Big data in China

Chapter 26 Big Data and Management Change (1)
Data Speaks——More Rational Decision-Making
Most of the human resource management methods commonly used today have many problems, such as too much subjectivity, judging based on experience alone, information asymmetry and opacity, etc. This artificial management method tends to examine the personal capabilities of HR, and big data. Appearance, it provides a powerful solution to HR headaches.

The IBM company in the United States is a model for applying big data technology to human resource management.As a multinational company, it maximizes the advantages of big data technology. The ProfessionalMarketplace database it creates contains testable information such as the skills, salaries, and recent schedules of IBM employees, and obtains a resource allocation through mathematical operations. the best way.

Obviously, this is of great help to people in building a team, especially when project managers need to build an all-round team, they can refer to the ProfessionalMarketplace database to find the most suitable employees.It's as easy as picking your favorite seat from a pile of tickets.

☆Talent assessment

Can talent be counted?of course can.In the era of big data and cloud computing, there is almost nothing that cannot be quantified and calculated. As a resource, talent can naturally be part of data and calculation.Moreover, this data-based talent assessment method can help HR find the required talents more easily.

In the final analysis, talent management is a part of resource management, which can realize quantitative and digital management in essence.For example, financial budget management is in charge of a company's economy and formulates the company's financial plan; talent management is the company's plan for talents. Through quantitative methods, it evaluates and takes inventory of the company's talents, so as to select target talents for training, training, Upgrade, and finally fully meet the requirements of the enterprise.

In an enterprise, is the existing talent situation in line with the industry requirements, where are the gaps, what are the advantages and disadvantages, is it a technical gap or a quality difference, etc., what is the analysis of these?Data, of course.If there is no large amount of data analysis, there will inevitably be huge errors based on HR's observation and intuitive judgment, and the scientific nature of talent management will not exist.

Especially when an enterprise develops to a certain scale, the human resource management of the enterprise needs the support of scientific data. The evaluation of talents usually involves many aspects, such as talent performance data, interpersonal relationship evaluation data, technical quality data, etc. If an enterprise has a database, it can draw conclusions by comparing the data, and can confirm based on which talents the company currently lacks, which areas employees generally lack, and which talents are really suitable for the future development of the company, etc.

In the era of big data, human resource management gets rid of the "confused" and "fuzzy" of traditional talent management in the past

mode, and then enter a new era of all-round digital management.As the largest provider of talent management and assessment solutions in China, Beisen has entered the field of big data and cloud computing after several changes.Beisen talent management platform began to adopt a new model to provide services to customers, such as the SaaS model, which uses cloud computing technology to build the platform.

This is a brand-new field full of unlimited prospects and potential, and it will also bring a more open and innovative talent management thinking to HR.

☆Performance appraisal
I think that there is no need to do any advertising for big data now, and the CEOs and CFOs of major companies are also very clear about this: big data is the development trend of this era, and it will become a stranglehold in the near future. The key to the survival of an enterprise's throat is a magic weapon for an enterprise to win.Because big data breaks through the IT gap between quantification and non-quantification, internal and external, management and execution in the traditional management of enterprises, it provides a strong basis for scientific management decisions of enterprises, and can greatly improve the quality of decision-making of enterprises.

Performance management is one of the indispensable management tools of modern enterprises. Its essence is to collect various performance data and then analyze them to provide support for various management decisions of the enterprise, and to predict the future of the enterprise and improve the process.At present, EPM (Energy Efficiency and Production Maintenance Management) will obviously be one of the main goals of enterprises' future competition, which will also lead to a new revolution in enterprise information systems.

At present, most enterprises still focus on the monitoring of KPI (key performance indicators) for EPM, which is obviously not enough. The basic premise for determining the effectiveness of EPM should first be how to choose appropriate and accurate key strategic indicators .A small number of large enterprises have realized this problem, and they have begun to extract more in-depth analysis results through data warehouse applications that aggregate KPI and massive financial and operational data.Companies like Google, Amazon, eBay, and Walmart.

An enterprise database in the traditional sense is like a product warehouse, which adopts a completely structured approach when storing and querying data.How to deal with unstructured data?Obviously, traditional enterprise databases can't do it.The development of social media and IT technology has made big data analysis possible. Big data tools can store and analyze massive unstructured data that cannot be processed by traditional databases.

Big data analysis means that enterprises will be able to expand the sources and types of data supporting decision-making to areas that were previously unattainable: to better understand employee behavior through search engines, social media, blogs, videos, etc. combined with structured transaction data.

Substantial growth in the volume and type of data
With the help of big data analysis, traditional KPI analysis tools will have the ability to analyze large amounts of unstructured data.For example, NetPromoterScore has been able to evaluate the user satisfaction and loyalty of a brand or product by analyzing information from social applications such as Facebook and Twitter.This is just a small start. As intellectual capital transitions to social capital, a company that cannot evaluate open social media information cannot conduct effective performance management.

The democratization of big data analytics

In the past, enterprises needed to invest a lot of money in IT and data analysis to acquire, process and analyze a large amount of data.But in the era of big data, this part of the capital investment will be largely "reduced", because there are many data processing tools for free.

Google Trends.A common analysis tool for trend data, any company can use GoogleTrend to analyze a large amount of trend data and apply it to market research.

Social Mention.This is a tool that allows companies to track brand mentions in social media, and can even conduct trend analysis on user evaluation information, such as whether most users respond positively to a certain brand or are generally dissatisfied .

TripAdvisor.This is a well-known online travel review website. TripAdvisor can collect tourists' evaluation information on hotels, restaurants and scenic spots through special tools, so as to provide a reference data for operators of hotels and restaurants, which is convenient for operators to analyze customers. Preferences and consumption trends, so as to make timely improvements.

The above tools are some good examples of data analysis.In addition, in other aspects of big data analysis, the most popular and typical example is the "Car Claims Prediction Competition" held by Kaggle.This competition will gather some car enthusiasts who don't understand the business at all. They can predict the accident probability of different car brands only by relying on the limited data provided by the manufacturer (specific models are replaced by symbols).Now you have to ask, is it accurate?It has to be said that the accuracy of the prediction results of this competition is 340% higher than the manufacturer's own prediction.

The Technological Change of Performance Appraisal

At present, most enterprises still continue the tradition and regard enterprise performance management as an analysis and reporting software based on data warehouse. However, in the era of big data, cloud computing and SaaS (software as a service) are gradually becoming enterprise performance management software. technology development trend.If the enterprise cannot realize this in time, it will definitely be eliminated by the market in the near future.In addition, Hadoop, as a framework capable of distributed processing of massive data and in-memory data analysis, is also a hot technology in the future of enterprise performance management.

☆Talent selection
The traditional talent recruitment method is usually that the human resource department of the enterprise or the government agency rents the venue and hosts the job fair. Although this method is direct and effective, its efficiency and the consumption of enterprise resources have to be said to be huge.In the era of big data, with the advent of big data new technologies that are abundant in the talent market, traditional methods of recruiting and selecting talents will gradually be eliminated.

Google receives more than 10 resumes per month. With such a huge number, if companies conduct interviews one by one, the cost will undoubtedly be huge.So, how to filter out the most suitable ones among thousands of job resumes?
To this end, Google has adopted big data technology, which will allow all employees to complete a questionnaire with 300 questions, and then establish a set of mathematical models based on the analysis of the results of the questionnaire.Obviously, the current employees have been trained to meet the company's requirements in all aspects, and the questionnaires made by these employees are the most representative of Google's human resource needs.This mathematical model allows Google to get rid of the recruitment defect of "knocking on the door" with a diploma, so as to start to discover those potential "best-suited" applicants from other aspects. They may not be good grades in school, but they may be more Has potential for development.

In 2003, American writer Michael Lewis wrote a best-selling book "Magic Ball-Wisdom to Win in Adversity". After that, Billy Bean, who was the general manager of the Oakland Athletics baseball team, became a favorite star.

Bean's baseball club relied on "scouts" when picking players, and in 2002 Bean decided to change that approach by using a mathematical model developed by a young statistical genius at Harvard University. Come to collect players, and later, this model became an exclusive model for Bean and his subordinates.

The miraculous performance of the Oakland Athletics proved the reliability of this model. Since then, this small team has continued to rewrite baseball records on a tiny budget and created the most in American baseball history. The longest winning streak, with 103 victories in just one season.What an impressive achievement!In the history of American baseball, only the veteran Yankees can match it.

The success of the Oakland Athletics sparked a revolution in professional baseball, and since then, more and more teams have embraced digital recruiting, with clubs willing to spend money on "mathematical models" to use predictions scientifically The model evaluates a player's potential and his market value.Most of the pioneers have tasted the sweetness, and they obviously have advantages over those relatively conservative counterparts.

If we don't understand big data and just read the story of this baseball team, we can only understand it to the extent of "an unforgettable inspirational legend of a baseball team", but now you can understand that it actually proves that A trend in big data applications: predictive statistical analysis and big data applications will change the traditional way of recruitment and give 500 million candidates a new opportunity for comprehensive evaluation.

Here, can we also think this way: using big data, it will become easier to form a strong and high-quality team.

Behavioral Information - Already Constitutes a Gold Mine

The development of the Internet and big data technology has made it possible to obtain human behavior information on a regular basis.Behavioral information has more depth and scope than other information.At present, more than 98% of the information in the world has been stored digitally. Compared with 2007, this data volume has quadrupled as a whole.

So, who are the people who provide these data?In fact, whenever you use the Internet, whether you're working in your own home or at a company, you're generating tons of this data.For example, in your daily work and life, you have generated behavioral information data by sending emails, browsing the web, using social media tools, or engaging in more other activities.And when generating data, you are the one who unwittingly helps launch a whole new social project.Big data affects the era we live in, and various industries have undergone tremendous changes because of big data.For example, Wall Street has been changed by computer program algorithms that predict stock price movements; traditional marketing methods have also been challenged by algorithms that use Internet browsing records.People tend to believe that big data will change the economic field, but for the job market, only a few people believe that similar data-driven methods can be widely applied.

It can only be said that the progress of an era always needs time to be confirmed, and those who are at the forefront of the era can always succeed before others.The HR of many enterprises has listed big data tools as the most powerful tool for human resource management.John Hausknecht, a professor at the School of Industrial and Labor Relations at Cornell University in the United States, once told the media that the demand for "labor analysis positions" in the United States has increased significantly in recent years.In order to adapt to the latest situation of supply and demand in the labor market, he himself modified the main subjects in his teaching courses.

So, does this mean that we can safely and boldly hand over our careers to data analysis?

People generally don't want to be the "first to eat crab".The use of "predictive analytics" in career analysis is the crab that may bite. For most professionals, this is still a very new field. A more appropriate name is "people-oriented analysis." (people analytics).How to apply theory to practice is the real challenge, and there are many people who have raised moral objections to this.

In order to achieve the purpose of forecasting, the "predictive analysis method" needs to establish a comprehensive and huge personal technical statistics table, which must contain all the content related to personal performance, such as business performance, work attitude, skills, KPI, etc., and its scale is beyond our imagination.

The more authentic the data, the wider the scope of personal performance that needs to be involved. In other words, this data is like a private monitor, which will spy on the most secret places of human nature, such as your growth experience, fertility, private life, etc. .Therefore, most companies in related fields are only at the stage of researching "whether crabs can be eaten", and they are exploring the possibility of such applications.

But it can be predicted that in the next five to ten years, new models will inevitably be born in the data analysis industry, and some large-scale new experiments will also come out.This will be of great benefit to the future corporate human resource management and personal career path.

The historical changes of the "selection of talents" model
From the moment the concept of "company" was born, the role of "manager" was born. Their task is to recruit talents for the company, select the most suitable talents from thousands of different people, form a team, and create benefit.

There is an interesting little story about the history of talent selection:
At the beginning of the 20th century, a manufacturer in Philadelphia, USA, was recruiting a large number of workers. At that time, someone came up with a strange method to determine the available candidates.The boss of this factory asked the foreman to stand at the gate of the factory and throw apples to the job seekers surrounding him, and the strong ones who could quickly receive the apples would be hired by the factory.Now it seems that this method is obviously not scientific at all.Times have changed, and today's recruitment concept has a completely different judgment from that era.Elite managers of some companies prefer to use a "cruel" method to achieve the goal of high-quality talents, and "survival of the fittest" has become the strongest voice of our time.

We can look back at the leading business giants of that era: U.S. Steel, DuPont, and General Electric, etc. They are all integrating in a way where big fish eat small fish, and one of their actions may affect the fate of the entire industry.A weaker firm is swallowed up by the competition, and a stronger firm emerges.Those who invent these "brutal" competitive models are often favored by the industry giants, and they are tapped to sit in higher positions in stronger companies.

For a century, this approach has worked unimpeded in strong crowds.People rely on their outstanding performance to become the selected talents, and managers also select the needed people based on people's performance.As Wharton professor Peter Cappelli writes in his treatise: “In the science of prediction and selection, there is no method as powerful as observing how people actually perform.”

(End of this chapter)