Ladies and gentlemen, thank you for standing by. Welcome to the second quarter fiscal year 2021 TAL Education Group earnings conference call. At this time, all participants are in a listen-only mode. After the speaker 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. I would now like to hand the call over to your first speaker today, Ms. Echo Yan, IR Director of TAL. Thank you. Please go ahead.
Thanks, operator. Thank you all for joining us today for TAL Education Group's second fiscal quarter 2021 earnings conference call. The earnings release was distributed earlier today. You may find a copy on the company's IR website or through the newswires. During this call, you will hear from Chief Financial Officer, Mr. Rong Luo, Linda Huo, Vice President of Finance, and myself, IR of TAL. Following the prepared remarks, Mr. Luo and Ms. Huo will be available to answer your questions. Before we continue, please note that the discussions today will contain forward-looking statements made under the Safe Harbor provisions of the US Private Securities Litigation Reform Act of 1995. Forward-looking statements are subject to risks and uncertainties that may cause actual results to differ materially from our current expectations. Potential risks and uncertainties include, but are not limited to, those outlined in the public filings with the SEC.
For more information about these risks and uncertainties, please refer to our filings with the SEC. Also, our earnings release in this call includes discussions of certain non-GAAP financial measures. Please refer to our earnings release, which contains the reconciliation of the non-GAAP measures to the most directly comparable GAAP measures. I would like now to turn the call over to Mr. Rong Luo. Rong, please.
Thank you, Echo. Good evening and good morning to you all. Thank you for joining us today on this earnings call. We continue to see a positive recovery in China's public health situation and economy in the second fiscal quarter. In the education sector, schools were able to reopen and start the school year on September 1st, following the gradual resumption of the offline teaching and tutoring during the summer months. At TAL, during the second fiscal quarter, we continued to execute on our online and offline strategy, which remains on track. More and more of our offline learning centers could reopen, and currently, all of our learning centers in China are back in business at a certain level. The second fiscal quarter financial results reflect the extended impact of the earlier COVID-19 outbreak.
Net revenue growth in the second quarter was 20.8% year-over-year in U.S. dollar terms to $1,103.3 million and 21.6% in RMB terms. Total normal price long-term courses student enrollments increased by 65% year-over-year, mostly driven by online enrollments as well as Xueersi Peiyou Small Class. GAAP loss from operations was $49.1 million compared to income from operations of $60.8 million in the second quarter last fiscal year. Non-GAAP loss from operations was $11.8 million compared to non-GAAP income from operations of $89.7 million in the same period of the prior years. I will now turn the call over to Linda Huo, our Vice President of Finance. She will give you an update on our operational progress in the second fiscal quarter. Next, Echo Yan, our IR Director, will reveal the second quarter financials.
After that, I will update you on our business outlook and on our business strategy and discuss our business outlook. Linda, please.
Thanks, Rong. I will reveal the various revenue streams of our tutoring business for the second quarter. Let me start with Small Class and Other Business, which consists of Xueersi Peiyou Small Class, Firstleap, Mobby, and some other education programs and services. These accounted for 67% of total net revenue, compared to 75% in the second quarter last fiscal year. The revenue growth rate was 8% in both USD and RMB terms. Xueersi Peiyou Small Class, which remains our stable core business, represented 57% of total net revenue in the second quarter, compared to 65% in the same year-ago period. The lower revenue contribution from Xueersi Peiyou was mostly due to the faster growth of xueersi.com online courses, which accounted for 26% of total revenue in the quarter, compared to 17% in the same period last year.
Net revenue from Xueersi Peiyou Small Class was up by 7% in both U.S. dollar and CNY terms, while our normal price long-term course enrollments increased by 31% year-over-year. In the second quarter, normal price long-term Xueersi Peiyou Small Class ASP decreased by 19% in U.S. dollar terms and 18% in Renminbi terms year-over-year. The decline was mainly due to the mix change of Peiyou Online and offline business proportions, as well as regular promotions. Our second quarter performance in the various tiers of cities reflects the relative larger impact from COVID-19 on the top-tier cities versus the lower-tier cities in our geographic network. Xueersi Peiyou Small Class revenue from the top five cities, which are Beijing, Shanghai, Guangzhou, Shenzhen, and Nanjing, increased by 3% year-over-year in U.S. dollar terms and accounted for 52% of Xueersi Peiyou Small Class business.
Revenue generated from cities other than the top five grew by 11% in US Dollar terms. The other cities accounted for 48% of the Xueersi Peiyou Small Class business. Next, I'd like to discuss our Zhikang One-on-One business. This business sector achieved year-over-year revenue growth of 6% in both US Dollar terms and Renminbi terms. Zhikang One-on-One accounted for approximately 7% of total revenue in the second quarter of fiscal year 2021, compared to 8% in the same year ago period. In this quarter, normal price long-term Zhikang One-on-One course ASP increased by 4% in both Renminbi and US Dollar terms year-over-year. Now, let me update you on our current capacity expansion strategy. Following our entry into 20 new cities in the first quarter, we added one more city, Xianyang, in the second quarter.
Year to date, we are well on track with the planned offline capacity expansion phases. In line with our long-standing approach, we will continue to pursue healthy and sustainable learning center network expansion by following government guidelines and market demands. In Q2, on a net basis, the number of learning centers was unchanged from Q1, at 936 in total. We opened 22 new Peiyou small class learning centers and closed 19 Peiyou small class learning centers. We also closed a net of three Zhikang One-on-One learning centers. During the quarter, we added 25 Peiyou small class classrooms. In all, by the end of August 2020, we had 936 learning centers in 91 cities, of which 90 cities in China and one Xueersi Peiyou learning center in the U.S. Among the total 936 learning centers, 716 were Peiyou small class and international education centers.
91 were the merged Firstleap and Mobby small class, and 129 were Zhikang One-on-One. As for Q3 of fiscal year 2021 until now, we have conditionally rented 14 Peiyou small class learning centers. As always, we expect to add a few more and close down some learning centers based on standard operations. We will continue to closely monitor the developments with regards to COVID-19. These estimates reflect our current expectations, which is subject to change. Turning now to our online business. Second quarter revenue from xueersi.com grew by 87% in US D ollar terms year-over-year and 88% in Renminbi terms, where normal price long-term course enrollments grew by 116% year-over-year to over $2.9 million.
Online contributed 26% of total revenue and 52% of total normal price long-term enrollments this quarter, compared to 17% of total revenue and 40% of total normal price long-term course enrollments in the same year-ago period, respectively. The growth in online business was supported by the structural growth trend in online education as well as sales and marketing efforts and retention of the previous quarters. In Q2, normal price long-term online course ASP decreased by 1% in both US dollar and Renminbi terms year-over-year. With that, I will now turn the call over to Echo Yan for the update on second fiscal quarter financial results. Echo, please.
Thanks, Linda. Let me now go through some key financial points for the second quarter of fiscal year 2021. Gross profit increased by 14.3% to $581.2 million from $508.7 million in the same year-ago period. Gross margin for the second quarter decreased to 52.7%, as compared to 55.7% for the same period of last year. Selling and marketing expenses increased by 44.3% to $379.8 million from $263.3 million in the second quarter of fiscal year 2020. Non-GAAP selling and marketing expenses, which included share-based compensation expenses, increased by 43% to $370.3 million from $258.9 million in the same year-ago period. The year-on-year increase of selling and marketing expenses in the second quarter of fiscal year 2021 was primarily a result of more marketing promotion activities to strengthen our customer base and brand, as well as higher compensation to sales and marketing staff to support more programs and service offerings.
Other income was $45.3 million for the second quarter of fiscal year 2021, was primarily due to the value-added tax and the Social Security expenses exemption offered by the government during the COVID-19 outbreak. Other expenses was $55.6 million in the second quarter of fiscal year 2020, mainly related to loss from the fair value change of an equity security with readily determinable fair value. Impairment loss on long-term investments was $4.6 million for the second quarter of fiscal year 2021, compared to $54.2 million for the second quarter of fiscal year 2020. Impairment loss on long-term investments was mainly due to a decline in the value of long-term investments in several investees. Income tax expenses was $2.4 million in the second quarter of fiscal year 2021, compared to $8.1 million of income tax benefits in the second quarter of fiscal year 2020.
Net income attributable to TAL was $15 million in the second quarter of fiscal year 2021, compared to net loss attributable to TAL of $23.5 million in the second quarter of fiscal year 2020. Non-GAAP net income attributable to TAL, which excluded share-based compensation expenses, was $52.3 million compared to non-GAAP net income attributable to TAL of $5.3 million in the second quarter of fiscal year 2020. From the balance sheet, as of August 31st, 2020, the company had $2,206.1 million of cash and cash equivalents, and $580.8 million of short-term investments, compared to $1,873.9 million of cash and cash equivalents and $345.4 million of short-term investment as of February 29, 2020.
As of August 31st, 2020, the company's deferred revenue balance was $1,172.5 million compared to $497.6 million as of August 31st, 2019, representing a year-over-year increase of 135.6%, which was mainly contributed by the tuition collected near the onset of part of the fall semester of Xueersi Peiyou Small Class and online courses through xueersi.com. Now, I will hand the call back to Mr. Luo to briefly update you on our strategy execution and provide the business outlook of the next quarter. Rong, please.
Thank you, Echo. Our business has withstood the unprecedented pressures of COVID-19 outbreak in the first half of this fiscal year and delivered a 30% revenue growth in Renminbi terms. We demonstrated our ability to provide quality education services to our customers, even in this unexpected circumstance. This underscores the resilience of our long-term vision and strategy, our comprehensive and competitive product portfolio, as well as our capacities, capabilities in technology and operations. Today, along with the encouraging progress in China's resumption of their activities and liveliness in all walks of life, our offline business is already in the process of gradual recovery. We will continue to pursue our offline development strategy to further expand our capacity at pace. We will cover more low-tier cities and further intensify our coverage in existing cities based on demand.
Meanwhile, we will keep following all relevant government policies and regulations regarding the national public health, and remain well prepared to deal with any public health contingencies in the coming period as always. As for online, under the current competition landscape, we made no fundamental change to TAL's current execution and the long-term strategies. We will keep working on enlarging our online market share by continued investment in technology, teacher supply chain and marketing, and consistent hard work of building all-around online services with superior customer experiences. We firmly believe that as an education player, our long-term success is defined by the quality of our products, service, and technology. Let me turn finally to our business outlook.
Based on the current estimates, total net revenue for the third quarter of fiscal year 2021 is expected to be between $1,061.1 million and $1,094.3 million, representing an increase of 28%-32% on a year-over-year basis. These estimates reflect the company's current expectations, which is subject to change. That concludes my prepared remarks. Operator, we are now ready to take questions. Please.
Thank you. Ladies and gentlemen, we will now begin the question and answer session. If you wish to ask a question, please press star one on your telephone and wait for your name to be announced. If you wish to cancel a request, you can also press the pound or hash key. Participants are requested to restrict one question at each time. The first question comes from the line of Alex Xie of Credit Suisse. Please go ahead.
Hi, management. Thank you for taking my questions. I'd like to ask about your thoughts on the growth expectations and strategies. In offline, we have seen that in last quarter you didn't really spend much, what's your plan for capacity expansion in the future? In online, since the competitive landscape has, I think, worsened due to more financing of the private players, what will be your strategy going forward? Thank you.
Thank you, Alex. I think it's a very good question. Right before we talk about some certain numbers in this quarter or next quarter, we would like to take this opportunity to walk through our business logic and strategies as a whole to give you some information, what we are doing and why we are doing. If you guys can remember, six years ago, even earlier, we are a traditional offline tutoring company. We hire high-quality teachers, we rent a lot of learning centers, we build our classrooms, we fit students in, we do a lot of stuff like that. Under that operation models, in the past year, for example, between 2014 to 2018, we have entered the cities from 16 to 42.
That is kind of the traditional models we are running today, which is also quite common to see in the offline tutoring markets. Coming to the year 2018, I think around two years ago, if you guys can still remember, we, as the first one in this industry, we invented or developed the new online education models. We introduced the online live platform into our products. We built dual-teacher online models in this status. I think that's the first time for this industry we can separate the learning and the practice. We can separate the role of master teachers and assistant teachers. The master teachers can teach a lot of students at the same time, while the assistant teacher is taking care of helping them in the online environment. This kind of dual-teacher online class models has grown very fast.
Three years ago, we only have tens of thousands enrollments, coming to today, last quarter, we have 2.9 million enrollments. I think the revenue contribution of this model two years ago is only around 8%, four years ago is below 3%. Today, it's now 26%. If I consider the Peiyou Online is part of the online, I adding these two numbers together, last quarter we have already reached almost 40% of revenue, four zero, coming from the online model now, and around 70%, seven zero, coming from the online model now. This model has a lot of advantages, and we have seen they provide us a lot of capabilities to consolidate this market. If we say in the traditional offline models, we have a lot of competitors or maybe counterparts in different cities.
We're running the company for 17 years, but today we're still below 5% of their market share. Coming to the online model, we have the chance to see some situations may similar to the server market. Maybe in the future, the top players will be consolidated into maybe top three or top four. This kind of consolidation opportunity, that is never seen before in this industry. Definitely in this sector, we definitely will keep our investment in technology, in teacher supply chain, and as well as marketing, to make sure we can maintain our leading positions over here. On the other side, this model is not a perfect model to resolve all the problems. Education, in the end, is not a simple game of marketing. That is again, the key to playing education is we need to focus on providing high-quality service to the students.
We need to put students in the first priority. We need to not only take care of the efficiency of the marketing, but also take care of the effectiveness of the students. If we run into this model, we can see currently the online model today, the one master teacher, they're teaching a lot of students, but it's a little bit difficult for the students to ask questions, because one versus so many asking questions, making it kind of difficult. Which means maybe this model is more useful to train the students in the intermediate level or maybe relatively easier contents. It's a little difficult to talk about their very difficult questions under these models. One thing we also see that is, if today, I think not only us, also some counterparts, you guys probably have seen, they have raised the funds from the market.
I think more or less this model today, we have been struggling in the new customer acquisition costs, which is very challenging to all of us. We, as a player who invented this model, and I strongly believe is online compared to the traditional offline, still is a very revolutionary product, and we believe that can help us to get more market share, much higher than what we can expect in the past. We need to be very objective on these models. They have some pros and cons. We need to leverage our branding, our contents we have developed, and some other advantages we have to try to reduce the operating costs for the online model. If we benchmark the case in several markets, I think maybe in the end, only the top guys can have most of the revenues, most of the profits.
We definitely will continue our investment over here to make sure we are the leading players even in the coming few years. I think, because of the pandemic in the past eight months, I think, by chance, they introduced a new opportunity to us, which frankly speaking is a little bit out of our imagination before. It's actually, we quickly figure out, right after so many students, they have to use the online model for few months. When the offline learning centers reopen, actually, we see new opportunities. Maybe we can use the word of online merge offline. Sometimes this OMO sometimes may be a little bit misleading because different companies may use different definition on this. I would like to use one example of Shanghai of Xueersi business in Shanghai to show you guys.
In Shanghai today, I think we have, in the summer term, we have around 120,000 offline enrollments, which means we have 120,000 enrollments, the students coming to our physical classrooms and seeing the classroom. At the same time, in summer, we have around 200,000 paid online enrollments. If I taking out the promotion enrollments, the number can still be around 150,000 enrollments. This number is higher than the paid offline enrollments. Besides the paid offline enrollments and the paid online enrollments, we still have around 300,000 Xueersi Online School enrollments in Shanghai. Which means, when we go back to see our number in summer in Shanghai City, of course, summer actually is the peak for the year. If we go into the number in the fall or winter, the number will be lower than the summer.
I use the summer numbers, Q2 numbers, as example. Actually, if we figure out in Shanghai, we have maybe around 600,000 enrollments in one city. If you guys have followed my company for a long time, you probably can see that, I think four or five years ago, when you guys asked me, 'Hey, what is your expectations on one single city? How many enrollments you have over there?' For example, in Beijing, at that time, my answer is, I was very reluctant to say maybe I can reach the number of 150,000 enrollments for one city. Today in Shanghai, if we consider all of these enrollments together, we have around 600 enrollments there. This phenomenon also happened in Beijing, in Guangzhou, and a few other cities.
Because we did not roll out these offers to all the cities, all the grades and all the subjects, we do pretty good in the tier 1, we have a lot of white space in other cities now. This showing us actually, that is the kind of the new opportunity, and that shows us the direction, how we can evolve the offline learning center in the future. We pay some very deep dive on the Shanghai numbers. We probably see that why the parents choose the Peiyou Online on top of paid offline. Number one, I think more frequency. For example, they come to our offline learning centers once a week, but at the same time, for example, math, but at the same time, they will take the other online sessions, maybe the shortened class, at the same time for the whole week. Second, maybe it is more subjects.
For example, they come to my learning centers to study math, but maybe they study my physics and chemistry in the Peiyou Online platform, which is the case already happening in Beijing and Shanghai. Number three, maybe more students. We are also seeing, maybe in Shanghai's numbers, in summer especially, around 10%-20% of their Peiyou Online students, actually, they're new students. They're new students, but they are quite close to our learning centers in Shanghai. Which means they come to our learning centers maybe for the test, or maybe for the events, or maybe for the activities, but most of the learning activities finish online. This kind of model is showing us actually, that's a good opportunity for us to develop more and invest more on the localized online class models. What does localized mean? Number one, local contents.
I think in China, different provinces, they have different versions. When we talk about education's contents, they have maybe three kind of dimensions. Number one is the different versions. Number two is progress, different progress. Number three is the different difficulty levels. In China, more than 30 versions in Gaokao, and we probably will have over 200 versions of Zhongkao. We have maybe thousands of thousands of different versions if we consider the tests in a primary school. If we only have one version to all, sometimes we cannot resolve all the problems. The local content, local progress, and the local difficulty levels will be more helpful for the students and for the parents to know what exactly happening in the city and do better to them. Number two, local teachers.
When guys are waiting in our Peiyou Online classes, we use the local teachers to teach the students. For example, the teacher in Beijing, they teach the students in Beijing online. They know more about this city. They know more about what's happening in this district, so they can provide more tailor-made and more kind of useful information to them. I share one small story. Right before the national holiday, our Beijing school will hold a small event, is we ask the Peiyou Online students to come to our office to visit our Peiyou Online teachers. The children are very excited to see all of this. When we invite the parents or maybe the students, even they finish their study online, but they come to our offline centers for some events, for the seminars, activities.
I think this kind of online and offline blending kind of experiences is much better than the other one. Number three is local classmates. I think when you are studying with your classmates, you know some of the people in the class. That will make the interactions between student and student much more efficient than the other one. Number four is local services. We have a lot of schools, we have a lot of counties and districts. Only we can provide some specific information or maybe the materials or maybe activities specific for them, and they can fill in the local needs, which is more helpful than the one version to all products. We believe all of this can help us to evolve our Peiyou kind of the small class models to next stage.
As a summary, I think for the Xueersi Online School online part, we will keep our investment in technology, teacher supply chain, marketing, and some other necessary investments to make sure we can continue being the leading position in this market, because this market has shown very clearly the opportunity of consolidating their markets into several players. We definitely need to be one of them. Of course, we will balance the consideration of investment and ROI to make sure we don't do something stupid. Secondly, we will accelerate our learning center expansions of Xueersi Peiyou Small Class. We will also try to enter more cities in the coming two quarters. I think one number can show you guys is, just now you ask why we don't adding any new learning centers in second quarter, because second quarter, that is the pandemic period, so we can't do anything.
Coming to Q3 today, we have already rented 14 new learning centers, and we also have over 50 in the pipeline. I'm not sure the reporting periods of these new learning centers, because we still need some time to do renovations, ask for license. We definitely will have more on our pipeline to do that. At the same time, we're also trying to roll out the Peiyou online offerings to more cities, more grades, and more subjects, which definitely is our strategy to go. If we consider all of the information together, I think looking to quarter-over-quarter performance, I think Q2 is probably top line perspective, Q2 is the lowest growth quarter. Q3, based on information what we see today, Q3 will be better than Q2. Q4 were better than Q3.
Of course, all of these estimates are based on one assumption, is we don't foresee any significant public health crisis coming back again. We are confident we can continue our growth structure, as I talked about just now. Thank you, Alex.
Thank you for the questions. Next questions will come from the line of DS Kim of JP Morgan. Please go ahead.
Hello, sir. Good evening, and thanks so much for the insight you just shared. It's really helpful. I actually have one question. Oh, actually two questions. First, as a housekeeping, may I check what's the segment performances and the currency assumptions you have baked in the November quarter guidance? My real question, more important question is, as you said in the prepared remarks, customer acquisition cost seems to be surging a lot from this summer, and some say it's as high as RMB 4,000 to acquire one paying user, one paying student. I was wondering if you could share with us what you see on the unit economics and LTV of a student today at this level versus pre-COVID-19 period, i.e., how long, in your opinion, does it take to break even on the cash tuition level after spending current level of customer acquisition?
How many years or semesters that you think we can retain student on average against the spectrum of higher competition? Thank you so much.
Thank you, DS. That's a long question, right? Let's come back a little bit. In the first place, we have to say, because the online provider opportunities can consolidate to the top guys, and the top players have chance to have much bigger market share compared to the offline models. Definitely everyone is competing in this space. I think when we talk about the new customer acquisition costs, we need to split them into two parts. Number one is the new customers we acquire from the open market, for example, the WeChat Moments or maybe Toutiao platforms. For this kind of the new students, definitely the cost is higher and higher because everyone is competing over there.
On the other side is we still have the other part we call branding channels, which means when the people know who we are, like Xueersi is quite famous in the top tier, the tier 1 cities. They will come to us or they will download our apps automatically. We have some traffic coming from the branding channels, and we also try to do more to encourage the current students to refer our products to their friends or other students. We are also doing more over here. In general, yes, we are seeing the new customer acquisition cost from the open market channels is increasing, so we are doing more and more, and our percentage of the students coming from our branding channels or maybe customer referrals is also increasing.
On the other side, sometimes it's not a simple work to say new customer acquisitions decides the user economies for all education models. Right after students sign up platform, they still have a lot of steps we can do. Number one, how to increase the retention. Sorry, how to increase the conversion. How we can convert more students from the maybe Renminbi 49 class to our normal price class. Second, how we can do more to improve our retention rates. What we can see is actually we're making a little bit progress in the retention perspective. Number three is we also see with the opportunity for us to cross-sell or up-sell some other products to them. Number four is how we can make sure they can stay longer and longer with us. The numbers also increasing in our review.
I think in the online today, we will keep our investment level over here, and we are the market leading position now. We will maintain these positions in the futures. That's our goal to do. All of this, on the other side, for the Peiyou Online will be a different story from the Xueersi Online School. We don't pay that much marketing dollars on the Peiyou Online, so they will highly connected to the Peiyou offline business, and we don't see any huge cost of new customer acquisitions from there. Our margin from the Peiyou Online is much better than the other one. In the last quarters, we grow our Xueersi Online School enrollments by 116%. By the same time, we grow our Peiyou Online enrollments more than 130%.
We foresee this trend will continue in the coming few quarters. We need to balance different drivers in our hands, but we're still very confident with our investment in the online space and with our new strategy evolving how to use more Peiyou Online plus the Peiyou Xianxia offerings to the students. We are confident to deliver our healthy growth in the future. Thank you, DS.
Thank you for the questions. I will take the next questions from Sheng Zhong from Morgan Stanley. Please go ahead.
Hey, thank you for taking my question. Just want to ask a question about the competition on the online space. You also mentioned that some private companies are getting more financing. Now already we are at the time post-COVID-19. At this time point, what do you see the competition going forward? You emphasized that you are targeting to maintain the leadership position in this market. What kind of investment level do you expect? Thank you.
Yeah. I think sometimes we will maybe underestimate the complexity of online education. Sometimes people may simplify the competition to new customer acquisition costs. Actually, the online education is a complicated process. Not only you acquire the customers, you acquire students, but how you can make sure you can maintain all of them. Looking into these coming quarters, Q3 and Q4, I think we are fully prepared for any kind of the investments, if the other players there want to do so. We are fully prepared for that. We definitely will keep our investment right over here. If you only talk about marketing or sales side, I think that's only one part of the business.
I think besides the marketing, besides the money we spend on marketing channels, we need to know is actually we are also do a lot how to increase our local, sorry, our internal technology, our internal teacher supply chain and some other operational efficiencies. The competitions are not only happening in the front end, it happening for the all life cycles. Education is now a kind of a battlefield which can be concluded very fast. That takes a long time. For us, we will not be disturbed by anyone how much money they get. We need to be ourselves. We stick to our own strategy. We stick to our own pace. We continue to hire more teachers as well we can train, and we continue our growth strategy to cover more and more students. I think the first quarter, we grow 143% in enrollments of Xueersi Online School.
In the second quarter, we grow 116%, and we foresee we can continue to the high growth in the coming Q3 and Q4. At the same time, Peiyou Online is also a triple-digit growth in Q1 and Q2. We also foresee we can grow very healthily in both the Xueersi Online School and the Peiyou Online. I think the competition cannot be concluded soon. We need to fully prepare for that. On the other side is, again, online is one part of the games. It's not all parts of the games. Besides the online, we need to focus how we can use our advantages in the Peiyou Small Class, how we can combine our offline learning centers and our Peiyou Live offerings to make sure we can build more advantages, especially in the top 50 cities.
That's the kind of most revenue and the most of the profit coming from. That's our key strategy now. I don't have more information to share with you guys, but what we can say is we are fully prepared for that.
Thank you for the questions. Our next question comes on the line of Felix Liu of UBS. Please go ahead.
Good evening, management. Thank you for taking my question. My question is on the sort of the relationship or synergy between Peiyou and the xueersi.com. I see that Peiyou Online is making very good progress, with the strong growth and very small amount of the sales and marketing cost. Is there any possibility that we leverage the success of Peiyou Online to xueersi.com? Any thoughts on sort of adopting the best practice there? Is there any room to build more synergy between your offline network and the xueersi.com operation? Thank you.
Yes, that's a very good question. I think by the end of today, because the online market is still very big, everyone is fast growth. We have a lot of potential. We can grow more market shares. We prefer is the Xueersi Online School will kind of execute a breadth strategy, try to cover more geography, try to cover more cities, and compete in the whole country level. While the Peiyou Online will highly connect you to Peiyou learning centers, and we prefer the Peiyou Online can be depth strategy, going deeper, going more localized, and provide more individual services to the students in the local cities. For the Xueersi Online School, we are willing to pay the marketing dollars to promote and to get more market shares, blah, blah. We are willing to bear some kind of the operational loss from there.
At the same time, for Peiyou Online, we will develop our new strategies, sorry, our new solutions within the local community to make sure we can capture the deeper opportunities. We not only maintain high growth over there, but we also maintain the very good profitability levels right over there. Based on what I can see today, they will share in the branding perspective, in technology perspective, in the content perspective. They may also share some kind of the operational leverage between the learning centers and the online schools because they are using the same name. For example, the branding channel, the students come to the Xueersi Online School is because learning with Xueersi. We still prefer the Xueersi Online and the Peiyou Online can be run a little bit independently to penetrate different markets.
The market has very wide potential now, so we wish we can leverage these two offers to get more market share.
Okay. Thank you.
Thank you for the questions. Next question comes from the line of Mark Li of Citi. Please go ahead.
Hi, management. Thank you for the time. I want to ask for our offline operation. We have seen quite a bit of promotion that impacting the ASP. May I know how is the promotion development heading to the next quarter, and what is our offline growth target for the next few quarters? How much recovery we will see for offline? Any color is appreciated. Thank you.
I think for offline, the coupons you mentioned just now, actually part of that is because in the first two quarters, because we moved the offline offers to the online due to the pandemic. We give certain type of coupons to the students. They can use the coupons in Q2 and maybe part of Q3. That is now the normal promotions or coupons you have seen in the past. I think looking into the coming two quarter and even next year, I think promotion and coupons is not our most important rampart to do this. I think we still need to do more work. It's how we can provide the online and merge offline models to the students.
For example, within the learning centers we have in one city, how we can find some ways to do marketing or penetrate the students, maybe three to five kilometers around the learning centers, or maybe 15 minutes, 15 minutes by drive, by car. How we can use these learning centers, and how we can attach the online or Peiyou Online offers to them. Today, frankly speaking, we're only running this model in the city level. We still did not go into the detailed management side. If our operations can be more detailed and be more kind of efficient than today, we can definitely see much more opportunities. I think we need to evolve or maybe reinventing out these models through products, through research and development, but not only through so-called promotions or coupons.
We wish current offers can provide students the new experiences, and we can use their data to provide more tailor-made or individualized solutions to them. That's our key to play in the future.
Thank you.
Thank you, Mark.
Thank you for the questions. Next question comes from the line of Lucy Yu of Bank of America Securities. Please go ahead.
Thank you, , for taking my question. My question is also related to what Mark has just asked. I'm aware that in the offline market, in the Peiyou Xianxia, so this summer we have offered some discount in addition to the coupon. I'm wondering how much of that discounting has impacted your GP margin or OP margin in your Peiyou offline. I am trying to understand whether your Peiyou offline business margin is flattish or declining or still improving. Thank you.
I think especially in Q2, definitely the Peiyou offline business, the margin is decreasing, of course, because the pandemic. That is not a normal quarter to compare sometimes. Coming to Q3 and Q4, where everything is coming to normal, we probably can see that they will recover from the second quarter's low numbers. Same as what I said just now, maybe if we look into all our situations, the second half will be better than Q2.
Thank you, Also, may I just follow up on your third quarter result? Would you mind break that down by business in terms of growth outlook? Thank you.
Sorry, we don't disclose that detailed information due to the competitive reasons. The direction is actually our Xueersi Online School growth will be similar as what we can see in Q2, plus minus a little bit. We also see a recovery in growth in the Peiyou Small Class business. That's the rough ideas we can share with you guys. Thank you.
Thank you. Thank you.
Thank you for the questions. We have another question from the line of Alex Liu of China Renaissance. Please go ahead.
Thanks, , for taking my question. Two questions. The first question is could you share with us TAL's strategy on the so-called very popular AI online class? I understand some of your competitors, including Banma, are quite aggressive on this model, and we knew TAL also has a dedicated product for this. Just wondering what's the plan there, and specifically, do you think this model will eventually reach similar revenue size to our current online model? That's the first question.
Okay. That's a very good question. I think maybe last year in my earning calls, I recommend one book to you guys. The book was called "On Intelligence," written by Jeff Hawkins from U.S. I think that's a very good book, you guys can pay attention to that. Sometimes AI products is a very good word, but sometimes too strong word for us. We are still way off from the AI. We're still far from AI. I think maybe in the future, if our online business models is getting bigger and bigger or maybe more and more people coming in, and our OMO models, especially pure online is also fast increasing, so we can penetrate a lot of students and we can collect more of their datas.
Not only their data, they study my learning centers, but also some data they study in the schools or maybe even at home. We can have more data. Maybe possible, we will have chance to develop a new type of maybe intelligent self-adapt learning models. I think that will happen definitely, today we can't figure out what's the right pathway to reach that level. What we can do is we invest in our online, both in the Peiyou Small Class and Xueersi Online School. We try to invest in our AI lab, we try to invest in more research like the brain science and something else to make sure we can figure out one way to do that. I think this kind of product still has a long way to be mature.
All what I can say just now, it's only comment the product of my own company. I have no comments on some other products because I'm not the right one to do that. In our views, we'll continue to invest in this area, we'll continue our investment, and we strongly believe in the future, the new type of products will come out. The only problem is when. We wish all the efforts we make in the online part can contribute to the next level products in the future.
Thank you, management. With that, ladies and gentlemen, that does conclude our conference for today. I beg your pardon. Do you have any follow-up questions, Alex?
Hey, Alex. I remember your second question.
Just quickly on the gross margin decline. Just wondering how much is due to the teacher compensation cost increase, and how should we think about this trend going forward?
Yeah. Let me answer very briefly. I think the gross margin, we have two factors. Number one is we continue to grow, we continue to develop more learning centers. Same as what I said just now, I think, in Q1, we are adding around 20 new cities, Q2 is one new cities. At the same time, we're adding more learning centers in the current cities now. Q2, the number is very small, in coming to Q3, we have a big number over there. Today we have 14, we have over 15 in the pipelines. With all of this, definitely will impact a little bit in my gross margins. On the other side, because Q2 is also a timing for us to increase our teacher accommodations. I think we continue our strategy same as before because we are education company.
We talk about a lot of synthetic technology, we still need to pool our teacher, pool our people in the first priority. We need to take good care of them. That's the normal practice every year. We don't have specifically how much coming from the new center kind of development, how much coming from the teacher accommodations. Both the investments on the new capacity and the investments on our teachers are very important for our long-term growth. Thank you.
Thank you. With that, ladies and gentlemen, we have now come to the end of the conference call. Thank you for your participation. You may now disconnect your lines.