How does corporate strategy of data management help business transformation? TADetails

In what Data Governance methodology essence

Data Governance ("Data management") is the strategy of work with corporate data and a set of solutions corresponding to it. Understanding of need of such methodology was developed decades, and therefore programs in the market exist for a long time. In Russia the idea gains popularity especially actively: many companies reached that level of a maturity at which work with data becomes one of the most priority tasks.

Data Governance is a system which allows business to understand what assets of data it owns as these data are connected with each other who and as uses them, allows to manage their quality. And in an ideal picture of the world all this information works at language of business terms — the expert of practice of Big Data of Croc company Egor Osipov explains.

Specialists note also that the demand of Data Governance is caused by popularity of Big Data and lakes of data. At the same time, if in the company there is already correctly constructed ETL system and information flows are competently organized, then implementation of Data Governance will not be technology difficult.

Data Governance as business function in the context of digital transformation

Data Governance is not a certain model of the relation to data which use guarantees momentary success. As confirmation to it serve the companies which take the leading positions in the segments of the market even at poorly developed approaches to data management. But if the enterprise intends to go the way digital transformation, then to do without Data Governance will be difficult.

However, in Russia of a talk it is still much more, than real implementations. Only separate components, but fully implemented projects — some tens are interesting to someone.

By experience of communication with our customers, I would tell that 10% of the companies already implemented similar solutions, 20% implement them at present, and 30% seriously consider such possibility. About 40% do not consider implementation of these technologies a priority task in the next several years — Egor Osipov tells.

What problems with data are solved by solutions of the class Data Governance

In the Russian market many companies really not always understand what to do to them with that data array which is at their order. Analysts select three main problems: quality problem, problem of knowledge and problem of control.

The problem of quality of data, probably, is clear to all — the expert of practice of Big Data of Croc company Egor Osipov says. — Many companies already implemented at themselves tools for quality management. If to say about a problem of knowledge, if a BI system allows to understand better what happens to business of the customer, Data Governance — what happens to data of business. It is tool kit using which business and IT understand what data are in an asset of the enterprise from where they appear as are connected among themselves as are calculated who is responsible for them. Plus, of course, when we speak about a control problem, we should understand accurately who and when got data access what data are sensitive and so on.

However, often it is difficult to business to decide on Data Governance implementation. Usually it is connected with the fact that already existing ETL system and the data warehouse were under construction long ago and by those techniques which do not allow to implement Data Governance easily.

The second complexity which the companies at decision-making face — appointment responsible. In methodology of Data Governance it is registered that all data require responsible. And very few people love responsibility — Vladimir Kolganov, the head "Storage systems" of Croc company notes. — The methodology does not allow to make in all errors guilty IT service because its implementation requires close interaction of IT and business. It is a lot of work which often requires creation of new governing body or consortium by data. Without such approach the solution will not be simple to react to business tasks competently.

What technologies provide data management

Data Governance is quite wide term, and technologies when implementing strategy the set is used. At the same time producers often have different vision of what solutions to carry to this strategy. Approaches of Informatica and Collibra companies are most popular, and among key software products Data Quality, Master Data Management, Metadata Management, Data Stewardship are called.

How to develop and implement the strategy of Data Governance in the company

The company should be ready to structural changes and to need of updating of outdated data warehouses. As this rather painful also requires process narrow specialists, and it is difficult to carry out implementation of Data Governance exclusively by own efforts. Infrastructure of data storage needs to be prepared for new calls together with the checked technology partner who will be able to book its correct audit and will help to define what high-quality changes will be required to go to Data Governance.

Everything depends directly on the level of a maturity of IT architecture which is implemented in the company. If it was designed not on the best samples which exist in the market, then data warehousing systems will change most seriously — Egor Osipov warns. — If it was designed according to the best enterprise-experts, then global changes will not be required. It is connected with the fact that the companies which develop products are also guided by the best examples, but do not create the solution in a vacuum.

Work with data which became a strategic digital asset demands from computing infrastructure of the maximum flexibility. Therefore building of architecture for productive work of analytics and platforms becomes the main task of CIO. Eventually it leads to creation of new values for end consumers and increases profit.

For example, Data Governance means creation of unified virtual environment for work with data. Correctly built capacities integrating productive computing platforms for big arrays of information, DWH for contemporary records, drives for storage of hot data and quick access to them at any moment are for this purpose necessary.

Taking into account this requirement, we offered the market special approach — "smart data storage" — Vladimir Kolganov tells. — It provides separation of data into hot and cold, and allows to differentiate them depending on use frequency. Hot data effectively are placed on productive carriers, for example, all-flash-arrays. And cold are stored on slow DWH. It allows not only to reduce costs on infrastructure for data storage, but also to balance management of them during decision making process optimization.

After creation ready to transition to Data Governance of computing infrastructure, the Chief Data Officer (CDO) will solve where to store data. It can be done in the company, and it is possible to transfer to outsourcing to DPC or a cloud. In passing the architecture for any platform, for example, using the HaaS (Hardware-as-a-Service) model – the option of outsourcing of computing resources gaining popularity can be created. The equipment in a configuration necessary for this purpose can be rented and be placed in commercial DPC without capital costs. Experts consider that it will make transition to Data Governance softer and economically reasonable.

What projects are implemented in Russia

Many Russian companies begin to prepare infrastructure and to pass to Data Governance, implementing projects separate blocks. For example, for the large company from the sector of mass passenger traffic specialists of Croc implemented the project on creation of the subject-oriented database (POIBD).

She collects information on sales of tickets and income from transportations and also helps to receive flexibly configured analytics hundreds of attributes using Big Data — Vladimir Kolganov gives an example. — And in the project on optimization of the data warehouse for federal insurance company "Croc" helped to accelerate data processing and to move to the platform with new architecture more than 10 TB of the contemporary records necessary for development of insurance products. Having reduced time of creation of the reporting by 150 times, the insurance company can quickly analyze a business status, quickly bring new products and services to the market, to use data for development of customer relations.