Scale. Depth. Transparency. Advanced Analytics that perform.
Our advanced analytics software is infused with cutting-edge, innovative algorithms that can help you solve even your most intractable marketing and sales problems. Make the best decisions possible, and identify opportunities and optimize outcomes you might otherwise miss.
Data Visualization & Analysis
Our Data Visualization and Analysis module provides a modern, integrated environment for advanced marketing analytics. Users, including those without advanced analytic skills, can examine and understand the sets of visuals and statistics needed to both verify the data and form initial insights and hypotheses on what's driving demand.
- Ensure data quality and alignment
- Comprehend information quickly
- Pinpoint emerging trends
- Identify relationships and patterns
- Communicate story to stakeholders
The core modeling module provides a comprehensive user interface (UI) for specifying a range of dynamic linear panel models at all levels of the client business. User-specified model structures are integrated with the model database to perform any required variable transformation which is then transferred to the proprietary model estimation algorithm housed within the Oxmetrics analytics package. Results from Ox are then returned for viewing in the UI with full graphical interface.
The modeling process comprises two unique elements:
- Network models of consumer demand estimated using Dynamic Hierarchical Bayesian econometric techniques for robust short-term price and media effectiveness across products, brands, channels and regions.
- Fully grounded in the economics of consumer choice theory – with both single equation and discrete choice demand system approaches
- Fully captures the dynamics of pricing decisions and media investments
- Unique blend of unobserved components modeling and HB techniques
- Network structure fully captures the off-online consumer purchase funnel
- Supplemented with Latent Instrumental Variable techniques to deal with endogeneity (selection) bias endemic in online search
- Assigns correct credit to each touchpoint on the path to purchase
- Dynamic modeling from market level to individual SKUs, helping to overcome dimensionality problems that often plague standard approaches.
- Covers a range of potential marketing response functions
- Aggregate cointegrating VAR analysis for long-term brand-building
- Network models of long-term base sales, consumer brand perceptions, paid and earned media
- Incorporates equilibrium (cointegrating) relationships between the variables
- Impulse response analysis uncovers long-term (persistent) effects of marketing investments
- Uncovers the true drivers of long-term brand building.
Full System Decomposition
The Full System Decomposition provides an elegant and automated solution to the challenge of turning a given econometric model result into a complete model report. The module calculates all the net parameters of a full system of linked models based on the estimated off-online marketing funnel structure and integrates these with the database to create the full report and visuals.
- Fully automated, allowing users to quickly sense-check model outputs
- Single equation and multi-equation panel decompositions
- Accommodates additive and multiplicative model structures
Simulation & Optimization
The Simulation and Optimization module supports clients with advanced and robust methods for optimizing marketing investments. Our Simulation Optimization approach integrates model results with a series of factors that future-proof results for greater planning accuracy. The web based infrastructure supports multiple users in a safe, secure and credentialed environment.
- Provides a potent set of optimization and simulation techniques to identify the actions that will produce the best results, while operating within resource limits and other relevant constraints
- Marketing organizations can consider more alternative actions and scenarios, and determine the best allocation of resources and plans for accomplishing objectives and managing activation
Forecasting & Reporting
The Forecasting and Reporting Module rounds out the software studio with a comprehensive tool for monitoring in market performance of the model in terms of ongoing forecast accuracy. In addition this provides an ongoing decomposition and indication of the impact of driving factors during the forecast period thus supporting ongoing diagnosis of performance.
- Ideal approaches for Marketing, Sales and Ops forecasting and collaboration
- Seamless forecasting by product, campaign, region or any other dimension
Developed in conjunction with Siem Jan Koopman, professor of econometrics at VU Amsterdam
“The application of state-space methods to Consumer Demand estimation has been a very interesting and stimulating project for me. Integrating the Hierarchical Bayesian component is a great advance in the deployment of these methods to large scale datasets. Working with Peter and the Marketscience team and seeing this deployed in the commercial space has been extremely rewarding”. - S.J. Koopman
We are proud of our association with Siem Jan and have a great working relationship in bringing this exciting science to the business community.
Intel selected Marketscience to revamp their MMM solution and deliver a more transparent, comprehensive analytic solution.
“The market.science studio and accompanying team support provided us with the necessary quality assurance to stand up to our stakeholders and partners in finance and marketing. This project was critical to re-establish trust in MMM. It also led to the successful adoption of the recommendations to improve performance”
- Yerjanat Khabai, Marketing Data Science and Analytics Manager, at Intel
For client Unitymedia, Truesight Marketscience designed and implemented a multi-phased marketing analytics program that would measure the impact of investments and other sales drivers.
“We had an immediate need to transform into a data-driven organization while at the same time, optimizing spend to maximize effectiveness and enhance collaboration between our departments. We also needed a partner that would train, fully enable and then support our in-house analytics experts to own, manage and operate the full analytics process once the project was completed.”
- James Oliver, Data Scientist Team Lead at Unitymedia.
PwC Partner Review
Market.science studio was selected by PwC after a competitive review
“It was the only solution that provided the necessary comprehensiveness to inform the complex financial models of the demand-side. Given our reliance on these models to make large investment and accounting recommendations, the academic and professional backing of the solution provided us with the leading edge and necessary support”.
- Dr. Jonathan Gillham, Director of Economic Modeling and Econometrics at PwC