Коэффициент угрозы Covid-19
Коэффициент угрозы - это показатель, который вряд ли можно найти на других ресурсах. Этот коэффициент имеет простую формулу - он рассчитывается как отношение позитивных результатов тестов на выявление инфекции COVID-19 к общему количеству таких тестов.
How and where can big data be used?
Politics After the successful victory of American President Donald Trump in the 2016 elections, the Internet blew up the term Big Data as something that changed the traditional approaches in the election race and showed the colossal effectiveness of the election campaign. The main point was that a group of scientists was able to segment voters based on data from social media profiles. For each group, its own electoral strategy was developed, corresponding to the opinions and principles of voters. In other words, information with all the advantages of the future president was presented to the audience supporting Trump, and information about the shortcomings of competitors was presented to the opposing audience. The effect was obvious - in the first case, there was an increase in Trump's electorate, in the second, a decrease in the electorate of competitors.
Retail Currently, the main task of selling companies is not just to sell now, but to retain the client so that he will come later, and more than once. The high competition of companies in improving the quality of service and services for the client has formed a new direction - direct marketing (a type of marketing communication, which is based on direct personal communication with the client). Simply put, the apogee of direct marketing is the following situation: A customer walks into a store, approaches the reception and thinks to buy an umbrella. He walks up to the counter, the manager greets him, calls him by name and immediately offers several umbrellas to choose from. Sounds like a fairy tale? However, not at all - the client came in, the cameras recognized him, sent a signal to the central server, the server retrieved the history of the client - calls, purchases, wishes, preferences, hobbies, in general, everything that is. Then the data was transferred for processing, the predictive modeling system calculated that the person most likely came for the umbrella and transmitted a message to the reception to the manager, who had many times greater chances of selling the umbrella to the client. And the client, most likely, will come again and remain loyal to the company for a long time.
Business It does not matter which one - production, service, construction, financial, insurance - any business in the course of its existence accumulates a certain amount of data. Very often such data are collected in thick heaps of archives, electronic at best, which can only occasionally be useful as reference information. In fact, this data is much more valuable than just a reference, because the history of documents and decisions reflects all stages of the company's development. This data stores information about all management decisions, victories and defeats, successes and failures. Correct decoding of such data will make it possible to assess the consequences of previously adopted management decisions, to avoid mistakes that were previously made, but remained unnoticed. In addition, by linking the dynamics of the company's indicators (profit, customer activity, costs) with external factors (gasoline price, dollar exchange rate, political events, individual geolocations), it is possible to identify a number of trend dependences that can be used to predict future activities. Modern technologies make it possible to carry out intelligent analysis of big data and reveal non-obvious (hidden) dependencies of factors that affect the company's success. Modeling such factors, using powerful intelligent systems and algorithms, makes it possible to effectively plan the operational activities of the company, which has never been under the control of an ordinary person.
How it works. Example for a bank - Probabilistic models of customer behavioral factors. Understanding general trends will allow you to more effectively implement the strategy, and anticipation of customer actions will provide an opportunity to improve communication and improve the quality of marketing; - Forecast of the elasticity of demand for banking products; - Recommendations for pricing; - Recommendations for customer loyalty; - Recommendations for new banking products; - Detection of systemic anomalies in transactions to prevent abuse and fraud, identify inefficiencies in business processes and improve security; - Factors or groups of factors that, to one degree or another, affect the financial results; - Identifying trends.
How to segment customers - by loyalty to the company (how long and how often they use the company's services). It will be useful for the formation of individual loyalty programs and promotional offers; - by solvency. It will be useful for approbation and implementation of new products, partner sales, formation of pricing policies - by profile. Creation of customer profiles (for example, to whom, how often and in what amount have to make payments for insured events). Identification of the so-called unprofitable clients for the company; - by location (how customers are distributed on the map). It will be useful for the development and optimization of the agent network. - Optional; - by sales channels (efficiency of sales channels - through agents, call-center, sms-marketing, e-mail, instant messengers, etc.). It will save time and resources when communicating with the client.