Use of modelling and forecasting methods

Management
Attribution of TV placements

Strategic partnership between ADV and Beeline. Building a predictive model of conversion using the Beeline TV subscriber panel creates opportunities for a detailed analysis of the interconnection between exposure to TV advertising and online and offline purchases, which in turn leads to optimising further advertising campaigns and increasing advertising ROI by 15-20%. Beeline TV has 1.9 million set-top boxes and 1.5 million mobile Beeline subscribers in 900,000 households.

Digital attribution

Optimising digital placements by evaluating the effectiveness of each point of contact between customers and the brand’s advertising. Identifying the importance of the point of contact in the consumer journey, optimising further placements to maximise returns from advertising budgets.

Econometric modelling

It is used to forecast sales and price elasticity, analyze the contribution of each type of media in sales, identify the media targets of advertisement placements, analyse and forecast the effectiveness of trade marketing activities, evaluate the halo effect and cannibalisation between SKUs, and carry out real-time scenario planning. The tool Havas Village makes it possible to automatically evaluate the real-time impact of TV commercials on website visitor counts.

Scenario modelling using dashboard

The development of fully customized data and reporting visualisation systems. Predictive econometric models can be developed and embedded in the dashboard. The models allow future advertising activities to be planned without any constraints on the values of variables in real time without requiring input from an agency.

Cluster and effectiveness analysis

Increasing the effectiveness of dealer and retail marketing. Building analytical models to evaluate the effectiveness of granular media investment (e.g. regional dealers). Effectiveness analysis, optimisation of further placements using current data.

SMART GEO by Arena

Increasing traffic due to new geotargeting. Distance-based district segmentation. Evaluating the potential of attracting consumers in a district depending on various factors: population size, area, distance, logistics and competitor influence. The tool makes it possible to optimise all marketing tools to be geo-tagged (outdoor, online, etc.). Measuring the effectiveness of each tool depending on the district and share of consumers coming to shop. Optimising the locations of media placement to maximise the impact in terms of increasing traffic.

Proprietary forecast models for media indicators

Use of proprietary, accumulated data about media consumption dynamics coupled with advanced forecasting algorithms ensures highly accurate precision during campaign placement.

Media planning using machine learning

A component of the Aizek platform that uses a complete set of industrial data about media consumption on TV and in digital. It automatically searches the optimum media plan of placement based on pre-set inputs (TV, OLV, TV+OLV) by looking through hundreds of thousands of advertisement inventory purchasing combinations. It finds the best option in terms of media effectiveness and costs. More about Aizek.