Dynamic marketing mix modelling

The marketing mix model is a widely-used tool to evaluate sales drivers and Return on Investment (ROI) of the marketing budget. Two key elements lie at the heart of the process: economics and econometrics.

At Marketscience Consulting, we treat both with equal importance.

Economics

Dynamic marketing mix modelling Marketing mix analysis is based on economic models of product demand, quantifying consumer response to marketing investments. Consequently, accuracy will depend on choosing a suitable demand model.

We work with you to build the most appropriate model for your business, ranging from discrete choice customer level analysis and single equations of product demand through to aggregated simultaneous equations of brand choice.

Our models reflect the full range of traditional and digital marketing investments, together with off and online media interactions, pricing, promotions and external economic drivers. More

  • Price and promotion 

We offer a complete price and promotional modelling service, addressing the efficiency and effectiveness of a range of promotional tactics covering price cuts, display, features and coupons at varying levels of data granularity.

Aggregated market level studies provide a broad view of  promotional effectiveness relative to above the line media investments. For more detail, account specific and store-level models across products and brands are ideal for accurately quantifying price response and promotional impact across the full range of tactics

Where the business focus is on highly detailed pricing strategyhousehold panel data, customer level and consumer conjoint analysis is preferable. This type of study provides information on specific aspects of consumer price and promotion response, ranging from product loyalty and switching behaviour, within-brand product cannibalisation, stockpiling and store switching through to willingness to pay.

At all levels, topics such as synergy, halo and cannibalisation effects across the product portfolio are thoroughly explored as part of all model structures.

  • Offline and online media

Traditional media investments such as TV, print and magazine play a key role in raising brand awareness and driving business outcomes. Digital marketing investments such as online display, paid search and social media typically impact outcomes via online natural search and webtraffic activity.

Our advanced model solutions fully incorporate traditional and digital media impacts for both off and online sales. Key features such as media and promotion synergies, halo, cannibalisation and category expansion effects are integral to the model structures.

Econometrics

Dynamic marketing mix modelling

All marketing mix models employ econometric techniques to estimate the sales response parameters. However, standard practice does not fully reflect the dynamic time series nature of the data. More

At Marketscience Consulting we undertake bespoke analyses of your business and data, providing a simultaneous decomposition of sales into key components such as long-term dynamic trend, season, cycle and off and online marketing.

This approach enables precise quantification of short to medium-term marketing ROI, accurately captures evolutionary long-term behaviour and leads to superior forecasting and simulation.

  • Marketing ROI

Dynamic marketing mix modelling

Quantification of the ROI of marketing investments requires an accurate decomposition of sales into base and incremental volume.

Base volume represents the long-run or trend component of the data. Incremental marketing volume captures short-run variation driven by promotions and media investments. Conventional marketing mix modelling imposes a fixed baseline – giving an artificial split into base and incremental volumes.

The Marketscience Consulting dynamic time series approach allows a more precise decomposition and more accurate marketing ROI measurement.

DECOMPOSITION

Dynamic simultaneous equation systems combine all promotions, offline and online media to provide a complete quantification of traditional and digital incremental marketing ROI. 

The long-term baseline, modelled simultaneously alongside the incremental components, fully reflects any evolutionary trend component that exists in the data.

  • Long-term behaviour

Brand Market Trend

Short to medium-term marketing effects are only part of the story. Clients typically need to quantify the long-term brand building effects of their marketing campaigns. 

Persistent long-term effects reside in the trend component of the sales series, impacting the evolution in base sales over time. The fixed baseline of the classical approach thus precludes long-term analysis by construction.

The evolving baseline of the dynamic time series approach is a pre-requisite to quantification of long-run consumer behaviour and ROI – and lies at the heart of brand equity modelling.