In the United States and Europe, the transactions of large shopping centers rely heavily on authoritative third-party data. But in China, such applications are not yet mature.
Respondent: Shanghai Huiner Information Technology Co., Ltd.
Respondents: Chairman Zhang Hongjun, President Ding Yao, Chief Data Science Consultant Hu Yu
Hanna Technology’s booth at the World Artificial Intelligence Conference (WAIC) Source: Huiner Technology
At the World Artificial Intelligence Conference from August 29th to August 31st, Hanna Technology demonstrated a new generation of data analysis based on Huawei Hisilicon AI chips and deep learning algorithms Products, and jointly set up in Huawei’s “Smart Retail” sector. In this regard, we participated in the media conference of Huiner Technology. At the meeting, Huiner Technology synchronized the latest developments and business ideas of the company.
According to the 2019 semi-annual report of Huiner Technology, in the first half of 2019, Huiner Technology Company realized operating income of 109 million yuan, a year-on-year increase of 42.64%; realized net profit attributable to shareholders of listed companies was 23.25.78 million yuan, a year-on-year increase of 64%. The net profit after deduction was 2,216,900 yuan, a year-on-year increase of 134.44%. The increase was mainly due to the income from the new fire video analysis system.
Public information shows that Hanna Technology’s main business is the comprehensive use of ABI technology (AI+Big Data+IoT), based on offline physical business scenarios, using sensors to provide users with data collection services, and then The foundation establishes a commercial big data platform to provide data management and data operation services. At present, there are 14 molecular companies in addition to the Shanghai headquarters, and over 170 service outlets covering more than 400 cities and regions across the country.
The following is the content of the interview, edited and revised:
Q1: What are the barriers to competition and the moat?
Zhang Hongjun: Since its establishment in 2004, Huiner Technology has started its application practice based on computer vision related technology in physical business scenarios, and has 15 years of industry experience. Hanna Technology currently covers more than 1,500 shopping centers and more than 50,000 branded retail stores, with an average annual statistical analysis of more than 10 billion passengers. In the whole year of 2018, Huiner Technology has analyzed and counted more than 12 billion passengers nationwide. The total number of passengers analyzed and counted since 2019 has exceeded 9 billion.
After
, we have collected data from these different shopping centers and different brands on our “Huike Cloud” platform. This platform is a collection of data models that can be used for data analysis. It can provide customers with a variety of data services, and finally the value of the data samples accumulated over the past 15 years can be realized. This is the moat of Hui Na.
(Editor’s note: At present, the exclusive data sample library on the Huike cloud platform includes: offline passenger flow database, offline customer group tag library, geotag library, mall basic information database, business brand tag library , store POI information library, etc.)
Q2: The semi-annual report shows that the company’s R&D investment has more than doubled in 2019, and what major aspects have progressed?
Zhang Hongjun: The first is the sensor, we are the first members of Huawei’s intelligent ecosystem. Our new generation of sensor products uses Huawei’s chips, we should be the first to apply this chip to the product. And this product is currently in commercial use, we have now used this series of products in part of the collection.
On the other hand, we are investing more in data platforms, including the construction of data platforms and the study of data models. We study how to apply these data to the operation and management of shopping centers, and how to use them in the management of brands. This is our greater investment.
It is worth mentioning that our essence is not to sell hardware products. Our hardware products are for personal use, and the ultimate goal is to provide better data to the data platform. Therefore, we will make special designs for many details according to the needs of the application of the scene, the needs of the data application, such as temperature, humidity, installation location, etc., and the corresponding algorithms will be adjusted accordingly.
Q3: Passenger flow data is largely owned by the mall and the brand side.Sensitive data, how is it handled in terms of data security?
Zhang Hongjun: First, we will sign a confidentiality agreement with each customer who provides data for the cloud platform. As a listed company, customers have certain Trust and recognition. Secondly, we are not using the raw data of each customer. We output the results of some models after processing. The data itself is in the black box, and we have also kept it technically confidential.
Examples, such as our benchmarking service. In the service of benchmarking service, we will not tell the customer the specific indicator data of the target mall, but the output of each indicator of the customer mall is in the same city or in the same business circle (such as the first 80%) .
Q4: What is the level of Huiner compared with overseas counterparts?
Huyu: In fact, our team has done a lot of research in the United States, Europe, and China, and found that Huinan has occupied a commanding height. At this point, China is in the United States and Europe. front. It is very likely that Huiner will intervene in international markets such as the United States and Europe, provide them with some data collection equipment and services, and push similar models to the US and European markets.
This situation may be due to the different stages of the Chinese economy and the US economic development. In the 1970s and 1980s, the United States built large-scale shopping centers. The offline economy was developed and its operations were very regular and organized. Now the United States and Europe are also affected by e-commerce, and offline operations may need to be changed. But for historical reasons, it is not so easy to lay out data acquisition equipment. However, China began to build large-scale shopping malls around 2000. When it was built, it is likely that companies that have already made passenger flow collection equipment have seen this opportunity to enter the market.
Zhang Hongjun: Yes, also because the US e-commerce shock is lagging behind China, Chinese e-commerce’s impact on the offline is coming earlier, the US is now Only when it started, the impact was not as big as China. Moreover, the closed loop of this business is to use the data model to guide every part of the mall operation, investment promotion, etc. It needs an algorithm model, and this model can not be derived from a single data, so the accumulation of data samples is very important.
Q5: What does Huiner Technology think about the future market?
Zhang Hongjun: Our target customer is a shopperField, retail brand. According to statistics, in 2018, the number of newly opened shopping centers in the country exceeded 530, and the number of large shopping centers in the country was nearly 5,000. We currently have more than 1,500 data, and there is still much room for improvement.
Ding Yao: At present, the number of shopping centers is steadily increasing, and passenger traffic is also steadily increasing (Huike Cloud Statistics). After that, the growth rate will definitely slow down, but the increase in stock competition is also certain. On the other hand, because of the intensification of competition, because of the fierce competition, their demand for data to assist their operations has increased.
Q6: What kind of planning is Huinan’s future business direction?
Zhang Hongjun: In the future, we will focus on data mining in the two major application scenarios of “new business” and “new retail”. Such as passenger flow forecasting, trend monitoring, evaluation of marketing activities, and marketing activity decision optimization services. In the future, Huishen will expand its services to every link in the industry chain. However, our whole idea is still about the flow of passengers and people, and also the relationship between people and the field. The location, investment, and operation generated in this direction include asset evaluation, and even future urban planning. And not just malls and shopping centers, like mobile business halls, banks, etc., can use a similar approach. This will be our future direction and market.
Ding Yao:We judge the future Asset trading may also be one of the goals of our application scenario, because The deal in a large shopping mall is a few hundred million, and the dependence on authoritative third-party data is very large. This applicationUS, Europe is very mature, and Singapore is much more mature than China. We judge that as soon as the time is up, the application in this area will be very broad in China.
But we have to figure out the different business routes and the different service targets. For example, the service customers of asset trading may be mainly funds, banks and other financial services institutions. ;The construction and location of the assets of the clients are mainly real estate developers, real estate construction, design, construction units; and commercial operations, passenger flow, investment, promotion,The target of publicity and planning is the business operation of the shopping center and the various brands in the shopping center.
Related reading:
The Age of Shopping Centers How to fight for the battle of passengers?
Hanna Technology: Does the new retail mean subversion or return to the essence?