Blog

Shared Bikes In China (2)

Cash and Deposits

At first, nearly all bike-sharing startups take deposits, usually ranging from ¥99 to ¥299.

The deposits are hard cash for startups; they are like life-time membership fees if the companies survive and users keep using the apps.

To put it another way, deposits are indefinite free borrowings.

Deposits and investors’ money are fundamentally different (in accounting), but both are actually (viewed as) cash sitting in the bank for startups.

Some may compare this model with gym memberships. When users pay upfront, gyms use the cash in expansion and operation.

Just invest (or borrow) and open one gym first, then one can use the membership fees collected to subsidize the opening of the second gym..  bike-sharing companies can use the deposits collected in one city to subsidize the expansion into the next city, to pay back any loans, to pay employees’ salaries…

Then comes the problems… for those startups that are not well-funded.


Closures in 2017

In June 2017, the first startup, Wukong Bicycle 悟空单车, announced to stop its operation.

Started in Chongqing, Wukong Bicycle has another layer of to collect operation cash – “operation partners” who would pay Wukong Bicycle an upfront fee to claim profits for a certain number of shared bikes. To be honest, “operation partners” are not treated as partners as Wukong can profit from selling bikes to them and management fees. Its “alternative fund raising” was a little concerning.

By the summer of 2017, more reports about the closure of bike-sharing companies ignited the concerns from users. When users were asking their deposits back, the real cash crunch came/intensified.

In November of 2017, it was reported that most of the bike-sharing startups have problems with their deposits. [60多家共享单车停运用户押金之痛难解 – 证券日报]

The largest of them (in 2017) was bluegogo 小蓝单车, which had raised ¥400 million. It started as a supplier for bike-sharing companies but then decided to enter the race. It had achieved a No.3 position in the space but things (financing) went south in June 2017.

As mentioned previously, it was acquired by Didi at the beginning of 2018.

Read more on 还原短命小蓝单车的365天 – 36氪  小蓝单车生死故事 – 36氪

Not a coincidence, an even broader problem in China emerged in 2017: P2P financing.


Meanwhile, the top-tier companies seemed to live well with millions of dollars raised on the capital markets.

Many required 0 deposits to entice users (and accelerated the failures of smaller players that reply on deposits).

Actually, China Consumers Association (CCA) encouraged bike-sharing companies to charge 0 deposits in December 2017.

 

 

to be continued..

Shared Bikes In China (1)

A History

ofo

The battle between 70+ bike-share startups (or the battle between their colors) could be traced back to the summer of 2014, when ofo (wikipedia) was started by students at the Peking University in Beijing.

Source: play.google.com

ofo initially focused on bicycle tourism before deciding on bicycle sharing. At first, it was only doing campus bike sharing. In May 2015, the team appropriated the investment fund for purchasing new bicycles and enticing PKU students to partake in bicycle sharing. [PKU news] [¥9 million seed/angel]

Shared bikes became crazy in 2016. ofo took off in 2016 with ¥15 million Series A in January led by GSR Ventures 金沙江创投 and followed by Dongfang Hongdao 东方弘道, then Series A+ of ¥10 million in April and Series B of “tens of millions” USD led by Matrix China 经纬中国 in June.

Yet the fund raising didn’t stop there. ofo raised another $130 million led by Didi (C-1), Coatue (C-2), and funds affiliated with Xiaomi (C-2) in October 2016, officially marching into city businesses instead of focusing on universities.

mobike

Two major differences were separating ofo and its main competitor mobike at the beginning stage: 1) mobike focused on cities from day 1 while ofo was for universities at first 2) the locks

mobike | Source: wikipedia

The first generation of ofo bikes has an unchanged passcode sent to users while mobike’s are unlocked by wireless communications between the phone, servers and the bike. mobike also uses GPS from the beginning.

ofo 1st Gen. | Source: tianjimedia.com
See the source image
mobike 1st Gen. | Source: eastday.com

Nowadays, shared bike companies are using similar product (lock) strategies for safety, management and data. [read more about smart lock technologies involved]

mobike raised its Series A of $3 million in October 2015 led by Joy Capital 愉悦资本, Series B of $10 million in August 2016 led by Panda Capital 熊猫资本, Series C of $100 million in September led by Warburg Pincus and Hillhouse Capital.


By the 2016 holiday season..

both ofo and mobike finished with their Series C with nine figures, while many other startups were just launching their services and raised their Series A, including bluegogo which would became the third largest service provider before went bankrupt and later acquired by Didi.

Image result for bluegogo
Source: crunchbase

At that time, another startup Hellobike 哈罗单车 which would be threatening to the first-movers, was also just preparing to launch its bikes (started with 2nd/3rd tier cities) and just finished its Series A with GGV.

Image result for hellobike
Source: crunchbase

The frequency of fund raising is probably the most remarkable part of the history (to me).

I actually wish that something similar won’t be happening again…

Businesses, investors, users/citizens and regulators all need some time to really think over.

For bike-sharing startups, they were going to feel something different in 2017…

 

to be continued…

Crazy Valuation For Tea Chains

Two tea chains hot on the capital market in China – Hey Tea and Naixue Tea.

Valuation are said to be ¥8 billion (2019) and ¥6 billion (2018) respectively.

Each may have 200-250 stores. (Say Hey Tea 250 by end of 2018 and Naixue 200 by the end of 2018)

An average store say has a revenue run rate of 250k * 12 = ¥3 million / yr

Then revenue run rate is 750 million for Hey Tea and 600 million for Naixue.

With a revenue multiple at 10.0x, 600 * 10 = 6 billion for Naixue…

and 7.5 billion for Hey Tea…


Seems “good” in numbers… thanks to the support from Luckin Coffee..

Luckin has $4+ billion market cap and 20-25 revenue multiple (not run-rate)

 

 

Hupu And Toutiao (ByteDance)

Today Toutiao acquired 30% of Hupu for ¥1.26 billion, valuing it at ¥4.2 billion.

Formed in 2004, Hupu provides marketing planning, sports events marketing and management, as well as events management. It also operates businesses such as offline eSports events, e-commerce and gaming co-operation. It owns Hupu.com, the sports site with the most page views in China, a retail site for trending sports gear, and the app for Shihuo.cn. More than 30 million users had registered on its websites and apps as of March. [yicaiglobal]

虎扑

A History

Hupu has been active in the capital market for a while.

It was pursuing an IPO in 2016 on the Chinese stock market with a revenue of ¥200 million in 2015.

[Hupu IPO Prospectus]

Before its IPO efforts, Hupu finished its Series C of ¥100 million in 2014 led by Greenwoods (景林) and Series D of ¥240 million in 2015 led by Guirenniao.

Then in 2017, Hupu’s IPO didn’t go through as planned..

..which led to a round of ¥618 million led by CICC (also the underwriter for Hupu’s IPO).

Toutiao + Hupu

Ways to cooperate:

    1. ads – precisely targeting a community with above-average purchasing power and distinguished tastes & shopping categories/habits
    2. video – Hupu’s sports video capabilities/rights/viewers will have synergies with Toutiao’s video infrastructure and recommendations; plus, Hupu will be one of Toutiao’s efforts to march into sporting business
    3. e-commerce: direct sale, and with the help from ads and video; Douyin’s video can lead to shopping on Hupu’s e-commerce platforms

 

 

「Video of the Week」How deep learning is poised to transform the life sciences and healthcare

Applying deep learning in the life sciences
Deep-learning algorithms have proven to be powerful across many aspects of the life sciences and healthcare. Here are eight examples of how they’re being applied throughout the pipeline:

  1. Deep-learning-powered lead optimization: Lead optimization is a key problem. Once you’ve gotten your lead, how do you explore a 1080 space to find a better molecule based on your lead? According to a recent publication, a type of deep-learning engine called an “encoder” can encode structures with a discrete library into a continuous gradient space. That allows you to explore a much larger area of compound space than you have in your discrete library. You can then drop novel molecules out of the gradient space, design small compounds and test them to see how they perform, compared with all of the molecules in your existing library.
  2. Predicting compound activity: One variant of deep learning is known as “one-shot deep learning,” in which you train the algorithm to identify differences, rather than similarities, in data. That requires a lot less training data, which is a critical advantage. A recent paper shows that one-shot deep-learning approaches excel in predicting compound bioactivity based on training with a small set of data. So, if you’re in a new space and don’t know much about the activity of the compounds you’re using, you can still apply these approaches to get a reading of a novel molecule.
  3. Cell assay imaging analytics: Applying deep learning to cell assay imaging is an obvious path to pursue. All of the papers that have been published in this space so far have shown that deep-learning algorithms can do as well as or better than humans in detecting things like phenotypes. Better yet, they’re more efficient. You still need a human who has to be trained initially and can understand the outcomes. But overall, people are freed up to think about higher-level matters, rather than sitting there looking at images.
  4. Toxicity prediction: Deep learning has been shown to be significantly more effective than existing methods in predicting the toxicity of any given molecule. Over time, it learns to look for the specific elements or substructures that are causing the toxicity.
  5. Counterfeit scanning: Deep-learning systems can be trained to detect counterfeit drugs on the web and other sources by examining the package labeling or the pills themselves. They can pick up on little differences between real and counterfeit with a pretty high degree of accuracy. For example, logos or lettering printed on packaging can be slightly off because the presses aren’t exactly the same as the original.
  6. Electronic health record (EHR) analysis: Deep-learning approaches have been shown to significantly outperform traditional methods in doing things like patient cohort identification, readmissions analysis, clinical trial recruitment and clinical predictive modeling from EHR data stores.
  7. Language translation: Clinical trial protocol translation is a critical issue. If you provide a protocol in one language and then have someone translate it into another language, how do you know if they have translated all of the important phrases accurately? You can use deep learning to translate back into the original language and figure out how to normalize all those key criteria in your clinical trial.
  8. Electronic laboratory notebook (ELN) analysis: Many life sciences companies know they have a huge amount of information in their ELNs, but no good way of getting it out. You can have deep-learning systems go in and pull out this information; they can literally read the text in an ELN and figure out what it means.

How deep learning is poised to transform the life sciences and healthcare

 

OYO Countered In China

To be honest, I haven’t tried either brand.

But the game of capital is on.

OYO is grown from India with series B ($100mn), C ($90mn), D ($250mn), E ($1bn) led by SoftBank. Huazhu has participated between D and E; Grab, Didi, Airbnb has participated in Series E separately.

From my understanding, OYO is pursing a model that provides minimum standardization with the least cost while getting data and digitalizing management.

The most valuable thing OYO provides is the traffic (if any), which is where OTA’s profits come from and where hotel chains are good at.

The brand itself tho, doesn’t have much power. China’s overall hospitality standard is higher than India’s I think (with players like Jinjiang, Huazhu, etc.)

OYO’s rapid expansion in China might make it worse.

But it is really big – said to have 10k+ hotels and 450k+ rooms on its website.

It now has a three-tier branding: 轻享,智享,尊享

Branding-up and providing more values is really important. Hotel owners may end up with less profit in the long-run.

It’s like imperialism in the hotel sector.


The traditional hotel sector in China has reacted with their own exploration in “light franchise” (but might be late for this game).

H hotel by Huazhu

轻简 by Botao

轻住 from MeituanDianping

OYU by TongchengYilong

 

 

China’s Fresh Produce E-commerce (2)

previous post – China’s Fresh Produce E-commerce (1)

Second Round

The two words that characterize the second (current) model is “front warehouse” (前置仓).

By managing more distributed front warehouses (mostly in cities), the fresh produces e-commerce companies can usually deliver within 2 hours after an order is placed.

The model could be exemplified by the current focus of Hema (盒马鲜生) and MissFresh (每日优鲜). The difference – Hema’s warehouses are also consumer-facing stores; MissFresh’s warehouses are expanding much faster and many have no “experience store” functions.

Hema is financed within Alibaba (consolidated in earnings reports) and MissFresh has raised several hundred millions from Tencent.

Hema

Hou Yi (侯毅), the CEO and founder of Hema, worked for JD.com and in charge of JD logistics, prior to joining Alibaba. He has rotated to the O2O (online to offline department) and was the founder of the predecessor of JD Daojia (京东到家), JD’s delivery team. It has been rumored that firstly HOU proposed to Richard Liu, the CEO and founder of JD.com, about the idea of Hema; unfortunately, LIU did not approve that idea at that time. Later HOU approached ZHANG You (张勇), the CEO of Alibaba Group, and get offered to join Alibaba and try out his idea. [equalocean]

Since its beginning in 2016, Hema has now expanded into 130 stores in 19 cities as of March 2019.

Tmall’s fresh stores have also been consolidated into Hema’s operations, announced in December 2018. (meanwhile, JD’s fresh produces team was combined with 7-fresh in the same month)

Its operation summary compared with competitors provided by EqualOcean.

MissFresh

Jan 2017, Series C, $100 million led by Lenovo Capital

Sep 2017, Series C+, $230 million led by Tiger Global and Genesis Capital

Dec 2017, Series D, said to be $500 million; another report said $200 million to spin off 便利购 (the unmanned shelves)

Sep 2018, Series E, $450 million led by 高盛(GSIP)、腾讯、时代资本、Davis Selected Advisers

Tencent has invested four times so far.

Its core competitiveness lies in its front warehouse network, inventory management system and local community operation. MissFresh has an average duration of inventory of 2.5 days.

MissFresh is targeting gross margin at ~20% in the long-term while maintaining an operating margin at 10-15%.


So we can see that attempts are made to locate “warehouses” closer to consumers and to shorten the waiting time to ~30min.

Alongside the 1st and the 2nd rounds, there is another attempt – not necessarily new but will take some time to stabilize, if possible) – to combine 1) mainline logistics, 2) city delivery networks, 3) front-warehouses + community stores, all facilitated by digitalized and AI-driven systems.

 

China’s Fresh Produce E-commerce (1)

Overall Market Size And Online Percentage

Fresh produce has long been seen as an undeveloped area within e-commerce.

The penetration rate of e-commerce in the Chines apparel market is ~35% in 2017, while the fresh produce category’s is only in the single-digit range (but growing fast!).

The overall size of fresh produce market in China is around 5,000 billion RMB. So the online sales is ~1% in 2015 and ~4% in 2018.


First Round

The first wave of exploration is an extension of traditional e-commerce platform + specialized logistics (cold chain).

Miao Fresh (by Tmall, Alibaba) and JD Fresh (by JD) are the two examples (and first movers) in this round.

Self-built logistics has been one of JD’s core capabilities for a long time. And the war in fresh produces, requiring an upgrade in cold chain, has made JD Fresh a very competent player in this field (besides the upgraded potential in logistics-as-a-service) .

JD Cold Chain Logistics | Source: JD Fresh Presentation

Tmall seems to have a better position in recruiting overseas/premium sellers (cherries, lobsters, crawfish, etc.)

Meanwhile, Alibaba (Tmall) has been investing in Yiguo continuously. In March 2016, Yiguo raised ~$250 million Series C from Alibaba and KKR.

Yiguo’s subsidiary ExFresh (安鲜达), China’s largest cold-chain company established to serve the fresh food via e-commerce, according to a new release from KKR in 2016.

Source: exfresh.com.cn

And Suning, a strategic partner with Alibaba and a very strong player in 3C retailing, led a round of $200 million (Series C+) for Yiguo in Dec 2016.

Then in Aug 2017, Yiguo finished its Series D of $300 million from Tmall (Alibaba).

But the battle in fresh produces is more than the battle in cold chain + traditional e-commerce. It involves exploring the new format of retail.

And so comes the Second Round…