Dada Nexus Ltd (DADA) 2020 Q3 法說會逐字稿

Good morning, ladies and gentlemen, and thank you for standing by for Dada's Third Quarter 2020 Earnings Conference Call. (Operator Instructions) As a reminder, today's conference call is being recorded. I will now turn the meeting over to your host for today's call, Ms. Caroline Dong, Head of Investor Relations for Dada. Please proceed. Caroline.

Thank you, operator. Hello, everyone, and thank you for joining us today. Our third quarter 2020 earnings release was distributed earlier today and is available on our IR website at ir.imdada.cn, as well as on global newswire services. On the call today from Dada, we have Mr. Philip Kuai, Chairman and Chief Executive Officer; Mr. Jun Yang, Co-Founder and the Chief Technology Officer; and Mr. Beck Chen,Chief Financial Officer. Mr. Kuai will talk about our operations and companying highlights, followed by Mr. Chen, who will discuss the financials and the guidance. They will all be available to answer your questions during the Q&A session that follows.

Before we begin, I'd like to remind you that this conference call contains forward-looking statements as defined in the Section 21E of the Securities Exchange Act of 1934 and the U.S. Securities Litigation Reform Act of 1995. These forward-looking statements are based on to management's current expectations and current market and operating conditions; and relate to events that involve known or unknown risks, uncertainties and other factors, all of which are difficult to predict in a manner of which are beyond the company's control. These risks may cause the company's actual results or performance to differ materially.

Further information regarding these and other risks, uncertainties or factors is included in the company's filings with the U.S. SEC. The company does not undertake any obligation or to update any forward-looking statements as a result of new information, future events or otherwise, except as required under applicable law.

It is now my pleasure to introduce our Chairman and Chief Executive Officer, Mr. Kuai.

Philip, please go ahead.

Thank you, Caroline, and thank you all for joining us today. We're pleased to report another strong set of results for the third quarter of 2020. Our total net revenue grew by 85% to RMB 1.3 billion on a year-over-year basis. Our industry-leading position has been further strengthened with market share expansion and widened gap compared with the second and the following players.

According to Iresearch, JDDJ maintains as the largest local on-demand retail platform in the China supermarket segment. With 24% market share in the first 9 months of 2020, increased from 21% in 2019. And Dada Now maintains as the largest open on-demand delivery platform in China with 24% market share in the first 9 months of 2020, increased from 19% market share in 2019.

I'm very encouraged by our progress as we execute our strategy to penetrate into lower-tier cities, drive synergies between our 2 platforms and continue to deliver on our mission to bring people everything on demand.

I'll start with exciting highlights from this year's "October 20th Supermarket Shopping Festival" and "Double 11" promotions and then provide updates on the performance of our 2 platforms. And then Beck, we'll go into greater details about our financial and operating results. So let's begin. We hosted our sixth annual "October 20th Supermarket Shopping Festival" on JDDJ platform with active participation of retailer partners. The number of stores joined in this festival achieved historical records. For example, all stores of Walmart China and [1200] (corrected by company after the call) stores covering all forms of Yonghui and more than 1600 stores under the CR Vanguard's umbrella participated in the promotion. We were particularly pleased to see our efforts in lower-tier cities pay off. During the October 20th promotions, our sales volume in the lower-tier cities was 3.4x more than that of the same period last year.

In terms of our Double 11 promotions, our data now platform covers more than 2,500 cities in the counties, ensuring timely on-demand delivery for omnichannel promotions and resulted in a new peak daily order volume records of over 10 million. There are over 100,000 stores participating in Double 11 promotion on JDDJ platform, covering more than 1,200 cities and counties. GMV on JDDJ platform more than doubled our November 11 on the year-over-year basis. Both platforms JDDJ and Dada Now achieved significant breakthroughs in terms of quality of partners, sales volumes and overall performance.

I will now give you some updates on our 2 platforms, JDDJ and Dada Now. Let's start with JDDJ, China's leading local on-demand retail platform. Our fast growth during the quarter was driven by a number of factors. First, we continue to execute our growth strategy in terms of expansion in geographic coverage and category coverage. By leveraging the scale and the strength of our retail partners and competitive advantage in terms of regional supply chain and branding, we newly entered 200 cities and counties during the quarter. At the end of the quarter, we now cover over 1,200 cities and counties, bringing more and more people everything on demand.

In this Q3, GMVs from lower-tier cities increased by over 170% year-over-year. We are also expanding our category coverage for consumer electronics and smartphones we have established partnership with dozens of authorized retailers of Apple and have large exclusive online stores on JDDJ platform. The first iPhone 12 order on launching day was delivered to consumer within only 14 minutes. We will continue to digitize more smartphone brands, retail stores and home appliance stores in the very near future.

In terms of pharmaceuticals, JDDJ established partnership with more than 70 out of the top 100 drug store chains in China, including all of the top 10 drugs store chains. For the fast-growing pet category, JDDJ covers a variety of subcategories, including food, toys, health care and body care. We also established partnerships with multiple well-known pet brands during the quarter as well.

Second, we are enhancing our collaboration with retailers and brand owners. In terms of retailer partners, during the quarter, we established new partnerships with more than 20 leading supermarket chains, including those of the top 100 supermarket chains and original dominant players such as Ouya in Jilin province and Zhenhua in Shandong province and Zhebei in Zhejiang province. There are all local winners. JDDJ had worked with over 2/3 of the top 100 supermarket chains in China by the end of September, which is a further demonstration of our advantages as a pure platform. We only enable and empower our retailer partners and create value to them as a partner rather than competing with them. The overwhelming number of leading retailers on our JDDJ platform compared with other competitors also presents recognition from retailers.

During the quarter, we also extended our strategic partnership with CR Vanguard in relation to omnichannel fulfillment product assortment management, consumer insights and marketing. During this quarter, our online marketing service revenue from brands continued to increase significantly by over 400% year-over-year. We extended strategic partnership agreements with leading brand partners, including Pepsico, Nestlé, Mars Wrigley, Yili and Mengniu and so on. We also continue to innovate marketing activities to create more values to brand partners. For example, our pioneered 1-hour delivery for live streaming program has been very well received by more brands during this quarter. In addition, we partnered with brands to create innovative marketing activities to improve sales and marketing efficiencies. For example, "Brand CP Day" is a joint marketing promotion with brands. During the Brand CP Day of Yili and Yihai Karry, the sales doubled on the week-over-week basis. And third, we are constantly developing innovative technologies to empower retailers and brand partners. As of October 31, 2020, Dada Haibo Omnichannel online retail operating system was adopted in over 1,500 large and medium-sized supermarket chain stores in China. We added new modules such as Smart Picking and Pack to optimize the order picking process, new category expansion assistant and an open developer ecosystem module, which have been adopted by all merchants.

Overall, our Haibo system enables merchants to efficiently integrate their off-line systems and supports omnichannel execution while ensuring operation stabilities. Meanwhile, our CRM systems have been adopted by more than of 35,000 stores compared to 30,000 stores as of end of Q2. In order to continue to empower our retailer partners and improve consumer satisfaction, we recently launched Dada Cross Sourcing picking service, also known as "Dada Jianhuo" in chinese. In addition to providing systematically recruiting, managing and training services, Dada picking also delivers efficient tracking of picking quality and personnel working status through the digital management of the entire process. We believe this will significantly improve the efficiencies of picking and fulfillment process while reducing store personnel costs. We also deepened our collaboration with JD Retail on "Wujingtianze", the key omnichannel collaboration program. During the third quarter, the GMV of Wujingtianze increased by over 470% on a quarter-over-quarter basis.

Let's now move on to Dada Now, the largest open on-demand delivery platform in China. First, I will touch on our chain merchants business. We extended our collaboration with industry-leading catering supermarkets, pharmaceutical and other chain brands and upgraded our dedicated delivery services. The customized delivery service, also know as "Quanxinda".

Dedicated delivery service serves the leading chain restaurants, pharmaceutical and fresh food chain retailers. And revenue from our key account merchants increased by more than 200% year-over-year during the third quarter. And turning to the last-mile delivery, we continue to benefit from enhanced synergies with JD Logistics as we further integrate our systems and delivery efficiencies. In addition to collaborating on delivery services with JD Logistics, we also expanded our partnership by launching pickup service. The premium writers who have passed our training and exams will be qualified to provide timely pickup service for the request on JD Logistics platforms. This value-add service will not only address challenges of frequent fluctuation of pickup demand based on the crowdsourcing model, but also significantly improved the user satisfaction based on the high-quality services.

So overall, we're very pleased with our progress during the quarter, and we think we have an exciting journey ahead as we execute our growth strategy. With that, I will now pass the call over to Beck Chen to go over our financials for the quarter. Thank you.

Thank you, Philip. We are pleased to deliver another quarter of strong revenue growth with significant improvement of operating margin in the third quarter as we continue to improve operating efficiency. And take advantage of economies of scale across our business. Before we go over the numbers, just a few housekeeping items in advance. We believe year-over-year comparisons are the most useful way to judge our performance. All percentage changes I'm going to give will be on a year-over-year basis and all figures in RMB unless otherwise noted.

Total net revenues increased by 85% to RMB 1.3 billion. Net revenues from Dada Now increased by 81% to 719 million, mainly driven by the increases in order volume of last-mile delivery service to logistics companies and intercity delivery service to chain merchants.

Total net revenues from JDDJ increased by 91% to 583 million, mainly due to over [91%] (corrected by company after the call) increase in GMV from the same quarter last year, which was driven by increases in average order value and the number of active consumers. The number of active consumers for the past 12 months ended this Q3 increased by 77% to over 37 million. Online marketing services revenues from brands continued to increase significantly by over 400% year-over-year as a result of increasing promotion activities from brand owners.

Moving over to the expenses side. Operations and support expenses increased to 1 billion, mainly due to an increase in rider costs as a result of increasing order volume for last-mile and intracity delivery services that we provide to logistics companies and various chain merchants on Dada Now platform and retailers on the JDDJ platform. Selling and marketing expenses rose to about $500 million, mainly due to an increase in incentives to JDDJ consumers on higher GMV, while the rate of incentives as a percentage of GMV declined second, an increase in advertising and marketing expenses, which was primarily attributable to increase in referral fees that we paid to staff at retail stores for their efforts to attract new consumers to the JDDJ platform. And the third an increase in personnel costs in connection with our growing business and increased share-based compensation expenses. G&A expenses rose to RMB 120 million, mainly due to increased share-based compensation expenses and also because of increases in professional services fees as we are a listed company now. R&D expenses rose to RMB 100 million, mainly because of an increase in R&D as we continue to strengthen our technological capabilities and increased share-based compensation expenses. The non-GAAP loss from operations narrowed by 22% to RMB 339 million. Non-GAAP operating margin was minus 26%, which was a great improvement from minus 62% in same quarter of last year. The improvement in our operating efficiency mainly comes from 2 areas. First, we were able to improve our delivery efficiencies because of an increase in order density and optimization of our Automatic Order Pricing Mechanism algorithms. Ratio of operations and support expenses as percentage of our total net revenue decreased to 78% this quarter from 94% in the third quarter last year. Operations and support expenses mainly include fulfillment expenses.

Second, the ratio of adjusted selling and marketing expenses as a percentage of total net revenue decreased significantly to 37% this quarter from 53% in Q3 last year. This was mainly thanks to the saving of consumer incentives because our platform becomes more attractive and the consumers become more loyal. In Q3, our non-GAAP net loss narrowed by 21% to RMB 324 million. Non-GAAP net margin was minus 25%, which was an improvement from minus 60% in the same quarter of last year.

The non-GAAP net loss attributable to ordinary shareholders was RMB 324 million versus RMB 621 million in Q3 last year. Non-GAAP diluted net loss per share was $0.36 compared with [$1.71] (corrected by company after the call) for the third quarter of last year. As of September 30, 2020, we had the RMB 3.7 billion cash and cash equivalents, restricted cash and short-term investments.

In terms of the outlook for the fourth quarter of 2020, we expect total net revenue to be between RMB 2 billion and RMB 2.1 billion. With a strong growth momentum, and we are confident that JDDJ revenue will grow by 100% year-over-year in the second half of 2020.

So this concludes our prepared remarks. And operator, we are now ready to begin the Q&A session. Thank you.

(Operator Instructions)

And our first question comes from the line of Ronald Keung from Goldman Sachs.

Maybe I have 2 questions this time, first natural selection, Wujingtianze, and the second one on your lower-tier city strategy. So for natural selection, I think, Philip, you mentioned about 400-plus Q-on-Q increase in GMV. So I just wanted to know how the cohorts，ticket sizes or, say, subsidy levels for these new orders from Wujingtianze? And how do we see that kind of fourth quarter and expect guided an acceleration in JDDJ. How much would that be coming from this Wujingtianze or natural selection? And longer term, how much do we see kind of our sales that will be driven from this Phase II channel, which as the users going in directly from the JD main app searching and having this natural selection. So for a bit of long question. Second would be just a lower-tier strategy, lower-tier city strategy. Do we see -- I know we're starting to kind of group orders for deliveries, so to keep that density and just how are we doing there? And do we see kind of self pickup from supermarkets or a possibility as well, given the different kind of self pickup options for consumers now in lower tier cities.

Thank you, Ronald. So I'll start and perhaps Beck something to add. So first of all, with the natural selection, we're quite excited to see this Wujingtianze collaboration with JD has been doing pretty well. So internally, we have streamlined our process and our technology backbones together with JD. So it's much more efficient and streamlined compared to a couple of quarters ago, and it's a much easier integration with the retailer partners. So at the same time, we're also happy to see that our category covered on the Wujingtianze has now expanded from supermarket that we originally start with and now to more and more categories, including but not limited to consumer electronics, smartphones, health care, cosmetics and so on. So it's a much broader coverage on the categories as of this quarter. And we anticipate this expansion on category will expand as well. And third, the product of Wujingtianze has been improved quite significantly over the quarter as well. If you have notice on the JD app, you have seen that on the search results page, there is a 1-hour delivery tab already and the product innovations around this tab and also the improvements for the entire search result page has been quite significant over the last quarter. At the same time, we're also very happy to see that -- I think just released yesterday on JD app. You can now see -- recently, I'm not sure if it was yesterday, but very recnetly on the JD app homepage, there's an entry point called local merchants, which is also an extension of this natural selection. So overall, we are pleased to see that the improvement of this Wujingtianze natural selection has been pretty well. So that is the result -- that is the reason why we can achieve 470% of GMV growth in Q3 on Wujingtianze. And we believe that the growth of Wujingtianze will continue to be strong. But still, it's still at a very early stage, as we have now just entered into this collaboration for few quarters. So it's still very new. So first, I think we will continue to improve and to grow the business.

Let's see, If Beck have anything to add?

Yes. For the first question, for Wujingtianze, just want to add more color on Philip's comments. So first, for the subsidy, our consumer incentive levels is generally the subsidy level is less than the level on JDDJ's own platform because generally, we believe for JD, it's really a large traffic source. So we don't need to subside a lot for consumers to like forming the shopping habits, for the one-hour delivery. And in the long term, we believe it could be even lower cost. It's just like the step-up stage. So we need to like to expose more to the consumers. So there will be some subsidies. But in the longer term, we don't think the subsidy is necessary.

And for the cohort or like for the like for the outlook of Q4, at least, as we have discussed in the previous calls that Wujingtianze is still in the early stage, and we expect to ramp up like quarter-by-quarter for Q4, we still believe that Q4 GMV could be still growing by multiple times compared to Q3. So it's still like a very fast-growing stage. And have seen early stage. Yes. So this is for the first question.

Yes. And in terms of the lower-tier cities, we have been growing pretty quickly in perhaps all of the lower-tier cities we have entered. And we being doing a few things in lower-tier cities. First of all, we continue to expand our geographic expansion coverage. And the second, so we work with more and more local winners. Those are the dominant players in each city or each province. And we are particularly happy to see this collaboration with the local government players have been pretty good. The progress has been pretty good in the last quarter as well. So at the same time, we have been experimenting various models in lower-tier cities. And -- But in general, we are seeing that the buying power from lower-tier cities is as strong as the Tier 1 cities. We're not seeing them, for example, having a much lower ticket size or much lower willingness to pay, we're not seeing that. We're seeing quite strong buying power from the consumer in lower Tier cities.

At the same time, we are experimenting, as we mentioned, like a self pickup and other bundled delivery to lower the delivery costs and so on. All those tests, we have been testing various cities. And I think it's still at the early stage, but we are very excited to see the growth. And yes, If Beck has anything to that.

Yes. Nothing to add from my side. Yes.

And your next question comes from the line of Eddie Leung from Bank of America Merrill Lynch.

Just wondering if you could give us an update of your thought on the various business models, handling these fast-growing grocery fresh food deals category. For example, we have seen more players getting into the so-called community group by business model. And we have also seen other business models from deals are growing pretty rapidly versus on demand. Obviously, you guys are going at triple digit. So wondering any thoughts on expansion of business models or continuously focusing on our instant on-demand model? And why? And then secondly, just a housekeeping question. I remember, Beck mentioned that the average order size, I think, improved EBITDA in the first quarter. So just wondering, on JDDJ, what's the average order size in the third quarter? And how should we think about the trend going forward?

I'll have Beck to handle the question about the second one, average SKU size, and I will share some thoughts on the first one. In terms of the business model and the community group buying. So first of all, I would like to start with a brief comparison between the community group buying model, with our model. So first of all, if you compare our average SKU size, which is over RMB 130 in the supermarket category. With around like RMB 10 of community group buy average vaccine size. So this is a very significant difference. The few reasons behind it, why there's a huge difference.

First of all, the community grew buying, they typically just rely on a few so-called hot Baokuan, hot SKUs. And so it's not a broad selection of SKUs. So therefore, the consumers just cannot buy much. At the same time, if the community group buying, would like to increase their basket size. They will have to very significantly expand their supply chain. And therefore, the supply chain and auto fulfillment costs will grow like exponentially. So this is a very key difference. At the same time, we enjoy the very broad product selection from the 2 markets we partner with already. Each of them has over like tens of thousands of very competitive SKUs. So -- and therefore, we can offer a much broader selection to the consumers. And we saw were much higher basket size, the profitability potential is also much higher. So this is #1 in terms of the SKU size and the logic.

And the second, we noticed that the product quality and also the customer satisfaction from community group buying has been pretty inconsistent. And it's also very difficult actually for them to achieve the quality and customer experience. But at the same time, because we work with the premium leading supermarkets, so the product quality and customer satisfactions are ensured. And also the other fulfillment can be also guaranteed. I think this is another key difference between our model and the community group buying.

And as a result, so the lower-tier city growth on JDDJ, in Q3, we grew by 170% year-over-year and take Hunan province, for example, because Hunan is one of the most competitive province that community group buying has been -- like all of them are there and compete very brutal. Even our growth in Hunan province maintained a 100% year-on-year and 40% quarter-over-quarter.

So as a result, we can continue to keep a very strong growth momentum in even the most competitive area. At the same time, we also enjoy some of the benefits from the competition and the pressure from the community group buying, for example, especially in the lower-tier cities, the leading supermarkets are now more and more willing or eager to work with us. And that's why recently, we have signed up with a few dozens of top 100 or the leading -- local dominant of supermarkets recently. So I think this is one of the reasons why we can sign up with so many dominant players so quickly because of this pressure from community group buying and at the same time, we're also happy to see that our Haibo omnichannel operating system has been very quickly adopted by those local dominant players. And it cuts off the pressure and because of the eagerness to work with us.

And in terms of the business model, we believe that our business model has very significant competitive edge and there is still a lot of room to improve. For example, the digitization of the entire process and also to empower our retail partners with our technology, including, but not limited to Haibo system. I think the growth potential there and room for efficiency improvement is very significant. At the same time, picking and packing is very important for the entire process, right? And that's why last quarter, we introduced the Dada Picking System for our partners. And we're very happy to see that leading retailers, including Walmart and Yonghui and CR Vanguard, all of them has very quickly adopted our data picking services in order to improve their picking efficiencies, lower their picking costs and improve the customer satisfaction. I think, overall, there are a lot more to improve around our current business model. And we're very confident with that. At the same time, some of the learnings from the community group buying, we also take advantage of that. For example, those group buying has been very focusing on leveraging the community, leveraging the Wechat group ecosystem and also to experimenting the self pickup model in order to lower the delivery costs and so on. So those are the things we have been learning from them, and we will apply in our future growth strategy as well.

Okay. For the second question about average order value. So in Q3, our overall average order value on the platform is RMB 150, which is a further increase compared to Q2 because the consumer electronics revenues GMV is growing very fast. I would like to emphasize that for the supermarket category, just like Philip discussed this minutes ago, fulfill and AOV is around RMB 130, and we expect RMB 130, very healthy and will be stable going forward. So -- Yes, so this is AOV information.

And our next question comes from the line of Thomas Chong from Jefferies.

My question is about the monetization trend on JDDJ. Can you comment about the commission advertising and other monetization models in the upcoming quarters. As well as how we should think about the take rate in 2021? And on that front, how should we think about the cost side, in particular, the subsidies ratio that we are seeing it continue to trend down on a year-on-year basis. So given the business trend, should we expect our profitability timing to remain unchanged in next year?

Okay. Thank you, Thomas. I will take the question. Yes. So for the monetization ratio, basically, we had multiple revenue streams from merchants, consumers and the brands. So first of all, we believe that for the online marketing revenues, and it will be continuing to grow year-over-year basis. So for year 2020, the monetization rate from online marketing is growing very much, almost like doubled year-over-year and more than double the year-over-year. But for the next year, we believe that we still just want to keep it stably growing instead of like doubling every year. So we believe it will be healthy for the platform's long-term growth. And for the delivery services revenues, we believe that for this year, especially for the Q2 and Q3, we have some incentives to encourage the consumers purchase. But like we said in the previous earnings call last quarter, starting from Q3, late Q3 stage, we are already to increase, it's increasing a little bit about the freight expenses -- freight revenues. So we believe that in the next several years, our freight expenses will be increased compared to year 2020. So we may have some consumer electronics category is growing fast. So it may have some drag factor to the commission rate. But because every year, usually, we will increase a little bit for the merchant standardized take rate. So we believe that it will be offset. So the commission rate will be a little bit like flat for the next 2 years. So overall, with the increase of delivery service and online marketing revenues, we believe that the flat of the commission rate because of the contribution from the consumer electronics category. So we believe the JDDJ overall monetization rate will continue to grow like by 30 bps each year year-over-year for the next 2 years. And for the direct margin level, we are still on the track to -- like to break even for the JDDJ direct margin level in 2021 next year. So on a year-over-year basis, like for the second half of this year on year-over-year basis, our consumer incentives is still decreased, compared to Q3 and Q4 last year. So for the overall profitability of the company, we believe that it will be a year later after the JDDJ direct margin breakeven next year.

And your next question comes from the line of Billy Leung from [Haitong International].

I just have one, I guess, sort of high-level question. As we mentioned in our prepared remarks that we gained market share for both JDDJ and Dada Now. Could I just get a -- can you help me understand where the market share gain has come from? Is it from are these local smaller players? Or is it from relatively larger plays? I just want to understand where the market share gain is from? That's my question.

Okay. So the market share we are gaining as per the statistic is mainly coming from smaller players and also like the players within Alibaba ecosystem. Hello, Billy?

Okay. Okay. Great. And just a follow-up. I think it's just a house keeping question is for JDDJ, correct if I'm wrong, but I realized that our take rate was just slightly down from quarter 2 to slightly. So I was wondering if there's any reason for that? I mean, is it because the advertising has been growing faster or whatever reason, it's just slightly down, I think it's 9.2% from 9.3%. Just any color on that would be great.

Yes. So it's mainly because like online marketing revenues as a percentage of GMV is like 2.5% in Q2, while it was 2.2% in Q3. The online market revenue is still growing very fast. And even on a Q-on-Q basis is still growing. But for Q2, there is like a lot of pent-up demand from the brand, they just want to spend huge amount of money in the June 18 next year campaign. While in Q3, there's no such kind of like P1 campaign -- P0 campaign, a big campaign in the quarter. So that's why, overall, the spending is bigger in Q3, but there's no big campaign Q3. So that's why as a percentage of the GMV, it seems that the brand online marketing percentage decrease. But just like we said or discussed in the previous calls that we believe that the absolute dollar amount will be still flat and a little bit increased as a percentage of revenues because overall, JDDJ revenue has increased very fast, and GMV is increasing fast. So the percentage is not necessarily going up so quickly every quarter. But still looking at like on a year-over-year basis, for the last year, last year, it's the monetization rate of online marketing is below 1% in Q3. For this quarter, it's like over 2.2% in Q3.

(Operator Instructions)

And your next question comes from the line of Alicia Yap from Citigroup.

I have 2 questions. One is the GMV for JDDJ was really strong this quarter. And given these record single days, any color we could expect in terms of the GMV growth into the fourth quarter? And also, category mix within the GMV, if you can share some color on what we have achieved in the third quarter. And how would you envision GMV mix into next year, given we are increasingly getting more and more of the electronics and the pharmaceutical category. So any color on the GMV mix into next year will be helpful.

And then just another question is I'm not sure whether I'm thinking it correctly. But just thinking about more the conflicts of interest in terms of the whole industry dynamic, right, between the on-demand delivery versus the increasing availability of the self pickup model. So would some of these self pickup model that could potentially cannibalize the on-demand delivery. So any color from management view in terms of how this industry dynamic could play out and how it would impact us? That would be great.

Okay. Thank you, Alicia. Let me take the first question. So for the Q4 and for next year, so first of all, for Q3 the consumer electronics growing fast, and it contributes to almost 10% of the total GMV amount and also helped to lift up the overall average order value of the platform as well. So going to the Q4 and next year, we believe that it's reasonable to estimate that the consumer electronics or other like new categories we will introduce will continue to contribute 10% of the total GMV. And in terms of the -- for example, the GMV -- the revenues of JDDJ platform in Q4, we believe that compared to Q3, the growth rate of the revenue will be accelerated and it should be able to growing by over 100%.

And for the second question on the comparison. Yes, I pass to Philip.

Yes. Right. So the self pickup model, I'm sure you're mostly referring to the Community group buying. So self pickup and delivery, I think there will be 2 major different models going forward. I think each one of them will continue to exist. But I believe the delivery model will continue to be the demand model because of the much better customer satisfaction, and this has been also demonstrated in the food delivery business already, right. So food delivery, restaurant food delivery also have the self pickup model in restaurants. But as you have already seen, the delivery model is the dominant. At the same time, we believe the self pick up is an added service for the stores. And therefore, we're also providing -- in fact, we also provide the self pickup this option for the stores and for the consumers as well. And we may experiment more models around this self pickup. But I think in the foreseeable future, our delivery model will continue to be the mainstream.

And your next question comes from the line of Wei Fang from Morgan Stanley.

I have 2 questions. First, could you please kindly share with us some information about the stickiness of the customer. For example, we have around like 21 million active users at the end of last year, September. And for this 21 million users, how much have they purchased on average during the past maybe 12 months or past quarter, are they like higher than the new users were lower than new users? This is my first question.

Okay. So for stickiness, I don't right now have the exact question for -- the exact answer for the question, but just like a little bit like the color to share with you is like the existing customer in 2017, year 2017, there is like over 77%, if my memory is correct, 77% is still purchase place order in 2020, the first 9 months. So this is like stickiness for some cohort analysis of the consumers' attention.

Yes. Just one quick thing to add on that. So we introduced the membership program for the retailers and help the retailers like Walmart, Yonghui, Vanguard and many others to establish the membership programs on the JDDJ, which also help to increase the stickiness and also the frequencies of those members on each retailers.

Okay. And maybe a brief follow-up on this. I thought I had read some numbers saying that the old users or maybe 2 or 3 years old users, and on average, they purchased something like slightly above RMB 1,000 per year. That number seems like if your AOV is at CNY 130, it's basically just 10 orders per year, which is 1 order per month, basically. So I think that it's not a very high-frequency behavior. So is that number true? Is that number -- I read the CNY 1,000 per year purchase is correct?

Okay. So first of all, so for our GMV purchase, I think right now, it should be -- has increased to like RMB 1,300 for the first 9 months of this year, and generally for new customers because we still like acquiring new customers online and offline. So they may be purchased only one time a month but for existing or like those repeating purchase usually, they purchase like 53x in 1 month, which is like more than 1 week or 2 weeks time.

Okay. And maybe my second question is a follow-up on the Wujingtianze. I just wondering if, for example, I purchased something from the JD platform, and I used the 1-hour delivery -- the 1-hour delivery tab in JD results -- search results and purchase from it. Are those goods sold -- the commissions are charged by the JD platform or it's charged by us? It's -- I don't know whether it's charged by JD or if Dada provides a delivery service, will also charge the commission?

Okay. JDDJ is a platform to integrate with like tens -- like hundreds of thousands of stores offline. So it's like we are still -- like the platform hosting the businesses. And we will charge the commission revenues to the merchant listed in the 1-hour tab on JD.com. So this is the revenue stream. And in the same time, we may pay like some nominal tax expenses to JD.com for the order is much less than the commission rate. So -- yes, this is the flow.

Okay. So that means if the merchant is simultaneously listed in the JD search result and also a merchant on our JDDJ platform, then when the customers search the results on JD and order will be guided to the JDDJ platform, and we will charge the commission?

So actually -- so it's like a little bit like a mini program on JD.com, JD app. And if you -- you're careful that you can notice that and the store and also the services that we provide, the store is hosted by JDDJ as listed on the page. And also the services or delivery services is delivered -- it's fulfilled by Dada Now. So this is like the flow of the businesses.

And your next question comes from the line of [Robin Liu (sic) [Robin Leung] from Daiwa.

This quarter, the user growth momentum is still very strong. Can management share more on the online traffic source. For example, as we are expanding into the electronics category, are we seeing more users from JD? And in 2021, can we maintain the similar growth level next year? Or should we even see further acceleration when we penetrate more into the lower-tier cities as I believe Wujingtianze is still mainly launched in higher-tier cities at this moment.

Yes. So the Wujingtianze certainly helped to increase the online traffic from JD, and as we expand to consumer electronics, smartphones, which are the dominant categories for JD in -- so most people will buy cell phones on JD. So we certainly benefit from that. And also, as you mentioned, we certainly benefited from entering into more lower-tier cities. For example, if you notice on the JD app homepage, there's an entry point called JDDJ on the top right side. And every city we enter, we will then open this entry point on JD because it's location based. So more cities we enter and the more traffic we will be getting from JD. So this is just JD. And certainly, we are expanding our traffic source to other sources and both from online and offline.

(Operator Instructions)

And your next question comes from the line of Hans Chung from KeyBanc Capital.

Congrats on the strong results. I just have one high-level question. It's about competitive landscape in the longer term. If we look at the e-commerce, right, it's pretty much the 3 race horses, right? And if we look at the food deliveries, only 2 players with 1 dominating. And so now if we look at the so-called offline to online, the supermarket retail as a whole, including the self pickup or the delivery. This is still in early stage, but it's also become crowded, especially we have seen big guys entering the market. So what's your view on the landscape in the long term? Will it be winner taking majority of share? Or perhaps the market will evolve into different types of the market segments? And each segment may have a one winner or there could be another different scenario?

So our vision from day 1, we choose to deliver everything on-demand. And we start off with the supermarkets, which is the single largest category for the offline retail. And at the same time, it's also the most difficult one. So we start with supermarket and then expand from one category to another. I think we are seeing that the competition in this O2O, so-called the instant retail or on-demand retail has this competition. I think there are 2 major models in this market. One model is basically represented by us. We are probably the only major players in this sector. We work with -- we support and we never compete with our partners. So therefore, we are strongly supported and backed by JD, by Walmart, by Yonghui, by CR Vanguard by all of the leading retailers because we never compete with them. We only empower them and bring new revenue to them.

At the same time, we're also seeing that other players like Alibaba or Meituan, they have retailers themselves. Right, they are more focusing on the first-party business, and they're competing with the retailers head on head. And with that, we believe that the platform model, and as a matter of fact, we are collaborating instead of competing with all the leading retailers will eventually win. And one of the key reasons behind it is that the China's supply chain is very fragmented. If you look into the entire landscape, even for the strongest retailer in China, grocery sector or supermarket sector, even the strongest, they can only be the winner, the dominant player in a few cities or a few province. Nobody, literally, nobody in the last few decades can ever choose to become the #1 in all province. Therefore, we believe that the best strategy is to really deeply working -- collaborating with the dominant players in each province and each city instead of being retailer ourselves. And we believe that this strategy will eventually win the competition.

There are no further questions at this time. I'd now like to hand the call or the mic over to concerned. Please go ahead.

Thank you, operator. In closing, on behalf of Dada's management team, we'd like to thank you for your participation in today's call. If you require any further information, feel free to reach out to us directly. Thank you for attendance today. This concludes the call.

And that does conclude our conference for today. Thank you for participating. You may all now disconnect.