Best-in-class marketers need a holistic understanding of the value of marketing on business performance, and a toolkit that supports optimization, scenario planning and forecasting of future investments across channels and tactics.

At Gain Theory, we do not believe in a one-size-fits-all approach.  Instead, we offer a number of marketing effectiveness solutions, and select the appropriate approaches based on the business questions we are looking to answer.  Importantly, we have extensive experience integrating disparate approaches into unified effectiveness solutions.

Gain Theory’s overriding mission for any engagement is to make a meaningful difference.  There is obviously a lot that sits behind this – models must be robust and make sense, all available data should be tested, and we should push hard to find suitable proxies for hard to find data.

But fundamentally, Gain Theory is passionate about making a meaningful difference to business performance – to improve marketing efficiency, to grow market share and to see baseline profits and share price grow.  Our goal is to deliver ongoing decision making support that drives real change, rather than a report that ‘collects dust’ on the shelf.


Leading Indicators

We live in a world where everything from the economy to a tweet can shift our priorities and behaviors. Insights that were valuable last year are far less significant today and may be useless tomorrow.  Hidden amongst the data available to marketers today are the metrics that really matter – those Leading Indicators that are most predictive of future business outcomes.  Gain Theory’s Leading Indicator Framework is an early warning system, that uses Leading Indicators to know, early on, whether business objectives are likely to be met, allowing marketers to re-calibrate marketing plans as necessary.

Marketing Mix Modelling (MMM)

Marketing Mix Modelling (MMM) measures the impact of all key business drivers on business performance.  Gain Theory builds fully-specified models to include: Distribution, Base Price, Trade, Above the Line, Below the Line, Owned Assets, Earned Assets and Non-Controllable Drivers (e.g. Holidays, Economic Indicators, Competitive, etc.).  A unique aspect of Gain Theory’s modelling approach is the AdModel, which converts spend, ratings and impressions across all media into a common currency – ‘Effective Cover’; this allows us to test a range of theories around the number of campaign exposures required to illicit a consumer response (drive sales), and over what time period… providing insights that are incredibly powerful for media planning,

Integrated Marketing Response (IMR)

We know a consumer’s path to purchase is complex.  While Search may be the final step on the conversion path, TV may have played a vital role earlier in the process, making Search more effective. Likewise, Out of Home (OOH) advertising may not lead directly to an incremental sale, but may amplify the effects of other media, both online and offline.  Integrated Marketing Response (IMR) is Gain Theory’s proprietary approach to understanding the consumer path to purchase and the role that marketing plays across the ‘funnel’.  Model output is used to optimize marketing investments, while accounting for direct and indirect effects, and short- and longer-term marketing impacts.

Digital Attribution

For organizations that invest in digital media, there is often a need to assign conversion credit to the digital channels of influence. Digital Attribution leverages user-level data and logistic regression to decompose Digital’s impact.  Because of the sheer volume of observations available in Digital, channels, publishers, sites, tactics and creative, and combinations, can all be explored using this approach.  Models go beyond the ‘last click’ to identify what actually drives conversion. A ‘retrospective attribution’ provides a learning layer around what worked and what didn’t, and Attribution outputs are combined with known budget constraints to inform future Digital investments.

Multi-Touch Attribution (Unified Measurement) – Inclusive of User / Event- Level Data

User-level Digital Attribution provides granular Digital insight, but rarely includes information from the offline world (where user-level data can be limited).  Unifying online and offline insights is crucial, as consumers are exposed to both online and offline messages, and often, interaction effects between these channels occur.  Our Multi-Touch Attribution solution (MTA) brings the world of event-level data and aggregate-level data together. It enables one analytical framework across all marketing channels, causal parameters and KPIs. This means traditional channels such as linear TV, traditional causal parameters like store promotions, and traditional KPIs like offline sales are considered in the same model as digital impressions, clicks and e-commerce sales.  Insights  inform upstream strategic thinking and downstream tactical optimization.

Multi-Touch Attribution (Unified Measurement) – Without User / Event- Level Data

Recent debate suggests that we may be moving towards a world where cookie-level information is no longer available for analytics. In order to future proof our Attribution offering, we have developed Sensor.  Sensor is a Multi-Touch Attribution approach that takes advantage of the natural geo-spatial variations in media and sales to provide sufficient data to decompose sales into tactically relevant granularity. Sensor can be performed by modeling an entire population or by modeling groups of individuals (down to what is permitted by privacy regulations).  Importantly, Sensor output is timely (it can be delivered within 2 weeks of the close of the sales window), sensitive to near-term events (can use as little as 28 days of recent data), sensitive to geographic differences in driver performance, granular, and all-encompassing (includes off-line marketing and marketing and online and offline sales).

Propensity Modelling (for Programmatic and Non-Programmatic Targeting)

Our ‘Next Best’ solution pairs Propensity Modelling and Machine Learning to score individuals based on their likelihood to convert. The approach helps marketers identify the ‘Next Best’ target audience, the ‘Next Best’ communication channel and the ‘Next Best’ message to improve conversion.  Model algorithms can be incorporated directly into an automated bidding process, for real-time scoring and optimization.

Agent Based Modelling (ABM)

While Marketing Mix Modelling and Attribution are useful in determining marketing’s ROI, they are less useful at understanding dynamic markets experiencing systemic change. Agent Based Modelling (ABM) allows us to simulate the impact of changing consumer attitudes, beliefs and behaviors. Using ABM models, we can run a variety of scenarios and understand the likely impact on brand sales and market share. ABM gives brands the opportunity not just to simulate external impacts on audiences but also how audiences in turn interact with and impact each other.

Bespoke Analytic Solutions

We believe that great value can come from creating bespoke solutions that address client-specific business challenges.  Whether delivering data analysis, marketing system integration or business foresight, our focus is on uncovering high quality insights that translate into meaningful business improvements.


Analytic Workbench (AWB) and ROVA

The Analytic Workbench (AWB) and ROVA are proprietary analytic platforms, used by our internal teams across our global network. Both platforms host libraries of pre-defined but customizable algorithms, enabling teams to quickly and efficiently conduct a variety of analytic use cases.

Machine Learning

The complexity of consumers’ path to purchase has always posed a challenge for marketers looking to acquire, grow and retain customers.  The volume of data available, particularly at the customer-level, brings simultaneously more data to better understand this path as well as challenges in processing the data and building models quickly. Gain Theory has invested in Machine Learning to not only speed up the process but to also enable us to test many models and algorithmic techniques in parallel. Advanced algorithmic approaches leverage large-scale computer processing power to interrogate large, disaggregated data sets to uncover patterns that can classify and predict human behaviours.  Today, we have applied Machine Learning to solve the following customer-centric challenges: churn reduction, product cross-sell, increasing customer lifetime value (CLV), customer sentiment analysis, and clustering.

3rd Party Technologies

We believe that the best solutions can come from anywhere.  We work closely with our clients to ensure that the best available data can be turned into the most meaningful insight.  In some cases, this means partnering with 3rd party companies to bring forward best-in-class technology solutions. Our approach is to be data, technology and approach agnostic ensuring we find the right solution best suited to our clients’ challenges.


Gain Theory has a long history of providing marketing effectiveness services across a diverse array of clients and categories.  All of our learning has been consolidated into an anonymized Gain Theory ‘normative’ database.  Data has been categorized using a single, globally unified classification system, which allows us to pull ROIs by country / category and against a number of ROI-determining factors (e.g. new product vs. established, long purchase cycle vs. short-, market leader vs. smaller brand, etc.). When providing output for clients, we include these norms to help ‘contextualize’ findings.



Managing Partner at Gain Theory, Russell Nuzzo, has written a piece for the @ANAmarketers highlighting six things marketers can do to boost effectiveness by being more data-driven. Read it online here -


Find out how we can help, demo our products and get more information about our global cross-sector experience.


If you think you have what it takes to make a difference at Gain Theory, please check out our careers page.
© Gain Theory 2023. All rights reserved.