Big Data Theory and Practice in Industries

Big Data Challenge Across Industries

By 2015, despite the short life of the Big Data sector, there are already estimates of the effective use of these technologies based on real examples. One of the highest indicators relates to energy - according to analysts, Big Data analytical technologies can increase the accuracy of generator capacity distribution by 99%.

Analysis of failed Big data projects

Analysis of failed Big data projects ^[1]

Big data in the public sector

Main article: Big data in the public sector

Big data for carriers

If we talk about big data methods aimed at obtaining an effect for the telecom operator's business, then in the general case four main areas are considered - the first three are aimed at improving the internal work of the company itself, and the latter is an additional market product for external customers^[2]

• precision marketing - targeted supply of products and services to those consumers who are most ready to purchase them (new tariff plans, additional services, payment terminals, etc.);
• Customer Experience Management to increase customer satisfaction to prevent user outflows
• Optimization of the operator's internal work and development planning (ROI-based Network Optimization and Planning) on the basis of taking into account all objective factors and consumer opinions in order to maximize guarantees of return on investment as soon as possible;
• monetization of information assets (Data Asset Monetization) - the sale in one form or another (including in the form of equity participation in projects) of data available to the operator to their partners so that they can solve their tasks with their help.

By deploying the big data solution, the mobile operator was able to begin to collect and analyze significantly more information about the behavior and interests of its customers, including the intensity of communication use and geographical location. Moreover, all this information could be linked with data on the operation of the cellular network itself, including its loading, failures that occur, etc.

The possibilities of using such methods are visible from the results obtained. So, at the beginning of 2013, the efficiency of marketing offers (for customers who accepted them) with a total mass distribution was 0.7%. By the end of the year, due to the simple segmentation of subscribers (by age, sex, subscription period), this value was increased to 4%, and during 2014 it was increased first to 11% (accounting for the intensity of service use and customer location) and then to 24% (taking into account preferred options for receiving an offer - voice calls, SMS, e-mail, social networks, etc.). During the year, it was possible to reduce the number of unsuccessful calls to customers by 11 million, significantly reducing the cost of advertising campaigns.

Based on an analysis of 85 parameters of subscriber behavior, a "risk group" was identified, potentially ready to leave the operator's services. Within it, a certain segmentation was also carried out, and for each category of customers a set of measures was developed to increase their level of loyalty (discounts, other tariff plans, gifts, etc.). The customer conducted a study, dividing the "risk group" into two subgroups: special retention actions were carried out from the first, nothing was done from the other. An analysis of such work over the year showed that the company was able to significantly reduce the outflow of its existing consumers, retaining more than 200 thousand subscribers; it should be borne in mind that the cost of retaining a customer is always significantly lower than attracting a new user.

Prior to the use of big data, the expansion of the operator's geographical network was actually carried out only on the basis of information about the density of buildings and the population, but having implemented this solution, China Unicom moved to develop its activities based on multifactorial analysis, which took into account indicators such as real traffic congestion and service demand (for example, taking into account the place of work of people), the "value" of customers (by standard of living), requirements for communication quality (distance between receiving stations), demand for different categories of services (the use of different equipment depends on this), etc.

In terms of monetizing client data for external partners, two examples were given: firstly, optimizing the placement of outdoor advertising, both geographically (the place of residence, work or transport communications of the right customers) and taking into account the time for dynamic advertising (depending on the time of day, days of the week and seasons of the year, the composition of the public may change), and secondly, similar proposals for the development of retail networks (taking into account location and assortment). In addition, it is very profitable to target mobile advertising in real time in accordance with the schedule of a person's employment, interests and physical stay (for example, sending information about action films that the client is interested in, precisely in his spare time and taking into account nearby cinemas). General industry experience shows that such targeted methods can increase advertising revenues at times.

Big data in banks

data analytics will allow banking organizations to better control information within the company and identify signs of fraud much faster than was possible before," Avivah Litan, lead analyst and vice president of Gartner Research, said in early 2014.

Massive adoption of big data analytics is complicated by the fact that banks often use disparate or simply outdated platforms. However, there are already examples of how employees responsible for information security prevented fraudulent transactions. In addition to Big Data technology, experts also believe that the introduction of modern user identification systems allows combating fraudsters. One example is the so-called continuous behavioral identification, which analyzes the behavior of clients over a long time. This is done by linking the account to a mobile phone.

Big data can solve almost all the key tasks of banks: attracting customers, improving the quality of services, evaluating borrowers, countering fraud, etc. By increasing the speed and quality of reporting, increasing the depth of data analysis, and participating in countering money laundering, these technologies help banks meet the requirements of^[3]

The main tasks for which banks use big data analysis technologies are prompt reporting, scoring, preventing dubious transactions, fraud and money laundering, as well as personalizing banking products offered to customers.

Big data technologies are mainly used to analyze the client environment. Dmitry Shepelyavy, the deputy CEO of SAP CIS, gives several examples: "The American bank PNC data on behavior of the clients on the websites, converts information on purchases and a way of life into policy of flexible charge of interest rates which as a result is expressed in capitalization growth figures. Commonwealth Bank of Australia (CBA) analyzes all transactions of its depositors, complementing this analysis by collecting data on them on social networks. By linking these data flows, the bank has achieved a significant decrease in the percentage of non-payment of loans. And in Russia, the experience of the Ural Bank for Reconstruction and Development is interesting - they began to work with information on the client base to create loan offers, deposits and other services that can interest a particular client as much as possible. In about a year of application of IT solutions, UBRD's retail credit portfolio grew by about 55%^[4].

At Alfa Bank, in 2013, the bank successfully completed the development of a prototype solution for interacting with social networks, now there are several pilot projects exploring various business hypotheses.

"Banks have a huge amount of structured customer information that can be successfully processed using big data technologies. This allows you to quickly make decisions and make highly relevant offers of banking products based on analysis of customer behavior, customer activity and client operations, "said Maxim Azrillyan, chief technical architect of the Alfa-Bank Center for Innovation and Technology of Electronic Business.

Big data in insurance

Insurance companies are interested in applying Big Data technologies, but only a few have begun to actively work in this direction. Such data in a joint study cite Bravura Solutions and Financial Services Council (spring 2014). Researchers interviewed a number of leading insurance companies about their plans to modernize and implement Big Data solutions.

According to the survey, 67% of insurance companies believe that they have only limited access to user data. According to respondents, these data are enough to personalize interaction with customers, but not enough to predict their behavior. However, for more than 56% of respondents, it is the creation of personalized campaigns that is the main goal of the development of marketing communications.

About 30% of insurance companies surveyed today already use Big Data technologies and analytics in order to predict the needs of customers and create personalized messages. The main problem for those companies that do not yet do this is the lack of necessary systems, the study says. Insurance companies have data sets, but so far there is no way to take full advantage of them. Most insurance companies are somehow interested in upgrading their IT systems in the next five years. However, for 23.7% of organizations, the issue of modernization is not yet worth it.

Big data in healthcare (pharmaceuticals)

Big Data Usage Areas

The introduction of mobile technologies in the field health care and the distribution M2M of devices will contribute to the expansion of the use of Big Data in the medical sector, but only a few healthcare institutions are ready to work with big data. This conclusion was reached by experts of the company MeriTalk, which conducted a survey among 150 top managers from the field of public health. USA The authors of the study (spring 2014) MeriTalk tried to find out how much medical executives are ready to work with Big Data and what steps have already been taken to meet these new technologies.

Less than 25% of top managers of state medical institutions believe that their departments are ready to work with Big Data. Only 34% of respondents noted that their department invested in technologies that optimize the data collection process. Even fewer managers (29%) hired IT specialists who are engaged in data management and analysis. 29% trained key managers to work with Big Data.

At the same time, more than half of respondents (59%) are confident that successful work with Big Data will be a key factor in increasing the effectiveness of their institution in the next five years. 63% of top managers believe that big data technologies will more effectively monitor the health of patients, and 60% say that thanks to large data, preventive work will improve.

In medical institutions, M2M technologies have not yet been widely used: only 15% of top managers mastered them. However, 53% of respondents plan to rectify this situation in the next two years. According to MeriTalk analysts, it is M2M technologies that can play the greatest role in improving the quality of patient care and remote monitoring of their health status.

Private and public institutions are already actively using big data to create personalized offers for their customers. But is this relevant for an industry like healthcare? The answer is yes! After all, understanding the needs of patients directly depends on how actively new technologies are used in medicine^[5].

According to a study conducted by McKinsey&Company in 2014, 75% of patients surveyed would like to use digital services - contrary to the established belief that the majority of the population is reluctant to turn to them in treatment.

To meet patients' need for quality health care, health care is increasingly turning towards smart technologies in many countries. In Germany, for example, today, thanks to Big Data technologies, oncological diseases or predisposition to them are detected by blood tests of patients and donors. As a result of timely diagnosis, the costs of the state and people themselves are significantly reduced, as well as the effectiveness of treatment is incredibly increased. After all, one of the most important enemies of the patient who launched the disease is time. Let us turn to the oncology mentioned earlier. Diagnosis and selection of the right treatment regimen can take precious minutes, which are so important in prompt response when malignancies are detected

Big data, in addition to already known and common tasks, can be used, among other things, to combat diseases and track the growth of epidemics, experts say. So, nine days before the Ebola virus outbreak was officially declared an epidemic, a team of researchers and scientists from Boston were able to detect the spread of hemorrhagic fever in Guinea using big data.

A picture of the movement of the deadly virus epidemic in West Africa was compiled by startup HealthMap, which runs on an algorithm that takes into account social media references, local news reports and other data available^[6] Network^[7].

Big data systems can be useful in the first place, not for detecting already manifested outbreaks of certain diseases, but for predicting potentially possible epidemics of this kind through the analysis of available information. In this case, virtually the same technologies that help marketers demonstrate targeted ads to consumers or offer music and videos to watch can be used to fight infectious diseases such as Ebola.

Big data in the automotive industry

Big data in e-commerce

Шаблон:Main 'Big Data in E-Commerce

Big data in retail

Offline retail uses big data to analyze the behavior of buyers, design routes through the trading room, correctly arrange goods, plan purchases, and, ultimately, increase sales. In online retail, the mechanism of sales is built on big data: users are offered goods based on previous purchases and their personal preferences, information about which is collected, for example, in social networks. In both cases, big data analysis helps reduce costs, increase customer loyalty, and reach a larger audience. All of these are just basic capabilities that can be realized with big data^[8].

Despite the economic crisis, the number of big data projects is expected to grow, including in retail. Although the introduction of new technologies threatens not only profit, but also high risks, companies have already learned about the successes of more decisive colleagues in business. In a difficult economic situation, the need to save and increase customer loyalty comes to the fore. These are the tasks that big data solutions are designed to handle.

In the struggle for a customer, retailers are increasingly turning to innovative technologies such as big data analysis, e-commerce, omnichannel services, RFID technologies, etc. In Korea, for example, the world's first virtual store was recently opened directly on the subway platform. By scanning QR codes from panels glued with images of various goods, Seoul residents put the selected goods into their virtual basket, which is then delivered to them home at a convenient time. Such technologies would probably resonate with forever rushing Muscovites^[9].

Tesco, the UK's largest retailer, is experimenting with augmented reality. For customers, an application was developed that allows them to quickly receive information about the calorie content of certain products and other information that did not fit on the price tag, simply pointing the tablet camera on the shelf and taking a picture.

Another example: until recently, online sales of clothes and shoes were not quite common precisely because of the inability to perform fitting in a virtual space. The buyer had a high risk of making a mistake with size or shape. But the situation is changing. Soon, a virtual fitting room will be available in the online store of eBay, allowing customers to "try on" the clothes they like from the network catalog to a three-dimensional model of their own body. A similar virtual primer project was presented by SAP at the Open Innovations exhibition in 2013 and was highly appreciated by experts. Thanks to such technologies, a person can use his photograph and the parameters entered (including height, size) to try on and order new clothes by paying with a mobile phone.

The development of mobile technologies can be called one of the main trends that affect the development of retail. The smartphone has become the most important tool in trade, analysts say, and IDC its importance will only grow. 69% of consumers already believe that it is smartphone simply necessary to make purchases and much increases the pleasure of the process. No one can deny the convenience of ordering from anywhere in the world and paying using the Internet or phone in any suitable way. As a result, the Omni Channel concept is actively developing - when real and virtual sales channels are combined into one. business process Already today, any customer wants to be able, for example, to start buying on the Internet by making an order there, and finish paying in the store and vice versa.

There is no denying the fact that competition in retail is increasing due to the emergence of "digital buyers." This new class of customers is used to choosing the best offers on the market with one click of a button and is in constant search of personalized stocks and promotional prices. Retailers are forced to look for tools that will allow you to create personalized offers and specifically promote the product. The customer interface Amazon.com is a textbook example of such a service. Each time you visit the site, the customer receives a variety of offers based on an analysis of the history of past purchases, viewed pages, left reviews, etc. The system processes huge amounts of information in a fraction of a second, each time converting them into a targeted offer, leading to an increase in sales.

Big data for the media industry

Requirements for storing big data in the media and entertainment industry increase very quickly as video resolution increases (July 2012). The spread of the standard HD and mobile video consumption stimulate the emergence of avalanche-like demand for relevant digital content. In this regard, the demand for storage solutions and for HDD for the creation of archival viewports is also growing, according to analysts Coughlin Associates^[10].

The penetration of flash drives in this industry has increased significantly - to 37% in 2012. Flash memory plays one of the key roles in the distribution of content and post-production, the researchers say. In the period from 2012 to 2017, the requirements for the capacity of digital data warehouses in the entertainment industry will grow 5.6 times, and the requirements for the volume of data warehouses involved per year - 4 times (from 22425 Pb to 87152 Pb).

Revenue from the sale of storage systems in the media and entertainment industry will grow more than 1.4 times between 2012 and 2017 from $5.6 billion to $7.8 billion. The maximum storage solutions in 2012 were used to save and archive new content (98%).

Coughlin Associates estimates that in 2012, 43% of the total delivered memory was in tape format, 41% in HDD, 16% in optical disks and 0.2% in flash (which is used mainly in digital cameras and some media distribution systems). By 2017, tapes will account for only 38%, and HDD - already 59%, optical discs - 3% and flash - 0.3%.

The total revenue from the sale of media and devices used in the media and entertainment industry will grow 1.3 times from 2012 to 2017 from $774 million to $974 million.

Big data in marketing

When competition escalates, it is important for companies to offer their services to customers at the moment when they are most in demand, and do it quickly. Therefore, the role of marketing is growing - this is no longer a side branch of the business, as it was before. According to IBM research, 63% of chief executive officers use the help of marketing directors (CMOs) in developing their business strategy. In terms of engagement in this process, CMOs are ahead of only chief financial officers with an indicator of 72%.

Now marketers can take advantage of modern big data technologies and powerful analytics, which repeatedly enhances the capabilities of marketing departments. If earlier they had at their disposal small fragments of data, on the main of which it was necessary to build a picture of the whole, and the data itself was sometimes stored where it was problematic to extract them, now the state of affairs has changed.

Marketing directors combine data from internal and external sources. First, people themselves report a lot of information, for example, on social networks. There you can track their preferences or criticisms of services. Analysis of such data allows customers to make personalized offers. This is particularly important for organizations in the SMB sector. Moreover, small companies are sometimes forced to open new areas of business if their customers need it.

IBM Corporation conducts a study annually starting in 2004, in which company directors are interviewed. The new study, titled "Taking the Challenge: How CMOs Can Begin Filling Information Gaps," involved more than five hundred chief marketing directors from 56 countries and 19 industries from around the world.

The results of the study showed that 94% of respondents believe that analytics will play an important role in achieving their goals. However, there was an increase in the number of directors (82 per cent of respondents compared to 71 per cent three years earlier) who felt that their organizations were not sufficiently prepared to benefit from the explosive growth of data.

The study also showed that when the director of marketing works closely with the director of information technology, the enterprise, as a rule, functions more successfully. Marketing directors' priorities now correspond to the needs of the digital economy. In 2013, for the first time, the item "technology development" came to 1st place among the priorities.

Another fact: 94% of marketing managers believe that mobile technologies will be the key to future success. Three years ago, 80% of respondents expressed this opinion. Many survey participants - 58% of respondents said they could deal with business affairs regardless of their location or device used.

Civil Aviation Big Data

According to analysts, 67% of companies from the aerospace industry implement projects based on Big Data, another 10% plan such projects. As for airlines, here the implementation of projects for February 2019 was announced by 44% of companies, and 25% announced plans for such projects.

These are the results of a study conducted in December 2017 by FlightGlobal regarding the role of Big Data for aerospace enterprises and airlines. Analysts also found an opinion on the sharing of data on the state of aircraft with manufacturers and companies engaged in repair and maintenance (TO)^[11]. The study involved 300 professionals from the aerospace and aviation industries. Most of them are confident that Big Data technologies can increase the operational reliability and efficiency of airlines.

Infographics with study results

About half of respondents replied that their companies use arrays of data on the state of aircraft, which helps them make more verified decisions. In the near future, the share of such companies will grow to 75%.

Data sharing with OEM/MRO still remains problematic. However, 38% of airlines believe that such a model can provide them with significant business advantages.

According to a May 2018 review by Honeywell's Connected Aircraft^[12], 47% of airlines surveyed plan to spend up to $1 million over the next year on each aircraft they operate. Most of these companies plan to meet the amounts of 0.1 to 0.5 million dollars. However, in the five-year term, 38% of air carriers announced investments in the amount of $1-10 million for each aircraft.

Until February 2019, when airlines invested in technologies related to aviation (connected technologies), it was primarily about providing satellite communications and Wi-Fi. Now, companies are ready to benefit from the data that they can obtain by using equipment directly on board aircraft. For example, such data can provide them with savings of 1% of fuel consumed, which is equivalent to $50,000 per aircraft per year, Honeywell analysts calculated. More details here.

Big data in logistics

• How to use Big Data for Forecast Analytics in Logistics

See also

• The catalog of BI solutions and projects is available on TAdviser
• Big Data
• Big Data in Russia
• Big Data Global Market
• Big Data: First Totals