In the travel scene, the taxi market always has its own emphasis, but few people pay attention to the freight O2O market that was born at the same time.

After eight years, the journey goes to the left, the freight goes to the right, and the two tracks take completely different roads. Freight O2O enterprises represented by freight Lala, Full House and 58 Express. It is not positioned in the freight market as quickly and on a large scale as the tourism market. Even though the head enterprises occupy a high market share, they are still the tip of the iceberg compared with the vast market.

The reason is that some media believe that the breadth of the industry service chain and the depth of technology are the biggest differences between the two industries, and this also makes it impossible for companies such as freight Lala to directly copy the growth model of the travel market, that is, relying on the huge offline legion. Rapid expansion.

Accordingly, freight companies need to focus more on a field that almost no one touches, but it takes longer to accumulate-technology.

First, the logistics park roadside freight O2O budding

Freight transport is a huge blue ocean market that is no less than passenger transport.

The data shows that there are 20.89 million employees in China's road freight industry, of which truck drivers account for 87.7%, and 99% drivers say that their families mainly rely on their driving income, or only have the source of driving income. If employees in related industries are added, the families supported by them will support more than 100 million people.

Typical Shenzhen Huaqiang North has 25,000 * * enterprises, 26,000 individual industrial and commercial households and 200,000 employees, with an average daily flow of 300,000-500,000, with a peak of 800,000. Countless goods are loaded by trucks and sent to the whole world at various time periods of the day.

Before 20 13, freight drivers often gathered at the entrance of the logistics park, staring at the blackboard at the door, looking for opportunities to take orders in the dense information.

These logistics parks generally have information departments, and staff members summarize user needs on the blackboard. The blackboard says "What kind of goods are there, how much they weigh, what kind of car to look for and where to go".

Master Chen in Huaqiang Beila 10 recalled that he used to "work" at the entrance of the logistics park. When he sees the correct information, he will call the customer to pick up the order. Sometimes, customers will come directly to the roadside to find them.

In addition to offline, logistics companies, drivers and shippers will also gather in the QQ group to release information. "Shippers generally need to contact the driver a few days in advance. If the driver sees that the information matches, he will call the shipper.

"Sometimes I will take the initiative to send messages in the QQ group. For example, if I shipped goods from Shenzhen to other places and didn't want to come back empty, I would leave a message in the group in advance, hoping to find the need for a return trip. " Master Chen said.

But the reality is not satisfactory, and it is not always possible to find the need for a return trip. With more and more freight, their income began to decrease. Sometimes they can't get one or two orders a day. Many times, several drivers sit together and play poker. "But I am very anxious because I have a family to feed."

Is there any way to match truck drivers and shippers more efficiently, so that shippers can easily find drivers and drivers don't have to "scrape" on the street?

Driven by such demand, freight O2O company came into being.

Freight Lala was established in 20 13, and officially entered the mainland market in the second half of 20 14, targeting Guangzhou and Shenzhen, two important freight centers in South China.

2065438+In September 2004, 58 Express App was officially launched, focusing on urban short-distance freight.

20 17 Yunmanman and Truck Gang merged to form Manbang Group. Three years later, Didi also set up a freight company and officially joined this battlefield.

Many drivers didn't have smart phones in the early days. In order to help the cold start of products, the product managers of freight companies often "hang out" with food stalls in the suburbs of Shenzhen and Guangzhou, and become good brothers with dozens of drivers to persuade them to use freight software.

In this way, the freight O2O market began to falter.

Second, the freight bid farewell to the laggards, but the subsidy will not work.

But no one thought that freight would be copied quickly and on a large scale, just like taking a taxi. However, in reality, this has poured cold water on freight companies.

Around 20 15, many companies hope to solve the freight problem by using passenger transportation online to offline, that is, burning money subsidies. At the peak, more than 200 companies aimed at this market, and the overall market grew at a rate of more than 20% every year.

Although both travel and freight seem to be transportation services, the problem is that the service chain of freight and the pain points that can never be solved are not solved by subsidies at all.

Generally speaking, there are two main online freight platforms: one is an information publishing platform that simply shows the freight demand and the contact information between the carrier and the merchant.

The other is to promote end-to-end freight transactions-a "closed-loop" platform from ordering, pricing, prepayment, freight matching, order tracking and payment settlement confirmation.

For the latter, the technical problems encountered are far more difficult than expected. The first problem before us is informatization.

Zhang Hao, the current CTO of Cargo Lala and former vice president of Hungry Technology, lamented in an interview that compared with other O2O platforms, there is still a lot of work to be done to make Internet freight intelligent.

The first challenge facing the intelligentization of Internet freight data is the matching of "cars" and "goods".

In the taxi market, people and cars are standardized, so every single service can be completed through homogenization. However, in the freight market, the mismatch between cars and goods cannot be loaded. On the cargo platform, the types of vehicles exceed 17, and the goods are even more varied.

The second challenge is road restrictions. In urban roads, there are a large number of truck restrictions, height restrictions, weight restrictions, axle load restrictions, and some policy factors. In some sections, small commodities can pass, but China commodities may not. These all need data to accurately match cars, goods and the environment.

This is a common problem faced by the industry.

At that time, many companies relied on huge subsidies to drivers to maintain the number of users, but they did not really get users. These subsidies come from venture capital funds, and the money for venture capital is limited. Gradually, some companies began to be unsustainable.

Soon, a large number of companies were short-lived in that war: goo goo express, No.1 cargo and other companies lost one after another.

Cargo Lala is a typical representative who has experienced the ups and downs of this track.

This company has also been involved in the subsidy tide for some time. However, their team believes that if subsidies are given blindly, it will attract many users. This is not the person that the platform can really retain, and it is not good for the platform and real users.

In the cold winter, the cargo girl also arrived in do or die, so the founding team at that time made a decision: stop subsidizing users.

This decision is like a broken arm to survive. Because other platforms are still subsidizing users, the order volume of cargo Lala has dropped from more than 1000 to more than 400 per day, and everyone is under great pressure, but the team of cargo Lala still believes in one thing-solving the real pain points of real users.

But on the other hand, a freight technology called no man's land in the world began to be born quietly in this company.

Third, technological breakthroughs have become the "main channel" for freight transportation.

The product evolution logic of Internet freight companies is actually not complicated, that is, how to improve the freight efficiency step by step to the extreme. In fact, the essence of logistics is the efficiency game from point A to point B.

Luo Zhaofeng, the current deputy director of cargo Lala products, is one of the earliest product managers of cargo Lala. According to him, around 20 14, Cargo Lala is just a simple information matching platform with very low efficiency. Drivers need to choose from a large number of orders that match their models and distances.

However, let everyone call online from offline. Although the product function is simple, it is already the biggest breakthrough for this industry.

From 20 15 to 20 16, with the deepening of market understanding and the accumulation of information and data of drivers and users, cargo operators began to carry out accurate matching to help drivers filter out orders that are not suitable for performance and improve efficiency.

After 20 19, with the improvement of algorithm ability and the expansion of production and research team, the algorithm began to be fully put into use in the freight station, which can improve the execution efficiency and balance the experience of drivers and users. Because of the use of the algorithm, the waiting time for a driver to complete a single order is reduced, and users can easily find a car that suits their own goods characteristics and preferences, greatly improving efficiency.

Today, the production and R&D team of cargo Lala has grown from 200 to 300, and now it is close to 2,000. From the beginning, the R&D team was only composed of several departments, such as the front desk and the middle desk. Now all departments are subdivided and the degree of specialization has improved.

In 2020, Zhang Hao shared a set of technical system created by Lala in an industry conference to reduce costs and increase efficiency. This middle-level system is called "intelligent brain" in Lala, and it is divided into four modules: supply and demand, invoicing, marketing and pricing.

The supply and demand engine predicts the distribution of transportation capacity, demand forecast and transportation capacity forecast;

The billing engine is matching, and the order is coming, which capacity is given;

The pricing engine sets the prices of different freight rates and different models, and sometimes the prices are different in different road sections and different time periods;

When does the marketing engine subsidize and promote demand?

The system is based on AI, big data, maps and other basic capabilities, and solves the core resource optimization problem through the self-developed operational optimization algorithm framework. On the basis of Internet of Things technology, all production factors of logistics are intelligent, and then the most accurate matching is carried out.

Among the production factors of freight transportation "people, cars, goods and roads", it is easy to realize the intelligence of people, but it is difficult to realize the intelligence of cars, goods and roads.

Non-standard goods, complex vehicle types, complex and diverse road restrictions, etc. Is a unique challenge in the field of vehicle freight.

"Many colleagues will feel a little uncomfortable when they first join, because the technical difficulty of the freight scene will make it difficult for industry experts to cope." Shi Lichen, the technical director of cargo Lala, who once worked for passenger O2O Company, said frankly.

Shi Lichen is in charge of the trading engine and map team of Lala, and once won the first prize of Wu Wenjun Artificial Intelligence Natural Science Award. In his team, there are more than 30 core experts, all from first-line companies such as Ali, Baidu and Meituan, which can be regarded as the leading figures in the domestic map industry. He admits that the digitization of the "road" in the freight scene is much more difficult than expected.

"Although the challenge is great, we have achieved industry leadership in two technologies." Shi Lichen introduced that one is the matching of vehicles and goods, and the other is the recommendation of loading and unloading points.

Take the matching of cars and goods as an example. After the user places an order, the platform realizes real-time matching according to the characteristics of the goods, the capacity distribution of the adapted models, and the comprehensive factors such as road conditions and traffic restrictions. Through the accumulation of big data and the iteration of intelligent decision-making system, the performance efficiency is continuously improved.

Another example is the recommendation of loading and unloading points. According to the user's order address, recommend the specific location suitable for truck loading and unloading. By recording and analyzing the historical distribution of loading and unloading points through big data, based on the topological relationship between the ordering address and the road network and the road restriction information, the intelligent recommendation system can recommend to users where it is suitable for parking, loading and unloading.

With the continuous accumulation of technology, in the scene of moving goods, the intelligent order-splitting system of cargo Lala can handle the real-time distribution between millions of orders and hundreds of thousands of drivers on domestic platforms every day.

Next, cargo Lala will promote the technology of cargo identification and volume measurement, and build an overall solution of Internet freight map, which will further advance in the direction of intelligent freight data.

After eight years, the hero ups and downs, freight difficult to move forward. Lala's case proves that in this O2O market where China is still a blue ocean, only the investment and persistence in technology can truly embrace it.