Many companies are exploring how to tap the value of big data, but companies have to have a good at doing this job, and that is: data analysis provider. The company's business model must be able to determine its effective and continue to provide customer is willing to pay for data products and services.
The following will share a great provider of data analysis - ComScore, how three key factors (a convenient, scalable platform, a deep understanding of data analysis team of employees, as well as a deep understanding of the customer) is achieved by large data the value of the excavation.
A convenient, scalable data platform
ComScore provide basic services is 14 PB of data from multiple sources online, real-time data collection from around the world, which is now called something big data.
Source of the data has four main channels. The first source is the sample library data from two million Internet users in the United States and abroad have 100 million users. ComScore sample allows library members secretly capture user behavior and demographic characteristics. Census data is the second source of data obtained from approved placed in approximately 90 percent of US digital media agency 500 companies inside the sensor. The third source of data is the use of an exclusive survey of the way, the cognitive data obtained from a sample library members. The fourth source is data obtained from a strategic partner. For example, ComScore adopts a membership card in-store purchase data, to help customers work closely together online and offline campaigns store buying behavior.
In the late 1990s, the plant can manage ComScore data mainstream commercial technology solutions has not yet appeared. ComScore on the development of an efficient and obtain patent protection technology platform to store and manage large data.
By 2013, the technical team has evolved the platform as a service-oriented architecture from a proprietary solution that supports three critical systems that run on Hadoop and MapR of Pivotal Greenplum database, each system has its own work mission and expansion needs. ComScore need uninterrupted stay in the top of the platform in the past 12 months, the amount of data grew by about 80%.
A deep understanding of data analysis team of staff
ComScore dependent on its outstanding staff to master the art and science of big data. It has grown into a global team of 1,200 people, everyone has different levels of "data scientists" capability. In order to keep pace with the company's growth and development and selection of these employees, the company implemented a number of leadership talent management strategy.
One strategy is: first to provide more analysis training from professional recruitment talent analysis and mathematical University Business School, and give her (him). ComScore has an ongoing internal company-wide training program -ComScore University, the majority of employees have participated.
Another practice is designed to develop a wide range of fields, deep understanding of data analysis team, set up different groups throughout the company depending on the skill level of employees, but at the same time, employees must be sufficient to ensure that these groups have a smooth communication basic skills. These teams need to communicate often work in the same place.
Leadership of the company also uses a matrix organizational structure, together with the different skill sets of personnel. For example, ComScore run several parallel teams, each team has a commercial product manager to understand the product roadmap, and analysts to monitor the quality of data and analysis methodology. This way is considered particularly important to the product development process.
Deep understanding of customer insight obtained action
ComScore released about 15 million copies daily online reports. The challenge is how to disseminate good report, so that customers can not only get insight, but also will work closely with this insight tasks. When the obtained customer insight into action, such as changing a policy or a campaign targeting new customer segments, they can get value from ComScore products and services. But as a provider of data analysis, ComScore way for customers to use data lack of control.
ComScore is able to control its own understanding of customer needs. The company is committed to the depth of excavation in areas of expertise, and understand their customers want to solve the problem. After Armed with this knowledge, ComScore to make big data becomes "consumable" and can actively help clients identify what is actionable insight. Have to admit, the data can be overwhelming in a short time, encouraging customers to focus on solving key issues, and then repeatedly apply.
ComScore to offer its customers have a graphical interface, integrated software tools to visualize and chart presentation; to support customer self-service through the extensive use of wizards and templates reporting; when you need immediate decisions, provides real-time access to data through the control panel; and create a knowledge portal to support the sharing of knowledge. Furthermore, the company also set up an organizational unit, to support its consumable products and services.
(Pictured, ComScore's typical customer)
All attributed to data
In the field of big data space, we must constantly adapt to keep up with its rapid development.
Years ago, ComScore also desirable to measure only the number of visitors hanging pages of advertising clients; today, it must report if this ad is actually seen. Early on, the company's non-US customer data has not been much interest; globalization gave birth to ComScore data covering the global market demand.
In addition, ComScore also need to follow to keep technology-related changes. For example, customers may need to analyze similar data from smart phones and video streams such burst data types. Another important example of change is the number of users using a number of different devices to access online content continues to grow. In this way, our customers need to be able to identify unique users across multiple devices.
ComScore platform, talent and a deep understanding of the customer gives the company enough energy to cope with the ever-changing industry environment of big data. Its convenient and scalable data platform enables ComScore able to adapt to the rapidly growing amount of data, and obtain a new data type. Cross team and a combination of skills, combined with deep analytical skills and team work, making it possible to smoothly adapt to changes in demand. The use of analytical tools to customers in good training, allows the company to expected future trends, and adjust their products and services. CISR (MIT Center for Information Systems Research) believes that big data to help develop the company's own interest to adopt ComScore method can be considered to get good service.
Recommended Reading articles
Big Data era need to know six things
Big Data framework hadoop Top 10 Myths
Experience big data development engineer salary 30K summary?