Big Data also has the West big drought

Big Data has developed into a critical phase. By 2017, the big data market will have grown to 50 billion dollars, but unfortunately 55% of the big data projects have failed. With the opportunity is the hype and false information, we are in the large data of the West drought stage. The big data industry is in a bit of a stalemate: people who understand it invest in the industry to collect, store and benefit from large data, while others pay for it with a dubious attitude and wonder how big data will affect their business.


Benign failure


large data allow error rate, this may sound contrary to common sense, but failure is also a benign failure and malignant failure of the points.


? And listen to explain: "Testing and learning" method can be used in the original state of large data. Companies have to discover these failures by making assumptions and then validating them. This allows companies to develop truly coherent strategic solutions by digging up big data.


these "bugs" are actually the necessary process for finding the right analytic results, creating significant opportunities for all walks of life, such as precision recommendation, risk management, equipment failure prediction, and streamlining logistics management processes.


in some optimistic cases, these companies are using big data to judge the development of new products, to open up fresh sources of revenue and even create a large data-driven corporate culture. To evolve to this stage, enterprises must break the rigid quantitative and cost control ideas, develop flexible analytical and judgment methods, and gradually realize the optimal development of enterprises through large-scale automated forecasting. Only in this way can they really find profitable business models with large data and develop new products based on data.


in the early days of large data use, speed was a key factor. The faster the data is completed and the more quickly the organization's expertise is built up, the faster it can create value and make more granular use of large data at new heights.


Quantcast perfectly interprets the four-step model of growing up as a big data company. The first step is to provide free web traffic monitoring services. Quantcast quickly overtook traditional database technology as market recognition increased-its daily calculations grew from thousands to 1 billion levels. And the expansion of the business step-by-step to provide it with a higher quality of analytical and judgment capabilities, better by virtue of the audience analysis for the enterprise value-added and maintenance of customer relations.


Quantcast soon saw the need to invest in large data science, because it was challenging to find the distribution of demographic data and interest maps in the vast range of activity data. Shortly thereafter, Quantcast tested many products and services, a very successful project called Lookalikes, which helped advertisers find new customers with a high degree of similarity to existing customers. Quantcast now earns $100 million a year, and its flexibility allows it to respond to opportunities in the development of large data in a timely manner.


Quantcast, Google, Facebook and LinkedIn are the forerunners of big data companies that have gone through these stages of development. When large data business is becoming more mature in internet companies, it is time for other companies to embrace large data to create value and build competitiveness. For example, large IT vendors are using refined technology product data and transaction data to generate analytic prediction models to improve the recommendations and optimize the trading experience.


Malignant failure


Unfortunately, many companies are still stuck in very rigid models, and they are simply using large data from the standpoint of cost control and storage scalability. Maybe they're still on the sidelines. "Flexible analysis" (Agile Analytics)--Break the traditional method of saving thinking and using data flexibly.


This means that many companies exploring big data are missing the opportunity to improve their business and optimize their services, and they are also missing the opportunity to exploit large data rather than assume the development of new products. They are entering a stagnant period of big data development-learning to store data but not extracting value from it.


Big Data requires human and resource input--from the human dimension, the need for more people to master this technology, and the traditional cost-saving method requires layoffs. For big data companies, the opposite is the case. To achieve a breakthrough in large data, enterprises have to be willing to vote for money. Companies that cannot react quickly to change and invest in time will take more opportunities from smaller, more flexible businesses.


Failure


Nothing is more embarrassing than the 55% failure rate mentioned earlier? What is the reason for the failure? One idea is that there are too many cheats in the West, and bombast can't see their results. They realised that the culture of big data was being emulated. Despite the lack of qualification, the estate advisers and system integrators have positioned themselves as experts in the field.


Similarly, many traditional businessmen tout the pattern of the last era as "big Data". Many of them are still using SAS on computers that are not networked (Statistics analysis System, which started in 1976)-Can it be big data? Others are focused on data mining and reporting, extracting, transforming, and loading small database content. These merchants typically use proprietary software that is disconnected from open source parallel computing programming tools such as Apache Hadoop.


we are at the key point of big data development--we need a steady stream of data to keep growing. and enterprises to think of outdated technology or skills as large data, ultimately because of the analytical ability to keep up with the injury or their business. If their projects fail or provide inaccurate information, they will lose the battlefield and cede opportunities to competitors who really know the technology.