New approach to big data analysisBarry Copeland 24 / March / 18 Visitors: 66
New approach to big data analysis - from expensive systems to affordable cloud solutions
Digitalization has covered almost all sectors of the economy and spheres of life of modern man. The spread of digital technologies has led to a significant growth in the sources of data acquisition and a rapid growth of information volumes. Big data now needs not only to be collected, stored and accumulated, but also to be able to process it in real time. And on the basis of this analysis to make strategic and operational decisions that allow business to work more effectively with customers, to keep a hand on the pulse of market trends, to respond instantly to any changes.
With the competent use of large data, the business will know everything about the client: his preferences, purchase history, solvency, geolocation and much more. Big Data technologies are used today in many industries. In industry, the construction sector and even in the transport sector, the information received from smart devices and sensors is mainly used. And in industries such as finance and especially retail, people, customers are at the heart of the big data flow.
Once a customer enters a store, he or she instantly becomes the object of information collection. Solutions in the field of video analytics determine gender and age, tracking the movement of a person in the sales floor. The most advanced retailers use geolocation beacons, thanks to which you can make a reliable thermal map of the buyer's movement in the trading hall. This, in turn, allows you to optimize the laying of goods and thereby increase revenue.
Shopping history and loyalty programs are no less promising sources for collecting big data. Today, almost all retail chains offer customers a discount or savings card or smartphone application, which must be presented to the cashier when paying. The store receives a "portrait" of the buyer, taking into account his interests and solvency, and the buyer becomes available to various personal offers and discounts, including promotions for his birthday and other holidays. Data on consumers and their preferences become the basis for generating analytics. They provide a comprehensive picture of the demand for product categories and individual items up to a single store. Let us not forget other sources of data acquisition for retail - internet orders, market research, customer activity in social networks, etc. These are just a few examples of how a competent and effective analysis of large data helps retailers to increase profits without much expense, to expand the range of products, to conquer new markets. Big Data is, in fact, also a commodity that can be sold and bought, and on its basis partnerships with other market players - banks, telecom operators, etc. can be formed. The largest retail chains are already on this path, and quite successfully.
The benefits and advantages that big data carry retail, we can talk about endlessly. But the question arises: how to analyze this information with the least cost and the most efficient? After all, business needs the results of this analytics not at the end of a month or a quarter, but today, "here and now".
There are various technological approaches to storage and processing of Big Data. You can use traditional server solutions and disk drives, but unfortunately, due to objective limitations, they cannot guarantee prompt delivery of results. RAM (in-memory) data processing technologies and hybrid solutions based on them provide significant acceleration. But their cost remains quite high and such products can not always pay off even in a large company.
As an alternative to retailers, we can recommend cloud platforms for processing large data. After all, it is the cloud that enables businesses to use IT as a service without having to purchase expensive infrastructure or invest in computing resources. At the same time, the cloud model assumes payment only for those capacities which are actually used for the decision of tasks facing business. On the Russian IT-market it is possible to find both the foreign cloud platforms including a corporate data warehouse and powerful tools of business analytics, and the domestic developments constructed both on Russian technologies, and based on components with an open source code. One example is the CROC Cloud Services service, which is based on a corporate solution that allows accumulating all company data in the cloud Data lake and working with them using BI-systems and data science tools. Cloud Big Data helps to reduce not only infrastructure costs, but also the speed at which business tasks can be accomplished. For example, the business processes of one of our customers have been compromised by delays in equipment delivery. To solve the problem and provide the client with a system for data analysis, it was decided to deploy software from the cloud. As a result, for those three months that the company in test mode used a cloud product, it was able to make sure that the IT costs due to the transition from CAPEX (purchase of iron) to OPEX (cloud) can reach up to 5 times, while the implementation time is reduced by 6-8 weeks. Another example from our practice: a large hypermarket chain decided to fight shoplifting with a synergy of clouds, Big Data and offline video surveillance and machine vision-based monitoring of violations. Since the customer did not have their own developed IT infrastructure for this task, the company's specialists implemented a cloud-based video analytics system in a managed service format. The appearance of this system made it possible to track visitors from the "black list" in real time. As a result, it led to a monthly decrease in theft in the network's hypermarkets. Losses from such activity at large retailers can be counted in millions of rubles. In what direction will the Russian market of Big Data Solutions for Retail develop? Most likely, it will be simplification and convenience for end users. To deploy and launch a data collection and processing platform in an enterprise data center, you need your own IT department, and you also need IT specialists to connect the cloud platform. But when you connect to a Big Data Cloud platform as easily as today's Internet Acquiring or any other service, the entire retail experience will change dramatically.