marketplace. We generated GMV of $252.8 million, a 48% year-on-year increase, and revenue of $80.5 million, a 55% year-on-year increase. Importantly, GMV growth accelerated approximately 800 basis points quarter-on-quarter, and revenue growth accelerated approximately 400 basis points quarter-on-quarter.
We are especially pleased that we drove our growth while once again driving significant marketing leverage. We also demonstrated meaningful leverage in our operations and technology expense line, which is important as we continue our march toward profitability. We are proud of the accelerating growth and operating leverage we demonstrated during the quarter, which we believe speaks to several unique aspects of our model, including high buyer repeat rates and our unique flywheel, where buyers become consignors and consignors become buyers. Matt will provide more color and the details of our quarterly financial performance later in this call. During the call today, I will briefly touch on three topics. Q3 performance strength, automation, and sustainability. First, I'd like to spend time on our Q3 performance. Q3 was a strong quarter across the board with strong financial results and significant progress against our priorities.
Our IPO had a positive impact on our top-of-funnel marketing activities, such as traffic and member growth. Q3 traffic increased 44% year-on-year, and members increased 76% year-on-year. As of September 30th, The RealReal has 14 million members. The strong traffic and member growth translated into strong active buyer additions. We ended Q3 with 543,000 active buyers on a 12-month trailing basis. That is up 43% year-on-year and a 300 basis point increase quarter-on-quarter. We added approximately 51,000 new buyers quarter-on-quarter, a record for us. Importantly, we achieved record new active buyer additions while also driving marketing leverage with only a 25% year-on-year increase in absolute marketing expense dollars.
Of course, because of the robust revenue growth, marketing as a percent of revenue improved significantly to 16.5% compared with 20.4% in the same period last year. Our buyer acquisition costs, or BAC, continued to fall with the Q3 BAC declining approximately 15% year-on-year. So far in Q4, new buyer additions have normalized to our pre-IPO levels. In addition to the top of funnel strength, Q3 AOV was also very strong and increased 5% or +$20 year-on-year. AOV benefited from a mix shift in the quarter that favored higher AOV categories such as watches, jewelry, and handbags. Matt will elaborate later in the call. Now let me shift to my second important topic, automation. It accelerates and facilitates our ability to scale efficiently and improve our unit economics.
As we discussed on our second quarter call, the first step in automating our inbound operations was in pricing. We exited the second quarter automating pricing of 52% of total unit volume, and by the end of Q3, pricing automation had increased to 61% of unit volume. While we have made significant progress on pricing automation, we also exercise human oversight over pricing to ensure we capitalize on opportunities in the market. In addition, in Q3, we began automating copywriting and photo retouching. We exited Q3 automating 17% of copywriting, including product and title and description, and 15% of photo retouching. We expect to see steady increases in the percent of items with automated copy and retouching, which will support improvements in inbound operations unit costs, and importantly, significantly increase the scalability of our operations. A big thanks to our teams executing on our automation initiatives.
Our automation progress contributed to our improving operations and technology expense as a percent of revenue. The anniversary of the L.A. store also contributed to leverage in our operations and technology expense line. I'd like to wrap up today discussing a core value at The RealReal, sustainability. Our value proposition represents strongly with the millennial and Gen Z consumers, who are powering the growth of luxury sales globally. Our survey data indicates that approximately 32% of buyers shop The RealReal as a replacement for fast fashion. Also, 56% of our consignor base and 64% of our millennial consignor base cite environmental impact or extending the life of luxury items as key motivators for consigning with us.
This is not surprising when you consider approximately 3,600 dump trucks full of textiles will be deposited in the landfills around the globe just during the time it takes to conduct this call. The good news is that building a business that extends the life of luxury goods, The RealReal is having a profound positive impact on the environment. We work with Shift Advantage, a sustainably focused consulting firm, to quantify the positive impact of extending the life of consigned products on our marketplace. We are proud to announce that since inception, consignment with The RealReal has offset 12,231 metric tons of carbon. Said another way, we saved 553 million liters of water. In addition, we are pleased to partner with luxury brands who also deeply care about improving the fashion industry's impact on the environment.
We have a long-standing partnership with Stella McCartney, and we recently announced a partnership with Burberry. We look forward to working with others who share our goal of making fashion and luxury more sustainable. Lastly, I'm going to close with a quick update on our retail expansion. We are excited to finally bring a retail store to our hometown of San Francisco. We signed a lease for a store in Union Square, and we expect it to open the first half of 2020. With that, I'll turn it over to Matt for the Q3 financial review.
Thanks, Julie. Good afternoon, everyone. Let's get into it and discuss the results for Q3, after which I'll provide an outlook for Q4. We had a very strong Q3, highlighted by accelerating top-line growth, which included a strong tailwind across our top-of-funnel metrics immediately following our IPO, and a significant operating leverage in marketing and operations and technology. Our Q3 results underscore our continued focus on balancing growth with a disciplined approach to driving operating leverage. Moving on to our key operating metrics. Trailing 12-month active buyers in Q3 were 543,000, up 43% year-over-year. GMV from repeat buyers was 81.8% of total GMV in Q3, reflecting strong buyer retention and accelerating new buyer growth in the period. We generated $252.8 million in GMV, an increase of 48% year-over-year. GMV growth accelerated 800 basis points quarter-over-quarter.
Trailing 12-month GMV per active buyer was approximately $1,700, flat year-over-year. Q3 orders were approximately 577,000, up 41% year-over-year. Consistent with expectations that we articulated on last quarter's call, AOV increased year-over-year to $438, a $20 year-over-year increase, or 5%. This 5% year-over-year AOV increase was driven primarily by an increase in average selling price per item, while units per order were up modestly year-over-year. Discounting was flat year-over-year. At a category level, all of our top-level categories experienced growth in excess of 35% year-over-year, with women's, watches, and jewelry all experiencing quarter-over-quarter acceleration. Men's and handbags were the fastest-growing categories in Q3, with men's strength in sneakers and streetwear. Our Los Angeles store continued to aid men's category growth.
Returns and cancellations were 26.2% of GMV and were down 150 basis points year-over-year, driven primarily by lower cancellations as a result of strong performance in our fulfillment operations. Q3 consignment take rate was 36.8%, an increase of 40 basis points year-over-year. The year-over-year increase in take rate was up more modestly versus Q2, as we experienced more pronounced strength in watches, jewelry, and handbags in Q3, which generally carry lower take rates, but also contributed to our strong increase in AOV. We expect year-over-year take rate to increase in the fourth quarter and a modest year-over-year increase in full year 2020, as we fully anniversary our February 2019 take rate adjustment. We also note that take rates can vary from quarter-to-quarter based on the mix of products sold, as well as which consignors had item sales.
In a steady state, we expect take rates to be highest in the second and third quarters of the year, and to decrease in Q4 with a higher mix of high-priced products. Moving on to the P&L. Total revenue in Q3 was $80.5 million, an increase of 55% year-over-year. Revenue growth outpaced GMV by 700 percentage points, primarily due to higher take rates, and accelerated 400 basis points as compared to Q2's year-over-year growth. Q3 consignment and service revenue was $69.8 million, up 53% year-over-year and accelerated approximately 900 basis points quarter-over-quarter. Consignment and services revenue includes approximately $5 million of revenue from shipping fees and our subscription program, First Look. Direct revenue was $10.7 million, up 75% year-over-year.
As a reminder, we primarily generate direct revenue when we accept returns from buyers after we have already paid the consignor. In such instances, we recognize the gross proceeds as revenue when the goods subsequently resell. Q3 gross profit was $52.2 million, an increase of 57% year-over-year. Gross profit per order increased by 11% year-over-year to more than $90. Our consignment gross margin was 72.1%, up 90 basis points year-over-year, driven by a higher take rate. Our direct gross margin was 17.6%, up 540 basis points year-over-year. Direct gross margin is lower than consignment gross margin because direct revenue is recognized on a gross basis with corresponding cost of sales. Moving on to operating expenses. Please note that I will speak about OpEx on a non-GAAP basis, excluding equity-based compensation and related taxes.
For our reconciliation to GAAP, please refer to our earnings release. Marketing expense was $13.2 million in Q3, an increase of 25% year-over-year. Marketing as a percentage of revenue improved to 16.5% compared to 20.4% in the same period a year ago. Importantly, we demonstrated approximately 400 basis points of year-over-year marketing leverage, driven by an approximately 15% year-over-year decline in buyer acquisition costs, while also delivering accelerating GMV and revenue growth. We believe there's continuing opportunity to improve marketing leverage going forward, driven by, one, buyer retention. 81.8% of GMV came from repeat buyers in Q3. Two, network effects within our marketplace, including our flywheel, where buyers become consignors and consignors become buyers. Three, continued media mix optimization, and finally, conversion to buyers from our long tail of members.
We note that marketing leverage will vary from quarter to quarter based on the timing of our advertising spend. In Q4, we expect a year-over-year decline in absolute marketing dollars as we executed on our plan to front-load higher ROI marketing spend earlier in 2019 and invest less in lower ROI marketing in Q4. Operations and technology expense, which includes costs relating to our stores, Luxury Consignment Offices, fulfillment centers, merchandising, engineering, and product management, was $36.3 million in Q3, an increase of 30% year-over-year. Operations and technology as a percent of revenue was 45.1%, an improvement compared to 54% in the same period a year ago. This improvement was driven by productivity in our inbound operations from the automation investments that Julie discussed and the anniversary of our L.A. store opening.
We drove approximately 900 basis points of ops and tech leverage on a reported basis. Excluding the impact of $2 million related to a settlement payment in connection with the early termination of a vendor services agreement. That was backed out of our EBITDA reconciliation in Q3 2018, we drove approximately 500 basis points of ops and tech leverage in Q3. We believe there's continuing opportunity to improve operations and technology leverage going forward, driven by automation and occupancy leverage. Selling, general, and administrative, or SG&A expense, was $27.2 million, up 81% year-over-year. SG&A as a percentage of revenue increased to 33.7%, compared to 29% in the same period a year ago, driven primarily by investments in our administrative function headcount to support being a public company, as well as growing our sales team in advance of an expected significant ramp in Q4 supply volume.
Our adjusted EBITDA loss for Q3 was $20.9 million, or 26.0% of revenue. Importantly, we demonstrated 470 basis points of year-over-year EBITDA margin leverage while driving accelerating top-line growth. At the end of the third quarter, cash equivalents, and short-term investments totaled $370.3 million. Moving on to guidance. We expect Q4 GMV of $292 million-$300 million, representing a year-over-year growth rate of 34%-37%. This guidance reflects quarter-to-date trends in demand and supply, a return of top of funnel metric growth to pre-IPO levels, and a year-over-year decrease in marketing expense that is consistent with our previously articulated plans. We expect our Q4 EBITDA margin loss percent in the range of 14%-15%, which represents a 15-16 percentage point year-over-year EBITDA margin improvement. We expect the following factors to drive significant Q4 operating leverage. First, gross profit per order.
We expect gross profit per order to increase year-over-year and quarter-over-quarter. Marketing. We expect a year-over-year decline in absolute marketing dollars to result in a year-over-year and quarter-over-quarter improvement in marketing as a percent of revenue. Ops and tech. We expect automation and occupancy leverage to drive a significant year-over-year improvement in ops and tech as a percent of revenue, consistent with Q3 as a percent of revenue on a reported basis. SG&A. We expect the de-leveraging we experienced in Q3 to moderate in Q4. For the full year, we are raising our GMV outlook and now expect 2019 GMV of $997 million to $1.005 billion, representing a year-over-year growth rate of 40%-41%. We are also revising our 2019 EBITDA margin range to a loss of 23%-24%, compared to previous guidance of 24%-25%.
A few final notes for those building models. We expect approximately $3 million- $4 million in stock-based compensation expense in Q4, and our fully diluted share count, including unvested options as of September 30th, 2019, was 95.2 million. In closing, we are pleased to report strong Q3 results, highlighted by accelerating top-line growth and strong operating leverage. We continue to focus on top-line growth and, more broadly, driving the growth and development of the authenticated luxury consignment market. We will do so while striking a balance with delivering strong operating leverage as we drive toward profitability. With that, we will open the line for questions. Operator?
Certainly. Ladies and gentlemen, if you have a question at this time, please press star then one on your touch-tone telephone. If your question has been answered and you'd like to remove yourself from the queue, please press the pound key. Our first question comes from the line of Scott Devitt from Stifel. Your question, please.
Hi, thanks. I have two. First, there's been a few articles recently just about the authentication process and the potential for inauthentic items to make it into inventory and be purchased. I was wondering if you could just talk about the process more broadly in terms of the percentage of items that you think actually do make it through that may be inauthentic, and in the few cases that that happens, how you actually handle that from a customer service standpoint. Secondly, it'll be interesting in your views competitively in terms of StockX, which is a business that's been focused more on the sneaker category initially, but has an authentication process as well. The company's expanded into categories that are somewhat similar to what you have and how you think of that business as a competitor or not. Thank you.
Hi, Scott. It's Julie, obviously the only woman in the room here. Look, I'm going to take both questions, and we're going to start with the StockX question. They're still a self-posting site. They really do differ quite a bit in the way we differ. They take possession of the item post-sale. More importantly, we're in a really large TAM. Our consignors are not self-posters. Our biggest competition is inertia, and that inertia is overcome. We do everything we can to encourage people to understand there is almost $300 billion of trapped value in their closet and get them to consign, both from an environmental reason and also, most importantly, it's just trapped resources in their house.
We actually don't really worry about the competition coming up, given our unique model and our moat, and we certainly welcome anyone authenticating goods in this huge TAM that's out there.
With that said, it's probably time to move into how we authenticate. This actually gives me an opportunity to toot our horn a little bit. We have a 70-point, actually, NPS score, which compares favorably with best-in-class companies like Apple or Costco. Costco's at 74, Apple's at 68. You don't get a 70 score on NPS without doing many, many things right, and we're counting authentication as one of that. Authentication is core and central to our brand. We're the only marketplace that authenticates this wide range of consumer products. I would say without hesitancy, our practices are best in class. Most importantly, they continue to evolve. They have to evolve because counterfeiters evolve. Let me just walk through how things work. We take physical possession of every single good.
The first step, once they arrive at one of our centers, our processing centers, is the receivers actually look at the item, they inspect it for condition, and to make sure it's one of our accepted brands. Receivers are trained on a base level for authentication. They use that knowledge to spot obvious fakes, poor construction, inferior material. They also separate the items considered high risk based on brand, style, value, source, things that are on trend, and they escalate them to one of our high-risk authentication team members. We have over 100 of these, and which include brand experts, GIA-certified gemologists, horologists, handbag experts, and art curators. They personally authenticate high-risk items, they build out authentication guides and materials, they lead trainings, and they work every single day to ensure that we have absolutely the best-in-class practices.
The gemologists in our LCOs, which are our luxury consignment offices, have on average seven years of industry experience, and it's important to note that most are GIA certified, and GIA certification requires six months of full-time training on a GIA campus. Other items are authenticated by our copywriters. This team, which is very large, are trained in specific categories and specific brands. Some only focus on women's contemporary fashion, for example. They receive daily, weekly, and monthly training to stay on top of the latest trends. We also audit and recheck items. We have an audit team to make sure that we're doing what we say we're doing, which is that we are the authenticated luxury consignment business. Our quality control team pulls items, does secondary and sometimes tertiary reviews to see how we're doing and uses the results to constantly retrain.
Originally, when I set up the business, my goal was to be the safest place to shop on the internet for previously owned goods. I still stand by that statement, and we certainly are. We're deeply committed to keeping counterfeit goods off the marketplace. Our business is really driven by amazing people who are dedicated to the mission of authenticity, and we're really, really good at it. Our processes are not static. They couldn't be static. We continue to invest in automation, training, and technology to stay ahead of counterfeiters.
Very helpful. Thank you.
Thank you. Our next question comes to the line of Eric Sheridan from UBS. Your question, please.
Thanks so much for taking the question. Julie, I wanted to know if we could dive in a little bit to the backstory of how the Burberry announcement came about, how we should be thinking about what that means for the platform broadly, not only just the announcement in and of itself, but whether it's a roadmap for additional partnerships with brands, how we should think about it arcing sort of the ability to drive more supply onto the platform and maybe even improve sort of some of the velocity we've seen of both shopping and conversion on the platform as you see those types of items get listed more. Thanks so much for all the color.
Sure. First of all, I personally have been talking to the brand groups for about five years. As you all may be aware of, the laws are changing in Europe and certainly in the U.K. about what they need to be in terms of sustainability. They're also hyper-conscious that producing luxury brands that are sustainable is the future. It's demanded by their customers, mostly millennials and Gen Z, which are driving the primary market. With that in mind, Burberry became the second large brand to join us in leading the charge in the circular economy. We do have a good relationship with many brands. We were particularly happy to see Burberry come along. We started talking to them about six months ago, and their goal was to launch when we had National Consignment Day, which was early. I think it's the first Tuesday in October.
We have about three weeks of results in. We are just compiling those results and sending them to Burberry today. It did drive incremental consignors for us, but I prefer to talk to them about the results before I announce it in the call. It was a really good first start, so we're excited about that. We do operate in the marketplace. I think it's good to keep in mind that we really aren't dependent on any brand for success. Having said that, we would love to see more brands come along and work with us. It's really important for the environmental impact of the luxury marketplace. We'll see. I would say that the brands on the whole are getting more enlightened, and certainly the new laws in place in Europe are going to force the issue. I expect good things in the future.
Thanks so much.
Thank you. Our next question comes from the line of Oliver Chen from Cowen and Company. Your question, please.
Hi, thank you. Regarding the environment that you're seeing now as we approach the holiday season, what are your thoughts for how the promotions look in the marketplace? Are you changing the marketing in terms of the timing of the marketing spend on a year-over-year basis? I thought I heard that. What do you think is going to happen? Because some of the inventories in the department store channel are mixed, and there are some recent bankruptcies. Would love your thoughts on how that may play out and also how you're positioned differently on a year-over-year basis as this important season comes. Thanks.
I'm going to start out, and Matt may jump in. As we mentioned before, our product mix is really diversified, and it helps us mitigate any impact of discounting on our marketplace. You certainly saw GMV in Q3 accelerate, and our AOV increased by $20 year-over-year. Our Q3 results really are a harbinger of things to come, that we can grow at healthy rates despite noise in the marketplace and discounting and competition. Again, the promotional environment isn't new to us. Over the years, we've learned to navigate in this environment. The other thing I want to say is most of our GMV comes from our high consignor repeat rate. We're well off to a good start. We do have a little seasonal impact.
We certainly don't have some of the big boom or bust periods, but we have a seasonal impact, and we're off to Q4. We don't expect the sad demise of Barneys to impact us on the grand scheme of luxury department stores. They were certainly important and trendsetters, but still relatively small. I'm going to turn it over to Matt if he has any other further comments.
Yeah. Let me just address the question around marketing timing in the fourth quarter. It's really not a change in our strategy versus where we were last quarter at the time of the IPO. It's been our plan all along coming into the year to emphasize our spend in advance of peak seasons for us. In the first quarter and the third quarter to spend in a relatively positive ROI environment and to de-emphasize our spend in the fourth quarter. That's what you see implicit in our guidance for the fourth quarter. Going forward, we're going to just react to what we're seeing in the marketplace in terms of our ROI trends and make sure that we're investing smartly and optimizing our spend as we drive toward profitability.
Okay. Lastly, regarding customer acquisition costs and buyer acquisition costs, what have you been seeing lately with the trends that you expect? It sounds like you're pretty pleased with TV as a media format. Would love your thoughts on what's been trending within that field and any thoughts on what we should pay attention to as you leverage marketing and grow revenues.
Yeah, sure. I'll start. The drivers of our marketing leverage are more or less consistent with what they were a quarter ago and going back even further than that. Top of the list is really having strong repeat rates from both buyers and consignors that make us less and less dependent progressively on acquisition to fuel our growth. Within that, having a high overlap between buyers and consignors, what we call the flywheel effect, sort of further perpetuates that trend. No doubt, our marketing investments have become increasingly efficient with our BAC going down about 15% year-over-year in the third quarter as we continue to lean into TV. I don't want to get into the specific tactical details of that for competitive reasons, but more or less, TV continues to be a very strong medium for us.
We've been emphasizing not only traditional cable and nationwide programming, but also increasingly OTT, like Hulu and Roku, that we've found pockets of success. That's what we're doing right now. Going forward, we're going to continue to stay ahead of the curve in emerging trends to continue attracting Millennial, Gen Z customers that perform very well for us. That said, we still do have very low brand awareness, so there's still lots of runway for us to amplify that aspect of the business. We'll do so while remaining disciplined about our overall level of spend and the leverage that it's driving.
Great. Congrats on Burberry. Best regards.
Thank you.
Thank you. Our next question comes from the line of Michael Binetti from Credit Suisse. Your question, please.
Hey, guys. Thanks for taking all my questions. Congrats on a nice quarter. Just one small question to think about for our model, and then I had a couple of bigger picture ones. On discounting, I think you said it was about flat year-over-year. I would have thought that would have been lower given some of the better analytics you guys have put in play to do things like remove the full site offers, like 20% off. Any view there, and any chance that the level of discounting can re-accelerate lower from here? Matt, on the take rate, I just wanted to straighten out a few comments that you made. It sounds like you expect it to be up in a year-over-year basis in the fourth quarter.
We already know the take rate is generally lower in the fourth quarter versus Q2 and Q3 every year. Since there's been some bigger and smaller increases over the last few quarters, maybe you could just help us, or maybe even just tell us how comfortable you are with consensus, which looks like it's up over 150 basis points in the fourth quarter again. I guess, Julie, just one last one, if I could sneak it in. As we look out to next year, you've got some big comparisons to anniversary now with all the progress you've made. As a percent of revenues on very important lines like marketing and ops and tech.
Maybe you could just give us some initial thoughts on the magnitude of how much progress you think you can make next year and what some of your big ideas are on initiatives for next year on those lines. Thanks.
Sure. All right, I'm going to take the last one, and then I'll let Matt go. One of the things you should understand, over time, our average selling price has been going up. We're always looking for pockets to move the ASP up, and part of it's on an absolute basis, meaning if you compare the same category or the same item versus the same item. Obviously, our AOV has gone up, and that's a combination of mix and diversity of product. On a line item, because we do a lot of analysis trying to raise the price and because we share in the risk with the consignor, we're always looking to sell at the highest possible price within a given time frame.
For next year, looking forward, we're pretty excited about next year, even though we've got to bring in this year, we want to keep focused on bringing in this year. Next year, we know we're going to have the San Francisco store opening in the first half. Whenever you get a store, it's always nice to have one where your headquarters are in the same place. It's sort of like the shoemaker's children finally get their shoes. That always has a positive impact on the marketplace. That store is exciting. As we mentioned before, we're going to have a couple stores a year, we're working on other leases. We don't have an announcement yet on the next city or cities. We're excited about San Francisco.
Automation our machine learning team is on fire, you can expect us to automate quite a bit more. The value is not just consistency in the way we describe things and saving money. It also speeds up things in terms of time to site. For example, most of our photo retouching is done offshore. That adds three days to our cycle. The more we can automate photo retouching, the shorter the period time to get an item on the site. We're really excited about that. The other thing that's been cool for me to see is the development of our Perth Amboy facility. I was just there a couple weeks ago. That's our largest 500,000 sq ft place in New Jersey.
That gives us the ability to make even more automation pushes, especially on all sides of the business in fulfillment, because there's room to work, and it requires us to think differently, and that's pushing it forward. I would say overall, next year, on one level, is going to be more in the same. On another level, it's also fine-tuning our business more, rolling out Perth Amboy, getting one store ready, another store lined up. I'll continue to knock on the door of the luxury brands, and hopefully we can actually bring in another brand partnership before the year is over. I think our goals are aligned with the luxury brands for sure, and that is actually reducing the carbon footprint. I feel really good about that. It's going to be a fun year.
Look, we still have more than six weeks left in this year, and it's been an exciting year. We've got to focus on bringing in the quarter. That's really where our focus is right now.
Let me tackle the discounting question, and I think the final one was the take rate trends. With respect to discounting, I don't think I want to add a ton more to the question other than to say we saw no unexpected trends in the quarter. Very much in line with what we anticipated as of the prior earnings call in terms of the overall environment and its impact on our business, which was negligible. With respect to, I think there was an element of your question around kind of our optimization efforts that, in theory, could impact discounting. In part, that's true, actually, a lot of our optimization on pricing happens upstream of that in terms of setting the optimal initial price. Discounting as a percent of that remains relatively consistent.
Where we've actually seen increases in our average selling price over time, and in the third quarter, it was about 4% year-over-year increase, was due to being smarter about how we price things initially. We watch things like a hawk in terms of the sell-through rate of product categories. You're right, those investments are paying off in terms of yielding higher average selling prices and ultimately average order values or AOV.
All right.
With respect to take rate, I don't want to get overly precise about take rate because as you've seen over the past couple of quarters, there's an inverse relationship between average selling prices or AOV and take rate. Take rate tends to bounce around a little bit based on what's selling, and so does AOV. That said, we've got enough visibility to know that we will see a substantial year-over-year increase in take rate. All things equal, unless the mix of product selling changes from here toward the end of the quarter, is more or less in line with our previous expectations and in line with what you were suggesting.
Okay. Thanks a lot for all the help. Congrats again, guys.
Thank you. Our next question comes from the line of Edward Yruma from KeyBanc Capital Markets. Your question, please.
Hey, guys. Thanks for taking the questions. I guess first, you guys highlighted some of the top-of-funnel benefits from the IPO. I know it's still early days, but any sense as to how these new customers are both behaving on the buyer side as well as on the consignor side? Second, as a follow-up, given how strong GMV was in the quarter, I guess, what's your comfort level that you have enough inventory to kind of support continued growth? Maybe as it relates to Perth Amboy, any kind of initial learnings from that facility? Thank you.
I'm going to take the Perth Amboy and turn it over to Matt. Here's what we had hoped when we took that space. It's a little further from New York City. We always know we get incredible talent when we pull from New York. To be honest, we were a little nervous about committing to a space that was quite a bit away from New York City. We are getting phenomenal talent, and we're pulling from other warehouses that have located from there. I don't want to give names, but other luxury online businesses that either have downsized or stopped hiring, or we offer a more competitive package. We've had nothing but positive surprises from Perth Amboy, meaning good quality workers, hard workers. I think we're up to about 300- 400 employees there. It's a beautiful facility.
It's all going as planned, a little bit better than planned. That's been a net positive surprise. We are testing out some still fairly lightweight automation there on the fulfillment side, but that's going really well, which will then roll out to our facility here in California. With that, I'm going to turn it over to Matt.
I guess the question is really around kind of the Q4 guide and how we feel relative to the supply required to deliver the GMV growth. Obviously, we feel good. It's embedded in our guidance. The good news is, you think back to Q3 and some of the strength we saw around the time of the IPO, that happens on both sides. Those top-of-funnel metrics, which of course was traffic, that led to strong member growth, which further drives strong, not only buyer, but consignor growth. We saw kind of the tide rise on both sides of the marketplace. We feel good about where we're positioned going into the quarter and as of this moment in the quarter, it's embedded in the guidance that we've provided.
Great. Thank you.
Thank you. Our next question comes from the line of Ike Boruchow from Wells Fargo. Your question please.
Hey, good afternoon. Let me add my congrats. Two questions. First question to Matt or Julie. The AOV increase that you saw in the quarter year over year, is it something you're doing tactically, or is it just the consumer opting more for these higher ticket categories that you mentioned, like watches and jewelry? Just kind of curious of the behavior there.
Yeah, I think we'll probably both chime in here. I'll start. What we saw in the third quarter, as we referred to in our prepared comments, was strength across the business. Every one of our categories saw strong growth in excess of 35% year-over-year. We saw relative strength in our higher price categories, watches, jewelry, for example. We're always focused on trying to drive sales across the spectrum of price points. I wouldn't say there was a particularly strong effort to drive high price point goods, because there's always going to be a balance between AOV and driving that up and maintaining strong customer growth and consignor growth, which are oftentimes categories of first purchase are on the lower end of the price spectrum. We need to maintain a healthy balance.
I would consider it more or less random in the quarter that we saw such a significant bump on a year-over-year basis. Certainly is in line with our objective to have a healthy mix of high price and low price products selling.
Just as a footnote to that, there was a slight impact now of having three stores where our average selling price is higher. Last year in the same quarter, we had only a little bit of the Melrose store in L.A. We actually had the Soho store, and we did not have the Madison store. This time you had a full quarter of three stores, still a small impact on our overall business, but stores do add to the AOV. The average order size is larger in the stores.
Just one follow-up. I think this one's for Matt. Looking at the margins, there's clearly a growing scale on the ad spend line as well as tech and ops, due to automation and other initiatives. You're still de-leveraging on the other SG&A line, in 3Q and your 4Q guide seems to point to improvements there as well. I guess what I want to ask is, at this point, is it too early to talk about that other SG&A line becoming an opportunity as well? Just how do we think about the costs that are in there and the ability to scale those over time?
Sure. I guess I should start by defining what's in there. The bulk of the spend in SG&A is what we consider G&A administrative headcount and related costs, which is semi-fixed. Once the initial buildup happens, which is particularly heavy this year, we'd expect it to grow significantly slower than GMV and revenue growth going forward as soon as next year. The other component, which is significant, is our sales organization, which will grow along with the business, and we're not expecting particularly significant leverage there as expressed dollars per unit that we bring in. That's obviously critically important for us to continue scaling our top line. I'm not calling for leverage in Q4, and I think I specifically call out just less de-leveraging in Q4.
As we get into next year, I think that's the appropriate time to start having the conversations about whether we think there's going to be leverage in 2020. We'll know more as we get closer to it.
Got it. Thanks, everyone.
Thank you. Our next question comes from the line of Aaron Kessler from Raymond James. Your question, please.
Thanks, guys. Just a couple of questions. Maybe just a follow-up to that question. Maybe you can talk a little bit about the product sourcing mix and maybe how that's been changing over the last few quarters, including maybe the retail stores. Additionally, if you could just talk about maybe other big milestones for automation that we should look out for over the next few quarters, or is it just more gradual changes there? Thank you.
Look, this is Julie, obviously. On the mix, the biggest change was the stores. We're still primarily in-home pickup, so it still skews heavily between 63% and 65% of all the product we get is from an in-home pickup. Stores, again, we only have three, small part of our business, but the stores are proving to be a nice source of high-value product coming in with a nice repeat rate. Still too small to call it out as a large section. The balance of it tends to come from our inside salespeople, where people just get it. We haven't seen that big of a shift yet, but I would expect there is, and the more stores we open, we'll see somewhat of a shift. With only three stores, one which has only been open since May, it's still too hard to tell.
On the automation front, really, we're going to keep doing more of what we've talked about, automating photo editing, copywriting. Those are the two big initiatives along with pricing. Some things won't get automated. Photography, every time we test an automated solution on photography, it's slower than our own people doing the work. Some photo editing, a small portion, will never get automated because it's fine jewelry and watches, which actually need a human to do it. I would say we're moving along our path of as much automation as possible. The biggest change next year is a bigger investment in automating our authentication. That also will help accelerate. Right now, it has a small portion of it's automated still. Everything's still going to have a human attached to it.
Over time, you're going to see, when you walk in, you're going to see a change in processes and more being automated and more experts being applied as the business grows. It's a cool transformation. Perth Amboy is our 500,000 sq ft warehouse, which required a different level of automation on the fulfillment than our 250,000 sq ft warehouse here in California. We're at a nascent stage of that, and as we fill up that warehouse and use the capacity, that will also be an area for innovation and automation.
Great. Thank you.
Thank you. Our next question comes on the line. I have Justin Post from Bank of America Merrill Lynch. Your question please.
Hey, this is Sean on for Justin. First question is, your returns GMV mix kind of came down sequentially, and it's at the lowest level I'm looking back in history. Want to think on what drove that and whether we can think of this sort of lower 26, 25 mix of returns as a new normal. The second question is, I just want to check in on supply acquisition. It sounds like you're ramping up headcount for your field consignor staff. Any puts and takes there? How have things gone, like the commission-only sort of consignor staff? I wanted to check on that.
Sure. First with return rate. The decrease year-over-year in return rate was primarily driven by lower order cancellations, which really had nothing to do with inherent product return rates, but just steadily and very strong execution on our fulfillment operations in the quarter compared to the same quarter a year ago. I wouldn't trend that decrease forward in terms of the return rate, but I think return rates do tend to be relatively steady but bounce around a bit from quarter-to-quarter. Second, with supply growth, I don't think we really want to add any more than we've already talked about. We're happy with the trends. The mix between channels is more or less consistent. Stores have been a nice, healthy addition. The commission-only sales program that you're referring to continues to grow, like all of our sales channels do.
All of them are doing nicely, and we're satisfied with where we sit in the midst of our most important quarter of the year.
Thanks a lot.
Thank you. This does conclude the question and answer session of today's program. I'd like to hand the program back to Julie Wainwright for any further remarks.
I just want to say thank you very much for dialing in, and I'm going to wish you all a wonderful Thanksgiving because we won't be talking, and a happy holiday period. We won't be talking till after the end of this year. Let's hope for a great 2020. I hope you all have a safe and wonderful holiday. Thank you again.
Thank you, ladies and gentlemen, for your participation in today's conference. This does