Author: Zhengfeng
来 源:正和岛(ID:zhenghedao)
"The company's transformation from good to excellent has nothing to do with whether the industry it is engaged in is in the trend. Many of the companies that have made the leap are not affiliated with booming industries, and some are even in very bad situations.
Excellence is not a product of circumstance, it is largely the result of deliberate decision-making. ”
– Jim Collins
In the business world, if you ask the successful entrepreneurs and managers, what is the secret of their business or the way to success? You'll hear all sorts of answers, such as efficiency, self-discipline, vision, strategy, pattern, and so on.
However, judging from the companies I have contacted and the business phenomena I have observed, there is another word that must not be ignored, and that is decision-making. Taking the above question as an example, even if entrepreneurs and managers have different answers, it has to be said that decision-making is indeed a high-frequency word in their mouths.
Why Decision? What exactly does the decision make? How can companies really make good decisions?
Before answering these questions, let's look at this story about decision-making.
1. Decision-making determines the trajectory of destiny
If you're sitting in your office right now, you can check to see if the company's photocopiers have a logo called Xerox on them.
As the world's largest manufacturer of digital and information technology products, Xerox has always been a hegemon in the global photocopier market, maintaining the first position in market share all year round.
In 1959, Xerox built the world's first floor-standing Xerox 914 fully automatic copier, which opened a new page in the history of office copiers around the world.
From the '60s to the early '70s, the copier sold more than 200,000 units, made a profit of more than a billion dollars, and was named "the most successful product in American history" by Fortune magazine. Xerox thus became the first company in history to make more than $1 billion in 10 years from just one technology (the second being Apple).
This performance led its then-president, Joseph Wilson, to say, "If this rate of growth can be sustained for decades, Xerox's sales will exceed the gross domestic product of the United States." ”
Its glorious history can be seen.
But frankly speaking, when talking about Xerox now, it is already a person who knows more about Xerox if it is a company that produces printers and copier equipment, and its popularity and brand influence cannot be compared with Apple, and the market value of the two is more than 1,000 times worse.
Judging from the development history of Xerox, the word regret is an evaluation of many people. Why? Because it's almost on par with a mega-company like Apple or Microsoft, and that "point" of difference is decision-making.
In the '70s, Xerox's laboratories gave birth to the world's first graphical interface technology, which is the basis for the Windows operating system that we are all familiar with today.
But unexpectedly, Xerox easily abandoned this "what you see is what you get" graphical interface technology and continued to cling to its own photocopier products. Why is there such a decision? Xerox's reasoning was "simple" - the technology was not directly related to the copier, which was the core business at the time.
The rest of the story is more familiar to us, Bill Gates and Steve Jobs used the graphical interface technology abandoned by Xerox to develop Ethernet and the first generation of graphical user interfaces, which are today's Windows and macOS systems, and ushered in a new era of personal computer revolution.
Therefore, when recalling that period of history, Bill Gates once said with emotion:
"I know I'm not the only one who sees Xerox's mistakes in decision-making, but I'm determined to avoid them at Microsoft. When it comes to the opportunities that research on computer vision and speech recognition might create, I've always tried to make sure that we look at the big picture. ”
Going back to the opening question, why has decision-making become a high-frequency word in the mouths of entrepreneurs and managers? I think Xerox's case would be a good answer. It is no exaggeration to say that the correctness of a decision directly determines the fate of an enterprise, and success or failure is often between this thought.
Of course, today, companies may no longer be as arbitrary in their decision-making as Xerox.
The reason is very simple, at the moment when the stock competition is intensifying, the downward pressure on the economy is increasing, and "living a tight life" has become a consensus, no matter which industry is doing business, the four words of reducing costs and increasing efficiency are a consensus on decision-making, or the premise of decision-making.
And judging from the trends in the corporate world, players from all walks of life are clearly aware of this.
2. Data-driven, data consumption first
In recent years, whether it is communicating with entrepreneurs or in major enterprise forums, there is a word that has frequently come to my ears, what is it? Digital transformation.
The so-called digital transformation, simply understood, is the use of various digital technologies to upgrade or even reconstruct traditional businesses, processes and models, so that enterprises can make decisions in a more timely and scientific manner.
To put it more figuratively, those enterprises that have successfully carried out digital transformation are like equipping themselves with a reconnaissance aircraft that can fly over the market, which can observe the market situation in real time, and collect various data in a timely manner and feed back to enterprise managers to guide them to make effective market decisions.
Why did digital transformation go from an option to a necessity? The answer lies in this. Especially in today's digital economy era, if you want to make decisions that can truly reduce costs and increase efficiency, it is no longer human intuition and experience, but flexible, accurate and efficient decision-making brought by digitalization.
At present, digital transformation is mainly done in three aspects, one is to standardize the work process, the second is to refine management, and the third is to be data-driven.
Among them, the first two can be better achieved through enterprise management systems such as SaaS and CRM. But it must be admitted that data-driven has always been a problem, and the reason behind it is not complicated, that is, a piece of data is collected, and even classified and sorted out in front of the enterprise, but what about after that? Can this data help a specific business? Can it provide a certain reference for the company's next layout?
A reality is that many companies have spent a lot of effort to collect and sort out the data, and business personnel and managers have indeed read it and made some reviews, but then these data have been sealed in an Excel sheet or PPT, and they can never "see the light of day" again.
As a result, data-driven work has become a mere formality, falling into a strange circle of data for data's sake, and failing to improve management efficiency and scientific decision-making. To put it more bluntly, in this way, not to mention helping to reduce costs and increase efficiency, it is the best result not to increase costs and reduce efficiency.
But let's be honest, these problems faced by data-driven can't all be blamed on employees at all levels, because people really don't know how to use this data and how to empower it to bring growth to the business, after all, data doesn't speak for itself. Moreover, each department has its own data, and it is even more difficult for the company to break down departmental barriers and integrate these data for analysis.
So there's really no solution to the data-driven problem? Won't data help companies make decisions that reduce costs and increase efficiency?
In fact, if you want to be data-driven, you must solve a core problem, which is data consumption.
What is data consumption? To put it simply, it is necessary to make full use of all kinds of data within the enterprise, so that the enterprise data flow can fully flow into the business flow and enhance the development momentum of the business.
When business development is reduced and efficiency is increased due to data empowerment, the frequency and scope of data consumption will be expanded, and the investment of enterprises in the construction of data assets will also increase, and the data supply capacity will become more and more sufficient, which will eventually constitute the data flywheel of enterprises, that is, data assets and business applications will promote each other.
More on data consumption and the data flywheel. Here, let's take a look at a few companies that have done a good job in being data-driven, and they have played a good role in the core driving force of data consumption, which has not only overcome the data challenges faced in the past, but also achieved visible results in reducing costs and increasing efficiency.
3. What did they do right to reduce costs and increase efficiency?
Let's take a look at Spark Thinking first, in terms of data, Spark Thinking has been facing two major problems before, one is that internal employees do not use the company's self-built BI system frequently, and the efficiency of data use is limited.
For example, in order to pursue a 100% completion rate of make-up tickets, Spark Thinking has added an additional management intervention process to the platform process. This requires the middle office to download an Excel list from the BI system, and then from the region to the team to the sub-team and front-line managers, personnel at all levels need to disassemble the list step by step. In this way, a lot of the employee's time is consumed on the task of splitting and distributing the table, becoming a "cousin".
Second, in the teaching process, Spark Thinking will generate a large amount of learning data, such as students' learning and class experience, but it is difficult for front-line business personnel such as tutors and operations to directly use these data.
In the past few years, Huahuo Thinking has strengthened the accuracy and richness of learning data through the introduction of AI, and integrated students' learning and practice results in different scenarios into a learning effect indicator - learning health level, which is used to guide the design of teaching content, teacher training, and communication with parents.
However, the problem is that these learning data are mainly used in the background for strategy configuration and product development, and it is difficult for the front-end business team to use them in daily teaching. In other words, front-line teaching staff who can hear the fire are often unable to make timely and accurate teaching adjustments based on this data.
However, judging from the latest news from Spark Thinking, these two major problems have been solved.
According to Zhang Junying, vice president of Spark Thinking Technology, at present, the number of monthly active users of Spark Thinking data system has reached 800, which is twice as many as in the past, and many front-line business managers have become daily consumers of data assets. In addition, the health of the learning situation can also be directly displayed to the teaching manager through the data dashboard, so that they can polish the curriculum and carry out mutual learning between teams based on the learning data. Thanks to this, the user retention rate of Spark Thinking has also been significantly improved.
So, how does Spark Thinking do it? An important reason for this is the help of Volcano Engine's data flywheel's series of landing products and services.
For example, with the help of intelligent data insight DataWind's visualization and intelligent analysis functions, Spark Thinking is more convenient in the construction of data dashboards, and employees at all levels can directly see and download the data they need from the data dashboard, and make the next refined operation actions based on these data.
Taking the problem of splitting the list table we just mentioned as an example, with the help of DataWind, the front-line business personnel of Spark Thinking can directly see the detailed reminders of the make-up lessons that their team needs from the data kanban, and no longer need to stare at the Excel sheet to split it step by step. This allows them to devote more time to teaching.
It can be said that with the help of the product of the data flywheel model, Spark Thinking not only improves the quality of teaching, but also reduces the cost of data construction and data training for business personnel.
Zhang Junying also explained this very thoroughly: "The number of our data dashboards and visualizations has been greatly increased without increasing the number of full-time BI and analysts.
This means that when a data tool is well used, people at all levels will be more interested in looking at the data, and the data is a process of exploration, and they can not only look at it, but also adjust the data. When new possibilities are discovered, they will look at and adjust the data again, so that there is a multiplier effect, so my understanding is that the cost is actually brought by the efficiency multiplication. ”
In addition, in terms of the health of the academic situation, the overall health of the current Spark Thinking has increased by 5 percentage points compared with the past, which is also closely related to the visual business kanban brought by the Volcano Engine Data Flywheel landing product.
Through this product, Spark Thinking can directly display multi-dimensional learning data to teaching managers through business kanban, so that each business team can see whether their courses have really promoted the improvement of learning health through a piece of data, and the middle and senior management teams can also let each sub-team learn from each other's successful experience and promote the improvement of the overall teaching quality.
DataWind shows the growing trend of academic health
With the successful experience of Data Flywheel in the data viewing scenario, Spark Thinking also introduced the growth analysis DataFinder and A/B testing DataTester system under VeDI, a digital intelligence platform of Volcano Engine, with its dynamic diversion and real-time data monitoring capabilities, Spark Thinking has also made a qualitative leap in efficiency when exploring some new businesses and new scenarios.
In the exploration of a new business, Spark Thinking tried to let the employees of the business perform the whole process of self-service operation on the homepage, instead of the past telesales, but what was the effect? This requires rounds of testing. In the past, it took about two or three weeks or even longer to go from sample statistics to analysis to results.
With the support of DataFinder and DataTester, the business team can directly monitor and analyze the progress of experimental data and the differences in user behavior paths in the background, even if you can't write code, you can still clearly see various data indicators through charts, and finally get the result that the success rate of new model registration has increased by nearly 30%, and this process took less than a month.
Another example is Deppon Express, in the exchange with Zhou Yu, the digital marketing director of Deppon Express, he used the word "black box" to describe the past user data status.
"The first-line region has a better understanding of regional users, but it stays in the processing and use process of traditional big data. Each marketing campaign needs to go through a long cycle, and the timeliness of data use is very lagging behind. The headquarters does not have a real-time understanding of the overall user data, and is in a black box state, only getting a non-real-time information when it needs to be used. ”
The reason behind this is that the express delivery industry involves a large number of users, and the role of each user is also different, there are large customers, direct customers, and offline stores, and the role and status of users have changed greatly. How to effectively identify the user portraits of these customers and timely and quickly target marketing has always been an industry problem.
Fortunately, after the introduction of the series of products of the Volcano Engine Data Flywheel Mode, Deppon Express has significantly improved its user identification and marketing efficiency, not only building a complete user graph and achieving unified management of all users, but also doubling the growth of monthly activity, and the number of users placing orders has increased by 13% year-on-year.
How did Deppon Express solve the previous data problem? The secret lies in the digital intelligence product customer data platform VeCDP and the growth marketing platform GMP that help the data flywheel model land.
Specifically, VeCDP's data integration and labeling system can present user data in each region in a visual chart, connect the IDs of each channel, and accurately identify the identity characteristics behind each ID. In this way, who the user is and which region he comes from, Deppon Express can see it at a glance on the screen.
GMP sends marketing information to target users in a timely and accurate manner through functions such as target group selection and sending timing selection, and truly achieves targeted marketing actions.
Zhou Yu also felt deeply about this: "Before cooperating with Volcano Engine, we could only carry out 3-5 activities a month, because it took a lot of time to gain insight into data and circle marketing users, and the marketing timeliness was poor. After using GMP, the efficiency has increased by about 5 or 6 times, and at the peak, about 100 marketing campaigns can be implemented a month. ”
Since the establishment of the Digital Transformation Department in 2014, BSH has strengthened its investment in digital infrastructure, not only building a data lake from top to bottom to achieve the integration, transformation and query of raw data, but also setting up a data middle platform to improve the efficiency of internal response to data needs.
However, with the increase in data and crowd size, BSH has gradually encountered bottlenecks in terms of data efficiency. Li Yifan, Director of New Business Development at BSH, said in a previous interview that BSH still relies more on the "semi-automatic, semi-manual" mode in many business links that emphasize data-driven, which is not efficient enough in terms of specific implementation.
Taking A/B testing as an example, BSH previously developed two versions through a mini program, and then manually circled people, marked, and then manually analyzed the data, which took a month. This is clearly not efficient and flexible enough in terms of the use of data.
After cooperating with VeDI, the team introduced a number of data products such as VeCDP, DataFinder, DataTester, and GMP, which not only greatly improved the efficiency of data use, achieved more refined crowd operations, but also expanded the boundaries of its original capabilities.
Taking A/B testing as an example, with the empowerment of VeDI, the digital intelligence platform of Volcano Engine, BSH A/B testing not only improves efficiency, but also applies it to multiple basic operation links. For example, after applying DataTester, BSH started A/B testing for push copy in this scenario, and after optimizing the copy, the open rate increased by 23%.
In addition, through the combined use of VeDI, the digital intelligence platform of the Volcano Engine, BSH has achieved an optimization and growth breakthrough in the existing business.
For example, by using the combination of DataFinder + A/B to test the DataTester, BSH saw that users were more interested in cleaning products and cleaning services, so it launched a combination of these two product solutions, which significantly increased the click-through rate and conversion rate of BSH Mini Program operation products.
It is not difficult to find that in terms of solving data problems and promoting cost reduction and efficiency increase, the above-mentioned companies have close cooperation with VeDI, the digital intelligence platform of Volcano Engine, and have achieved remarkable results with the help of its series of data products.
Therefore, a key step in doing a good job of data-driven is to use good tools and the power of technology to solve the problems of large, complex, and difficult transformation of data.
Obviously, VeDI, the digital intelligence platform of Volcano Engine, is a good helper for major enterprises in terms of data-driven.
Fourth, the "data flywheel", the right medicine
It is worth mentioning that the related products and services of VeDI, the digital intelligence platform of Volcano Engine, mentioned above, are actually included in the digital intelligence development model of Volcano Engine Data Flywheel.
Before we explain the data flywheel in detail, we might as well understand the "flywheel effect" proposed by management guru Jim Collins:
"In the transformation of an enterprise from good to excellent, there is no single decisive initiative, no amazing innovation, no lucky mutation, and no miracle moment. Instead, the whole process is like constantly pushing a huge, heavy flywheel to turn. ”
In a nutshell, the flywheel effect refers to a company's need to find a sustainable and virtuous cycle of business operations.
This mode is like a heavy flywheel, and when it starts to push, it will be very laborious. But through continuous efforts, the momentum of this flywheel will become greater and faster, and it will eventually become an unstoppable, strong and efficient business model. The data flywheel is actually to produce such a flywheel effect through data.
On the whole, it is composed of two parts: the upper business application wheel and the lower data asset wheel, with data consumption as the core driving force, which not only allows the business side to use data, but also relies on data to make scientific decisions and agile responses. At the same time, with the increasing frequency of data consumption, the data flywheel will further promote the construction of enterprise data assets. (It is not difficult to understand that when the business side truly feels the empowerment brought by data, enterprises will increase their investment in data assets, and the content, quality, and precipitation speed of data assets will be improved.) ）
Eventually, the upper and lower flywheels form a positive cycle driven by data consumption, and the internal data and business can also interact in both directions.
It is important to emphasize that data consumption is always a core part of the data flywheel. As we said earlier, data consumption is about making data useful, or making data "live". This also makes the data flywheel very different from the previous data middle office concept.
The middle office is more about building data assets, doing a good job of data integration and presenting these intuitively and clearly, but the problem is that the pain points faced by the business layer cannot be directly answered from these data.
The data flywheel, with data consumption as the core, is more like prescribing the right medicine, starting from the business pain points, seeing what problems or needs the business has, what role data can play in it, and then giving specific products and solutions. This enables the data flywheel to solve the practical problems faced by the business very effectively and systematically, and ultimately achieve cost reduction and efficiency increase for the enterprise.
Fifth, directly hit the business pain points, data can also speak
In the communication with the technical leaders of the above-mentioned enterprises, I can feel that they have a deep understanding of this.
For example, Zhou Yu summed up this point in three words:
"The Volcano Engine Data Flywheel is a bit like a holistic approach that solves the pain points of your business first, and then empowers the product. For example, the whole system of VeCDP+GMP+A/B testing DataTester+N is gradually being applied, and the integrity of this product is eye-catching, from data collection to construction to operation, it is a whole system. ”
Zhang Junying also talked about the business pain point of "data island". In the past, each business flow and business unit of Spark Thinking had its own data, but these data could not be interconnected, and the crux of some problems was difficult to analyze in combination with data.
For example, the data of the teaching team and the data of the sales team are in two systems, if the user retention rate of a course is low, is it the teacher's teaching problem or the sales channel problem? You can't make a judgment based on the data.
Now, with the help of the Volcano Engine data flywheel, the data of the two teams can be correlated for in-depth analysis, and it is clear which link went wrong in front of you.
It is worth mentioning that Zhang Junying also mentioned a key point, that is, in addition to the bottom-up changes, the Volcano Engine Data Flywheel also provides a top-down perspective.
What does that mean? For enterprise managers, when these data resources are precipitated, they can make certain assumptions based on the data to scientifically and reasonably predict the future, and then make decisions that can affect the future.
Zhang Junying gave a visual example:
"For example, if I want to predict what the company's business volume will look like in 3 or 5 years, one logic is to make some assumptions about each process and each link from front to back according to the company's business flow, but the final calculation result of this method will have a big error with reality.
Another logic is to look at the business as a whole, with growth and attrition, and turn it into a pool model, on the basis of which the assumptions are often more accurate. But the problem here is that you need to turn the horizontal business process into a vertical overall model, and you need a lot of data to analyze from many angles.
With this (Volcano Engine Data Flywheel), it is very easy for me to think in different dimensions and perspectives, and when I go up to aggregation and analysis, the platform also provides a large number of analysis tools, such as table calculations, etc., to help us predict the future through data. ”
In addition, with the support of large model technology, data consumption is not as complex as imagined, even non-professional data operation and maintenance personnel who can't understand the code can retrieve and analyze data in a timely, convenient and fast manner, which greatly reduces the threshold of data consumption and truly penetrates data consumption into all levels of the enterprise.
In this regard, Zhang Junying's "very good-looking" feeling is also a very accurate description.
In his view, a very important point of To B products is to do a good job in user experience, to reduce the difficulty of user use, the more users can fool self-learning, the fewer problems they will face in use, and the more they can realize the self-service data consumption of employees at all levels.
As he said, "User experience is the barrier to entry for business adoption, and if colleagues are not willing to continue to explore, they can only become passive consumers of data, not active producers of data." ”
Therefore, if you want to summarize in one sentence why Volcano Engine Data Flywheel can help enterprises do a good job of data-driven, I think the core point is that it flexibly uses fixed data through frequent and convenient data consumption, so that these precipitated data can really "live".
To put it more figuratively, it can let the data speak, tell managers and business layers what the data means, what changes can be made accordingly, and give corresponding solutions, and ultimately promote enterprises to build data-driven business capabilities, and make scientific and effective correct decisions at the market and business level.
Come to think of it, this is also just as the co-founder & CEO of APP said (since 2020, the company has started to cooperate with Volcano Engine, and the whole company has established that the data system should be treated as a "hard currency" for dialogue and consensus):
"In the company group, there will be different data dashboards, the purpose is to let people 'look at data' instead of 'looking for data'. The purpose of looking at the data is to provoke discussion within the company and achieve consensus.
Unlike many companies, we do not take a lofty approach, but only send the data effect with a long picture, which can be viewed directly with the picture, and turn the data into a consumption set, so that people who are completely unprofessional and unaccustomed to using data can also use data with a very low threshold, instead of using more lofty systems to block these people from using data.
In the past two years of working with Volcano Engine, the data-based consensus has been greatly improved, from awareness to behavior, to the way they discuss it when they meet with each other. ”
6. Conclusion: May the "beauty of data" illuminate the lights of thousands of homes
Finally, I would like to share a particularly favorite quote, which is a motto left by Mr. Drucker, the "father of modern management", which is also considered to be the best description of "the beauty of business":
"Some people think that a business should be a money-making machine. For example, if a company makes shoes, all the people will have no interest in shoes, they think that money is real, but in fact, shoes are real, and profit is only the result. ”
The moral behind this sentence is to ignore the pure business purpose, retain a reverence for customer value, and on top of this awe, with your own ingenuity as an offering, pour your life.
In the same way, judging from the above enterprise cases, data itself is not the ultimate goal, the key is to be able to use it, solve business pain points, and help enterprises make scientific decisions to reduce costs and increase efficiency.
In other words, the value of business growth, efficiency improvement, and scientific decision-making brought by data is real, and data itself is just a result, which we might as well define as "the beauty of data".
Obviously, the Volcano Engine Data Flywheel interprets such a "beauty of data" for us. We also sincerely hope that more and more enterprises can feel the "beauty of data" and achieve scientific decision-making, reduce costs and increase efficiency with the help of the Volcano Engine data flywheel.
In this wave of digital transformation sweeping the world, I hope that all enterprises will make good use of tools to win the competition, and may this "beauty of data" illuminate the lights of thousands of homes.
References:[1]. "BSH's "Intelligent Manufacturing" Leap: Seeing Numbers, Seeing Wisdom, Seeing People" Finance and economics are unscrupulous
Typography | Edited by Shen Wangwang | Zhengfeng Rotating Editor-in-Chief | Xia Kun