MCtalk · CEO Dialogue with Guandata: How Data Guides Both Business Intelligence and Organizational Efficiency | Linear Capital Portfolio
Let data come from the business and be used for the business.
This is episode 9 of MCtalk · CEO Dialogue
The origins of Business Intelligence (BI) can be traced back to the 19th century. Though still a nascent concept at the time, it had already become a powerful tool for commercial success in banking, exemplified by Sir Henry Furnese. It wasn't until the 1990s that Howard Dresner, an analyst at the globally renowned consultancy Gartner, crafted a precise definition: BI is the concepts and methods for improving business decisions through the use of fact-based support systems. Since then, with the continuous evolution of information technology, BI has entered the digital age. Any decision, especially rational business decisions, relies on objective, fact-based insights — and data is the best tool we have in today's digital era for understanding objective reality. When facing outward-facing operational decisions and inward-facing managerial decisions, how can a single dataset serve simultaneously as a forward-looking BI tool and as real-time management metrics for improving organizational efficiency, thereby maximizing value? For this episode of MCtalk · CEO Dialogue, we are honored to welcome Chunyuan Su, Founder and CEO of Guandata, as our guest. Guandata is a domestic one-stop intelligent analytics platform and service provider that has been deeply rooted in the BI field for many years, bringing a unique perspective and profound insights on data. This is an entirely new topic that MCtalk · CEO Dialogue has yet to explore. Coincidentally, NetEase Digital Intelligence's NetEase Shufan, as a technology and service provider focused on full-link data development, governance, and analytics, also has deep expertise and excellent products in the data business. In this episode, we will see an open exchange between Ruan Liang, NetEase Vice President and General Manager of NetEase Digital Intelligence, and Chunyuan Su, Founder and CEO of Guandata — two seasoned veterans in the B2B space sharing their perspectives without reservation.
Key Takeaways
- Chunyuan Su: Data usage needs to "return to its roots" — only by returning to the business can it generate value.
- Ruan Liang: The opportunity cost saved by BI for business decisions deserves to be recognized.
- Ruan Liang: Data powers intelligence, intelligence enables agility, and agility improves fault tolerance.
- Ruan Liang: Data metrics, unlike human management, show no favoritism. In other words, data is objective and fair.
- Chunyuan Su: Letting those who hear the gunfire make decisions is essentially about enabling business teams to use data and make decisions.
- Ruan Liang: Beyond decision guidance, data opens up more perspectives, elevating strategic dimensions.
- Chunyuan Su: Using data is like "prescribing the right medicine": diagnosis is qualitative, prescription is quantitative — diagnose first, then prescribe.
- Chunyuan Su: Building products requires walking on two legs: one "standardized", the other "configurable and composable".
01 Technology and Tools Are Fundamentally About Solving Business Problems
Ruan Liang: Welcome, Mr. Su, to NetEase. Since we're both in the B2B business, we have plenty to discuss, and we have similar products as well. I've been in the internet industry for a long time and consider myself quite sensitive to numbers — whether operational data, product data, or financial data. I know Guandata is a company that does BI very well. So I very much hope to explore data-related topics with you in greater depth — our views on business intelligence, business data, how to use data effectively, and how to build good BI products. There should be many interesting stories here. Please share first, Mr. Su.
Chunyuan Su: Thank you, Mr. Ruan, for this opportunity to discuss B2B and SaaS together. Guandata is a company focused on business intelligence, founded in 2016 — eight years ago. Since our founding, we've worked with many advanced enterprises, including numerous foreign brands like Unilever, Budweiser, McDonald's, as well as more domestic leaders such as ANTA Group, China Merchants Bank, and Genki Forest. Foreign enterprises didn't used to use domestic BI products much, but our cooperation with them has deepened considerably. The main reason is that these global companies' businesses in China have become increasingly localized.
Ruan Liang: That reminds me of a NetEase Digital Intelligence client — Yum China, the parent company of KFC. Yum China became independently listed precisely because it localized so well in China. In the Chinese market, KFC even surpasses McDonald's — a differentiation from foreign markets.
Chunyuan Su: Indeed, many foreign enterprises' operations in China are becoming increasingly different from abroad — in areas like e-commerce consumer perception, omnichannel construction, and so on. All of these require real-time data to support decision-making. Therefore, many foreign enterprise users highly value the "agility" of data-driven decisions, with particular emphasis on mobile BI experience and empowerment, because it allows them to analyze and make decisions anytime, anywhere. The mobile BI we provide can naturally and seamlessly integrate with WeChat, Lark, DingTalk, various internal OA systems, and more — enabling decision-makers and frontline business personnel to view data and make decisions in specific business scenarios through mobile devices.
Extending further, we can see that the value of technology and tools is rooted in the business layer — they are fundamentally about solving business problems. Over the past years of working with these industry giants, we've not only collaborated with their IT departments to build platforms but have also gone deep into their business units — marketing teams, supply chain teams, operations teams, finance teams — to meet the diverse decision-making needs of different business teams, helping them identify business problems and opportunities through data analysis. Our philosophy is captured in a simple phrase: "Let the business use it."
Ruan Liang: I deeply resonate with this point. We also have big data products — though perhaps more focused on data middle-platform solutions. Similar to Guandata, we frequently interact with clients' IT teams. The IT team ultimately makes the purchase, but engaging with IT alone is far from sufficient. Because IT teams also serve the business — or rather, the ultimate value generated must be in the business processes and business teams.
02 Data-Driven Agile Operations, Reducing Enterprise Decision Costs
Chunyuan Su: I'd like to take this opportunity to share some observations about the retail and consumer goods industry, which has been one of our key focus areas in recent years. In recent years, this industry has undergone waves of internet-driven transformation — from the previously popular "new retail" to the current rise of domestic brands and the new guochao (China chic) trend — reflecting the iterative evolution of socioeconomic patterns and consumer preferences.
The biggest change is that the rise of e-commerce over many years has transformed people's lifestyles and shopping habits. Everyone has grown accustomed to online shopping, creating ever more data connections between merchants and consumers. For merchants' business operations, data has become increasingly important. In this environment, we've served numerous retail and consumer industry clients, assisting them in building efficient data analytics systems. This industry is the quintessential example of "business is data, data is business."
Take ANTA, with whom we've partnered for five years. We helped build ANTA's own "business command center" — unlike e-commerce command centers that focus mainly on online business, this covers online and offline, direct sales and other distribution channels, livestreaming, and more — all presented on a unified platform, empowering ANTA to drive agile operations with data.
Ruan Liang: From your perspective, Mr. Su, what specific scenarios does ANTA use Guandata's BI products for? Or which functional points do you think are most valuable to them?
Chunyuan Su: ANTA shares many commonalities with other companies we serve, such as Skechers, McDonald's, and Mixue Ice Cream & Tea — they are all what we'd call "ten-thousand-store chain" brands today, with extensive networks of stores as their nerve endings. For business operators, in the current era, they all face a core challenge: how to enable decision-makers to penetrate through data to gain business insights. Traditionally, decision-makers might review presentation after presentation of PPT reports to understand performance and its drivers. Or they would conduct store visits in person — from VPs to mid-level managers, large numbers would personally inspect various regions to assess frontline operational issues or opportunities. But the pandemic in recent years changed this approach — what do you do when you can't visit stores in person?
During this period, ANTA's digital applications experienced explosive growth. They used data for "virtual inspections," leveraging data to gain insights into performance across different regions, different store formats, various new products, inventory of bestsellers, and more. Problems that might have taken monthly reports to discover in the past could now be identified and fed back in a timely manner on a weekly or even daily basis. After the pandemic, this approach has become solidified as part of ANTA's daily agile operations.
Ruan Liang: That's really excellent. First, efficiency has improved. Second, it saves substantial costs — costs that, from an economics perspective, we call opportunity costs, right? The time saved can be spent on more analysis or decision research.
Chunyuan Su: Exactly. This is also a concept we began promoting last year, and one that the industry is increasingly converging on — it's called "agile operations." In today's complex economic environment, it's becoming increasingly difficult for enterprises to achieve profitability and growth. A critical aspect of profitability and growth is cost control — but what is the biggest cost? Fundamentally, it's the cost of decision-making. The core of agile operations lies here: whoever can achieve breakthroughs in the timeliness and granularity of decision-making, realizing more agile decisions, can identify and solve problems faster, and achieve growth.
Ruan Liang: This is enormous leverage. I wrote an article last year with a point I wanted to make about agility — when we use AI, many business processes become incredibly agile. This brings an additional benefit: it's not just efficiency improvement; even when decisions or execution are wrong, because of agility, the cost of errors becomes smaller too.
Everyone marvels at how fast the world is changing, and we're facing all kinds of uncertainties. Therefore, given so many uncertainties, I believe agility is essential. Regardless of enterprise scale, we must embrace this change and adapt to it agilely. This also brings to mind that after large language models emerged last year — or as the new AI era arrived — we observed that with AI empowerment, agility has become even more feasible, provided of course that organizations actively embrace this mindset.
03 "Let Those Who Hear the Gunfire Make Decisions"
Ruan Liang: There's one incident that left a deep impression on me. We had an excellent employee who once told me: when he was an ordinary programmer, he worked very hard and was quite smart. But at that time, because he wasn't particularly good at self-promotion and perhaps wasn't especially close to his manager, he consistently received only above-average performance ratings. He believed he deserved better, so he communicated with his manager about what he needed to do to achieve excellent performance in the next quarter or second half. However, his manager couldn't give him a clear answer.
Often, this isn't really wrong. Usually, if an employee and manager have a closer relationship, performance ratings tend to be better — provided the work performance is indeed decent, right? While this may be reality, it creates objective unfairness, even if the employee is genuinely excellent. Therefore, we implemented a reform last year: we made engineers' work effectively measurable. We don't measure based on nominal lines of code, but rather on tasks assigned to employees. For example, working on a BI product, today's task might be building a certain front-end feature or other requirements. Based on experience, we assess roughly how much time this requirement should consume for an engineer at this level, how many days, what the effective code volume should be. Based on this, we then use a points system to effectively measure engineers' work output.
Chunyuan Su: Yes, many of our teams — R&D, marketing, customer success — have developed good habits of using data.
Ruan Liang: For example, R&D teams, frontline R&D teams are quite suitable for this, as are delivery teams.
Chunyuan Su: Right, because delivery teams bear numerous responsibilities — requirements communication, system deployment, implementation delivery, training and go-live, and various other types of work. Against this backdrop, quantifying their work into specific standard operating procedures as much as possible can greatly help identify best practices and improve team effectiveness.
Huawei founder Ren Zhengfei once said, "Let those who hear the gunfire make decisions." Everyone agrees with this, but the real difficulty, in my view, is how to give those who hear the gunfire the capability to make decisions? And data is the most important enabler for empowering frontline personnel to actively participate in decision-making. Often, the crux of the problem isn't that frontline employees lack the willingness to be data-driven, but whether they possess the corresponding decision-making capabilities and matching data. When we empower frontline business personnel with the data and analytical capabilities needed for analysis and decision-making, we visibly unleash their initiative, enabling them to confidently use data to improve their work efficiency. For example, like the case you just mentioned — when we manage R&D teams traditionally, we can only look at results over long cycles without timely feedback, but with a data foundation, we can quantify task difficulty, develop reasonable work plans, and enable frontline R&D personnel to complete work more efficiently under data-driven guidance.
Ruan Liang: Indeed. Previously, a relatively outstanding employee who couldn't get fair treatment would most likely leave the company. After our reform, not only do employees receive fair performance evaluations and compensation, but many talented individuals have also been discovered. We have a very young employee who has been the points champion for several consecutive months. If we were to tag him: first, young; second, physically healthy — he works out, very fit; third, extremely smart. Why do I say he's extremely smart? His code volume isn't the highest, and his working hours aren't the longest, but he has the highest points. Because he's exceptionally skilled at using tools, leveraging various AI tools to boost his own efficiency. Therefore, within the same time frame, he ultimately earns more points than others, with more efficient output.
04 Data Opens More Perspectives for Managers, Elevating Strategic Dimensions
Ruan Liang: There's another topic I'd very much like to discuss with you, Mr. Su. From your perspective, you must regularly review various data about your own company. When looking at data, which dimensions do you pay most attention to? And why?
Chunyuan Su: The core revolves around balancing growth and profitability — this has been the central strategic goal for us in recent years. Then following this main thread down, I look at the key process indicators within it. This represents an evolution in our thinking this year. Previously, our company was also quite results-oriented, believing in "brute force miracles," but in the current market environment, pure results orientation no longer works. While ensuring moderate growth, we must also guard against the trap of losses, so our attention must particularly focus on process indicators.
The inherently long-cycle nature of B2B means that while pursuing annual targets and final outcomes, we must also follow this main thread to pay attention to process indicators in shorter cycles — quarterly, monthly, weekly. For us, the most important category of process indicators is "ICP," or Ideal Customer Profile — referring to the ideal customer segments that align with our core competitive advantages and market positioning. In today's market environment, to achieve our strategic goals of profitability and growth, the most important tactic isn't "casting a wide net" but "precise strikes." To this end, we've carefully selected about 10 IP focus segments internally, striving for intensive cultivation in these advantage segments.
For example, assuming I have 10,000 potential opportunities this year, we would lean toward 2,000 high-quality leads from our ICP domains, as these yield the highest conversion efficiency from a process indicator standpoint. Then we go further, ensuring frontline teams' weekly process indicators continuously strive to capture these high-value opportunities from our ICP focus segments.
Behind these initiatives are actually some bitter lessons from our past, but also an internal process of continuous iteration and evolution — so in retrospect, it's actually a good thing, which I can share with everyone.
In an earlier quarter, one of our teams had very poor performance. Without data analysis, the intuitive conclusion would have been insufficient opportunity volume. But after truly dissecting the data and analyzing it, we discovered that although opportunities appeared to have shrunk, the volume in our ICP advantage segments actually hadn't decreased. The real crux causing poor performance lay in problems within the conversion funnel, as well as our pricing strategy during the conversion phase.
Because this year's overall macro environment hasn't been great, there have been some price wars, but internally we have certain threshold points — at certain discount red lines, we instinctively abandon certain potential clients. This is actually normal; this team was following our internal policy, and at certain discount boundaries, frontline teams have the authority to autonomously decide whether to proceed or retreat.
Ruan Liang: The organization has rules in place, and gross margin must be protected first. This is data guiding decisions at the strategic dimension.
Chunyuan Su: Yes, when we established this policy last year it was correct, and because profitability is an important strategic goal this year, gross margin is naturally a key parameter in decision-making. But when we penetrated through the data layer by layer and discovered this issue — or rather, when I as the number-one leader looked at these real data — I gained a better understanding of what state the business was in, which links had problems, and how to implement targeted solutions.
The process of data analysis prompts us to continuously fine-tune strategies. Take the pricing strategy example I just mentioned — getting drawn into price wars often stems from our value proposition not being sharp enough. So facing the temptation of price wars, we took a different path. For opportunities with huge potential, we have our best solution experts, experienced delivery team leaders, and company leadership engage in effective communication with the other party to strengthen value-based partnerships. Through this shift, we move beyond relying purely on price competition, winning client favor through professional strength and personalized service.
The core of this series of actions remains the pursuit of growth and profitability — the end goal is clear. But in the decision-making process, we need to continuously make trade-offs, identifying and optimizing process indicators to shorten decision cycles. For example, we identify key opportunities weekly, and when opportunities reach critical turning points, we mobilize all available resources to form joint forces, doing our utmost to maximize our grasp of every opportunity.
Ruan Liang: When I review data myself, I often have similar realizations to yours, Mr. Su. Recently, I've been contemplating and attempting some reform initiatives that I'd like to share with you. For example, when we serve a client, there are various stages — opportunity stage, sales stage, pre-sales stage, implementation and delivery stage after contract signing, and finally customer success stage. We look separately at the costs incurred at each stage — for instance, how many person-days delivery consumes, then calculate costs.
From a management accounting perspective, there's a method called activity-based costing. Previously, our most common way of evaluating the efficiency of a certain type of client, industry, or process was using financial accounting methods: how much time did I spend, how much money did I spend, and comprehensively calculating this client's gross margin. But from a management accounting perspective, we don't necessarily have to calculate purely from a financial accounting angle of how much money and time. Instead, we can evaluate how many key actions we performed in the process of serving a client, with actions corresponding to processes.
For example, in the implementation and delivery stage, there might be several key actions. First is deployment, second is pre-training the knowledge base, then some key data governance actions. We define standards for these actions — what is the cost each time we perform this action at this company? Then by applying these activities to this category of client, we can calculate the true management accounting-level gross margin or efficiency for this client category. Finally, by doing variance analysis with other client categories, we can identify the true advantage clients and advantage scenarios. Because from a financial accounting perspective, Client A is profitable and Client B is not, but from a management accounting perspective, it's possible that Client B is actually more efficient.
05 Building Standardized and Configurable Products: "Walking on Two Legs"
Ruan Liang: When business develops to a certain stage, our standardized products are based on the logic of meeting most clients' needs. Due to so-called growth pressures, or the need to create benchmark effects, we also make certain compromises to do some special customization with major clients — this is unavoidable to some degree.
We strictly control the proportion of customization. But this remains a long-standing vexing problem. Frontline teams need to win major clients, and major clients want functional customization on top of standardized products — features that most other clients may not use. Ultimately, whether to accommodate major clients' needs becomes a question of trade-offs. What's your view on this, Mr. Su?
Chunyuan Su: I think our path has been somewhat different. As a startup going from 0 to 1, we are very firmly committed to building standardized products. Why did we choose standardized products as a startup? It relates to our team's DNA and convictions. First, our core founding team has spent many years building standardized products, witnessing the entire journey from concept to market validation, so we firmly believe this path is viable. A product we previously created served one-third of the Fortune 500 — using a standardized product. Second, from a commercialization perspective, while there may be short-term difficulties, we firmly believe in the enormous potential of standardized products in the long run. Good BI products are fundamentally about continuously distilling and abstracting best decision-making practices — they have universal value and clear long-term commercial prospects.
In the process of building standardized products, we've constantly faced many trade-offs. To summarize, I believe there are two key points to building good products: first, we continuously adhere to the fundamentals and achieve product standardization; second, something we'll continue strengthening in the coming years — configurability and composability.
Configurability is a direction we identified from accumulated experience over many years. For instance, BI needs to connect to various new data sources, many of which may not have existed before. Through years of practice, we've continuously distilled from leading client needs to develop an effective configurable solution — whether for connecting different data sources or for diverse configuration options within the same database, such as customizable data access frequency settings. We meet unique needs across different industries and application scenarios by designing multi-layered configuration options.
Take banking clients as an example — they have particularly high requirements for system performance and reliability. Considering senior management's daily dependence on timely data presentation for decision-making, robust information barriers are needed to effectively insulate the availability of critical data streams from frontline operations. This process involves complex engine scheduling and resource isolation strategies, but we've made it into intuitive, easy-to-understand configurable options to meet client needs.
We've been very committed to configurability in recent years, which entails substantial short-term investment but has also made our product more vibrant and powerful. Take a Fortune 500 commercial bank we work with — internally, over 60,000 people actively use the Guandata BI platform, making it one of the largest BI projects domestically. With such a massive user base, there was intense debate about whether to adopt customized services. But fortunately, our understanding of BI aligned very well with the client's, and their BI application was quite sophisticated. We jointly determined that we must adhere to the main thread of standardized products — overly customized products lack vitality, are difficult to replicate and scale, and weaken the product's intrinsic energy and market appeal.
Ruan Liang: I very much agree. If we go down the customization branch, based on our conversations with many major clients, the clients themselves are also aware of this: if they use our specially customized version, there may be usage failures, various problems, or even very serious issues, and when updates come, they can't be updated in a timely manner.
It's been my good fortune to invite Mr. Su from Guandata to NetEase at this point in September. Today we've discussed numerous topics — related to data, management, R&D, sales, and finally product customization. I've personally benefited greatly and received much inspiration. I believe this has been a wonderful exchange of ideas.
Chunyuan Su: I'm also very grateful for Mr. Ruan's invitation and sharing. As fellow entrepreneurs in the B2B space, we both hope to make China's SaaS market and enterprise services market better, and to make the world a better place.
About Linear Capital
Linear Capital is an early-stage investment institution focused on "frontier technology + industry" — that is, frontier technologies represented by data intelligence, digital new infrastructure, next-generation robotics, and technological transformation in traditional fields (such as biomedicine, materials, energy, etc.), applied across vertical industries to substantially improve industrial efficiency, empower them to solve pain points, and complete industrial upgrading — achieving excess returns through substantial increases in industrial value. Currently managing ten funds with total AUM of approximately $2 billion.
Our investment stage focuses primarily on leading angel to Series A rounds, with individual investments ranging from $1 million to $10 million (or RMB equivalent).
To date, we have invested in over 120 entrepreneurial teams at early stage, including Horizon Robotics, Kujiale, Sensors Data, Tezign, Rokid, Guandata, Agile Robots, and others. The combined valuation of Linear's portfolio companies is approximately $20 billion.
In the near term, Linear Capital is striving to become the best "Data Intelligence Technology Fund," and in the long term, gradually build itself into the most influential "Frontier Technology Application Fund."