VK Predict

The main articles are:

• Business Intelligence Definition
• Geoanalytics

VK Predict geoanalytics are client analytics services and decision support systems based on analysis of data arrays, machine learning technologies and artificial intelligence.

2024: Optimization of Telecom Radar service

VK has expanded the functionality of the Telecom Radar service, a solution for telecom operators that allows you to analyze the quality of communication and, based on this, make decisions on infrastructure modernization. The company announced this on November 27, 2024. The solution is available to broadband operators: they will be able to assess the share of the wired Internet market in a specific region and city, compare the level of services with competitors, identify problems with the Internet and improve the quality of services for subscribers.

𝓲

VK

The Telecom Radar service from VK Predict is used by cellular operators to analyze the quality of communication in relation to geolocation and comparison with competitors, as well as to determine market share in the region. Communication quality analysis is formed on the basis of video consumption - one of the most popular and heavy types of content, which imposes high requirements for network quality.

Telecom Radar takes into account the average video start time, average download speed, resolution and content freeze. Based on this data, the operator can evaluate the user experience of subscribers, real network coverage in different regions, conduct competitive analysis and, if necessary, modernize the infrastructure. This will increase the level of satisfaction and loyalty of subscribers and reduce the risk of their outflow.

Thanks to the high accuracy of geo-positioning by Telecom Radar, operators will be able to assess in which settlements and areas there are problems with coverage, check and, if necessary, plan the modernization of the infrastructure, and then assess the effect of its implementation. The service allows wired Internet operators to see abnormal problems with the quality of Internet communication, for example in apartment buildings, when one user overloads the network. And also analyze the dynamics of the number of subscribers and assess how marketing campaigns affected the market share, compare themselves with competitors and assess the prospects for expansion into new locations, "said Roman Styatyugin, director of the VK Predict analytical services center .

2023: GeoCursor ML service to select the optimal location for business

On April 24, 2023, VK (formerly Mail.ru Group) introduced the VK Predict geo-analytics ML service - GeoCursor. Retailers, banks, chains of pharmacies, restaurants, beauty salons and other companies can solve problems related to assessing the locations of their outlets with high accuracy. The service helps to search for a target audience, analyze factors affecting revenue, predict turnover, assess the competitive environment, develop territories and choose the optimal locations for business development.

Offline business development requires the analysis of hundreds of parameters. Unlike online resources and online stores, offline points have more limitations and less scope to scale. Analyzing all the necessary criteria manually is difficult and long. To do this, you need to take into account the indicators that can change every month. Technologies simplify this path for business, reduce the burden on employees and the risk of error, allow you to make objective decisions about business development and increase profits from new points, "said Roman Styatyugin, director of the VK Predict Analytical Products Center.

GeoCursor's machine learning models take into account more than 500 parameters: impersonal audience signs, district infrastructure, pedestrian and car traffic, purchasing activity. Based on them, analyzes and predicts sales information, turnover and average check.

Algorithms analyze data within a radius of less than 40 meters from a given location. This reduces the impact of side indicators on miscalculation. For example, it helps to separate traffic at a subway station or public transport stop from real traffic at a shopping center or a potential opening point for a new store.

The service forms a general portrait of users who are often near the selected point. It includes socio-demographic characteristics, income level, interests in various topics and other criteria. Information about the district's infrastructure helps to assess how close the location is to competitors, high-traffic facilities and transport hubs. Due to service analytics, the business can also predict the impact of a new store or department on the indicators of already open points.