Original link: https://www.latepost.com/news/dj_detail?id=1749
A year ago, the latest independent business sector in the Ali system, Lingyang Intelligent Service Company was established , with Peng Xinyu, the former head of Alibaba’s data center, as CEO. Lingyang integrates multiple data technology teams such as Ali’s data center, business center, customer service system, and supply chain management system. But its goal is not completely consistent with Ali’s main business – it not only helps merchants do business in Ali’s domain, but also serves the needs of merchants for multi-platform and multi-channel operations.
The establishment of Lingyang occurred between Ali’s implementation of the “operational responsibility system” in 2020, advocating each business to go to the market more independently, and the “1+6+N” major adjustment two months ago. Lingyang began to explore independent operations a year in advance.
In May of this year, Lingyang moved its office to Gaoqiao Yungang, 5 kilometers away from the Xixi Park of Ali headquarters. This is a two-story building transformed from a factory. At the time of starting a business, everyone gathered together instead of being scattered in various buildings in Xixi as before. The team really became a “group of people” and “a meeting can be held with a shout”.
One year after its establishment, Lingyang released a new product, Lingyang One, which integrates different systems of pre-sales marketing, in-sales orders, inventory management, and after-sales customer service in a single product. Not only large customers, but also mid-waist customers provide lighter standardized services.
Lingyang One supports merchants to do business through multiple channels, not limited to Ali, but also almost all mainstream platforms in the industry, such as JD.com, Pinduoduo, Douyin, and Xiaohongshu. Lingyang hopes to provide a more open choice for merchants who seek growth on multiple platforms at the same time.
“LatePost” interviewed Peng Xinyu, CEO of Lingyang, and talked with him about how Lingyang stepped out of Ali’s perspective to realize omni-channel data intelligence services, how data can truly drive growth, and Lingyang’s past experience and new learning.
From “5 Clouds” to Lingyang One, lighter and more standardized products
“Later”: When Lingyang just became independent last year, it launched “5 clouds”, covering pre-sales, sales, after-sales and other business processes. Now it has launched Lingyang One that integrates these capabilities. What is the core difference?
Peng Xinyu: In the original “5 Clouds” product portfolio, there are more customized project delivery, which is more suitable for medium and large customers, because even if we withdraw people, their own technicians can make good use of this system .
Today, Lingyang One can be more targeted at mid-waist customers. It is lighter and cheaper in terms of manpower deployment and product use. The smaller the business, the less it spends. It focuses on data inclusiveness. I hope small and medium-sized businesses can also use it. Up, affordable.
“Later”: Why do decentralized functions first and then unify, instead of launching more integrated products from the beginning?
Peng Xinyu: We wanted to make a product like Lingyang One on the first day. In the past year, using the “5 Clouds” to serve large customers like Mengniu was to verify whether the previous data technology, product capabilities and experience were feasible. Through this polishing process, we know how to abstract and refine common functions to make more standardized products, so now we combine 5 into 1 and become Lingyang One.
“Later”: What kind of products are more standardized?
Peng Xinyu: It turns out that many software companies in China try to standardize their processes, but the processes of sales, production, and scheduling are determined by entrepreneurs. Some people think that approval should be divided into three steps, and some think it should be divided into eight steps. Difficult to standardize; to truly standardize, we should start from the data and extract standardized dimensions and structures, which are controllable just like the TCP/IP protocol.
“Later”: Some small and medium-sized customers may also have some customized needs, how to meet them?
Peng Xinyu: Set the big structure and leave blank. Just like decoration, we will do fine decoration, let you move in with your bags, but will leave room for you to modify, such as changing the process and authority according to the company’s characteristics, but if you insist on moving the location of the toilet, I will remind you, maybe After three to five years, there will be problems with the sewer.
“Later”: Most other companies launch products that meet a single need, such as data analysis and data marketing products. Few companies launch a product that integrates multiple functions like Lingyang One. If this is really a good model, why are there only a few companies? doing?
Peng Xinyu: Can others do this market? Yes, but it is difficult for a start-up company to do it, because the investment is high, and it is difficult to make money and survive in the short term. Most of the platforms will not do cross-platform products and services, because merchants are their advertisers, and they still hope to bring the most business to their own platforms.
“Later”: What is the difference between the products provided by Lingyang and the existing data tools on many e-commerce platforms, including Ali Business Consultant and JD Business Intelligence?
Peng Xinyu: Lingyang One is to solve the needs of cross-channel operation. If a merchant only does business on Alibaba or JD.com, it is enough to use the platform’s native tools.
“Later”: Lingyang put the original five functions into one product today, while Ali itself has been dismantling a middle office into multiple business departments in the past few years. How do you understand this difference?
Peng Xinyu: Ali was undoubtedly right to be a middle-end platform in the past. It brought cost savings in units of “100 million” to Ali and the industry. Later, because the data governance capabilities, membership capabilities, and logistics capabilities of the middle office were all established, the data capabilities had been integrated into the business, and some people were dismantled into the business. It’s like when a house is built, the scaffolding and construction workers will be removed, but the house is still there. This house is the middle platform, not scaffolding and workers.
I hope that one day some customers can complain to me that your Lianyang One has too many functions, which means that the market is large enough and customers like it enough. At that time, a new stage will be reached, and I will have the opportunity to make new changes.
Get out of Ali’s standpoint and think about demand from the perspective of merchants
“Later”: Over the past year or so, as the CEO of Lingyang, not just the person in charge of Ali Data Center, what new understanding do you have about the opportunities in the data intelligence industry?
Peng Xinyu: The general trend is that China has not experienced real IT and dataization, and it will suddenly enter the intelligentization. This has a huge gap and a huge opportunity.
In terms of specific needs, Lingyang One is mainly to meet the omni-channel operation needs of retail and other enterprises, such as brands, that is, to do business in both online and offline channels such as Taobao and Douyin.
At the end of last year, when I visited small and medium-sized businesses in various places, I realized that there were various platforms in the topics that businesses talked about. What should be concerned is their operating problems in multiple channels.
“Later”: What are the specific difficulties in multi-channel operation?
Peng Xinyu: In terms of the process, if you use products from different companies for traffic flow, membership management, inventory management, etc., and the overall performance cannot be completed in the end, the products in each link will play football with each other.
In terms of channels, in many companies, there is competition among departments that sell goods on Taobao and Douyin, and it is impossible to transfer goods across channels and adjust distribution in a more reasonable and timely manner. In the end, the performance of the department may be achieved, but the company’s performance may not be achieved.
“Later”: How do you implement these demand insights into your product design?
Peng Xinyu: For brands and retailers, the three biggest common divisor requirements are: pre-sales marketing, in-sales order management and inventory management, etc., and after-sales customer service, which corresponds to our pre-sales marketing cloud, in-sales Production and sales cloud and after-sales customer service cloud.
Merchants often have a function called store manager or channel manager, who wants to control the overall situation, so we launched the analysis cloud, which is not to analyze reports, but to help store managers make decisions after processing data.
The data governance needs of large enterprises in various industries correspond to the development cloud, and they can use their own technical teams to do secondary development of some data applications on the development cloud.
This is the design of the “5 Clouds” when we officially debuted last year. Lingyang One is a further integration and refinement, and we want to make it easier and more convenient for merchants to complete omni-channel operations.
“Later”: 2B product design, a key is to distinguish what is just the customer’s request, what is the real demand, how do you do it?
Peng Xinyu: Over the past decade or so, Ali Zhongtai has participated in or supported many products, such as data grinding, Taobao index, business consultant, etc., hoping to make it easier for merchants to do business on the Ali platform, which itself has accumulated common needs refinement.
When we talk about needs with customers, there is a principle that there is no right to speak without investigation. Only by understanding the details can we know where there are pain points and problems. Come back to report, if there is data, it is data, if there is no data, it is a case, and if there is no case, it is observation.
“Later”: How do you replicate this service awareness and ability to discern needs to the team?
Peng Xinyu: The first of Lingyang’s values is to require every employee to “be a good person before doing good deeds”. 2B business is not a one-off deal. When someone calls me, no matter how busy I am, I will still have a chat with him. Don’t focus on his current money, but focus on his growth behind him, his work difficulties, and possible opportunities for cooperation in the future.
Another value is to come from customers and go to customers. Don’t force customers to push things, find out their real needs, collect and solve the needs, and then send them over.
“Later”: To meet the needs of merchants for omni-channel operation, it means that you have to analyze the data of Douyin, Pinduoduo, and JD.com in addition to Ali. Will there be any obstacles to obtaining cross-platform data?
Peng Xinyu: What we use is the business data authorized by the merchant to us, and its own business data on various platforms or channels, which is originally the merchant’s. We also have some publicly available data, such as weather forecast data, which can guide the stocking strategies of down jacket merchants.
“Later”: What types of snacks sell well on Douyin, and what cosmetics are most popular on Pinduoduo? Can you provide merchants with the internal trend data of these e-commerce platforms?
Peng Xinyu: If some customers purchase these public domain data reasonably and legally, they can be integrated into the operation console of Lingyang One. We will not crawl the data of JD.com and Douyin by ourselves. Compliance and legality are the bottom line.
Past experience is initial velocity, new learning is acceleration
“Later”: From Ali Data Center to Lingyang, what core data capabilities have you accumulated in the past?
Peng Xinyu: The first is technical ability and experience. For example, “Double Eleven” and “618” data processing at such high-concurrency time points, we can monitor at the second level and give accurate feedback on which channels and regions the orders came from. In this way, when the orders in East China are out of stock, we will adjust the delivery as soon as possible. Strategies divert traffic to the southwest, and merchants can intervene in time. Accurate data and low latency under high concurrency, it is difficult for a team that has never done it before. The performance of these big promotions will affect the business’s annual performance.
Secondly, we have the breadth of the industry, how to do the first launch of the new brand and the new brand of the old brand, how to change the track to make the first launch, etc., we have done it.
Finally, we have done professional data management. For example, for merchants, if they want to analyze why the sales volume has dropped by 5% this year, it may be no problem to just look at the KPIs of each department, because the problem is indeed not in a single department, but in the intersection between departments, in the “urban-rural fringe department”. Data management manages these intersections and ultimately leads to business growth.
“Later”: How to use data management to solve the problem of “intersection” that no one cares about?
Peng Xinyu: The starting point is the standardization of data. Taking the data of “active buyers” as an example, platform A may think that buying once a week is active, while platform B buys three times a week. What exactly is active? To figure out the original definition of the data, and then do the data calculation. Similar to exposure rate, bounce rate and so on.
If it can’t be solved, it will happen that the merchant wants to count the annual active buyers. At the summary meeting, all departments have reached the standard in this dimension, but in the end the company’s revenue KPI has not been completed.
Ali Data Center launched the “One Data” data methodology very early, which is to unify data standards. Only when the standards are unified can we talk about the quality of data and the realization of business based on data.
“Later”: Most of your past experience comes from shelf e-commerce. For new forms such as live e-commerce, is this experience still applicable?
Peng Xinyu: Lingyang One specially designed a function for live streaming this time, which can help the anchor recommend the selling point of the product, and review the highlight moments during the live broadcast.
A single platform such as Douyin is not all content e-commerce. Merchants often compare the performance of a product on Douyin and Xiaohongshu. Wherever it sells best, they will continue to launch. We provide content analysis and content analysis of the entire platform. distribution tools.
And no matter the shelf or live e-commerce, behind it is the understanding of the “people and goods market”, which is not fundamentally different from the e-commerce ten years ago.
“Later”: Most of the above talks are about how to serve retail companies to do full-platform operations. Do you plan to expand to other industries in the future?
Peng Xinyu: One of Lingyang’s business principles is not to do things beyond the scope of competence, not to make money beyond the scope of knowledge, which is responsible to customers.
The scope of our current capabilities, one is data technology and engineering technology capabilities; the other is the experience, contacts and insights we have gained in the retail industry for many years, which supports us to serve customers within the scope of large retail. Today, selling cars and insurance also needs to be launched on Xiaohongshu and Focus. They are not retail companies, but their operations are similar, and our products can also be used.
“Later”: This year you also connected to Alibaba Cloud’s Tongyi large model. How do you make good use of this new capability?
Peng Xinyu: If the customer has a large model, he will call his own after connecting to our product; if not, he will use the large model in our product, which is also market-oriented, and we will accept which large models are useful and cost-effective; In the future, Lingyang may also use the open source ecology to build its own industry vertical models, such as retail models.
“Later”: At present, what needs to be improved in Lingyang’s technology and products?
Peng Xinyu: Some customers’ businesses are not familiar to us. For example, fresh food companies focus on delivery radius and delivery capacity, while the e-commerce we are familiar with focuses on goods and inventory, but we are learning and improving quickly.
“Later”: Experience is an asset, and it may also bring prejudice. How to learn on the basis of experience and expand the boundaries of experience?
Peng Xinyu: The experience of the past ten years is our initial speed, and the acceleration of learning will be faster after that. Just like a large model, only when your knowledge is large enough can intelligence emerge.
We will use A/B test to test repeatedly to ensure that the direction of learning does not deviate, or deviate for too long.
A more neutral Lingyang, from a planned economy to a market economy
“Later”: In the adjustment of Ali this time, Lingyang did not become an “N” company, but a company under Cloud Smart. Were you surprised when you heard about this arrangement? What are you thinking?
Peng Xinyu: Unexpected, but reasonable. But I also immediately thought that this adjustment is to enable each company to have sufficient ability to respond to the market, and to establish its own organization, operation, and decision-making capabilities. This was the case when Lingyang was established last year, and our changes have been very small.
“Later”: How has the cost accounting method changed before and after Lingyang’s independence?
Peng Xinyu: Originally we reported the budget, and now the group will give money, but it is not a budget, but an investment. Now we need to ask for growth and opportunities from the market.
It is equivalent to a planned economy in the past and a market economy now. In the future, as long as you do well enough in the market, external shareholders will definitely come in to support you.
“Later”: Lingyang used part of Alibaba Cloud’s sales team at the beginning of its establishment last year. Will this part of the cost be settled separately with Alibaba Cloud?
Peng Xinyu: At present, we are not just a sales relationship, but a value connection. When Alibaba Cloud’s customers have digital construction needs, Lingyang’s products can become part of the comprehensive solution. This is not the diversion thinking of the past. Diversion alone cannot attract customers.
“Later”: Now you are part of the cloud intelligence group instead of a more independent “N” company, which may mean more resources, but does this identity also bring some restrictions?
Peng Xinyu: The adjustment of the whole group has made Lingyang more neutral. On the night when the news of the split came out, some customers and friends sent me messages, saying that I was worried that you would cooperate better with Taobao and Tmall, but we are doing business on many other e-commerce platforms today, and now you are more neutral. The cooperation between us can be more in-depth in the future.
We are indeed still working on cloud intelligence, but when we serve customers, we don’t necessarily have to deploy on Alibaba Cloud. Some multinational companies use AWS, and we can also deploy on AWS.
“Later”: What is the value of Lingyang to Ali today?
Peng Xinyu: Going back to Ali’s slogan, let the world have no difficult business. Today, many merchants have a lot of scattered data on each platform, which makes it difficult to do business through all channels. Helping them solve this problem is the birth of Lingyang. original intention.
“Later”: From the head of Ali’s department to the CEO of an independent company today, how has your role and positioning changed?
Peng Xinyu: I used to be a problem solver, and I solved problems within a fixed range; now I have to solve problems for myself and for the team.
“Later”: The positive feedback of 2B business is relatively slow. Some things you insist on, such as the value of data standardization, will take time for the market to accept. How to face internal and external doubts in this process and let the team work hard in one direction?
Peng Xinyu: To be number one, I think there are several states. One is temper, “I don’t believe we can’t do this, do it!” The other is courage, “I believe we can do it, everyone go!” The third is to rely on confidence. I have done it, disassembled it, and calculated it carefully. I probably have a little bit of it myself, and I’m not very swayed when someone questions us, because I’ve been through that situation.
When I joined Ali as a data warehouse in 2004, many people felt that data was dispensable. Ten years later, only the part of data governance will save Ali in the billions of costs every year.
The top entrepreneurs we met today no longer need to teach them about data thinking. They are all thinking about how to do better with data, and they will no longer question whether it is useful.
“Later”: What is the core of data thinking? How is an entrepreneur and company with data thinking different from one without data thinking?
Peng Xinyu: Data thinking is data-centric, but not just data and technology, but the comprehensive capabilities of data, technology, and business. It takes business development and growth as the origin of thinking, and the ability to define problems, solve problems, and continue to iterate.
This article is transferred from: https://www.latepost.com/news/dj_detail?id=1749
This site is only for collection, and the copyright belongs to the original author.