Enhance marketing decision making by answering the right questions

We are combining knowledge from multidisciplinary backgrounds with high-end analytics to solve real business issues. We often hear from clients that we effectively fill an analytical need in a way they have never experienced. Clients trust us to talk directly to end clients because we know how to take out the ‘geek-speak’ and we keep the business need top of mind.
The combination of tools at our disposal allow for unique customised solutions to answer your questions and optimise your decisions. The results are presented in an actionable What IF Tool. There is no one size fits all with our approach. We'll work with you till satisfaction.
Drivers of Brand Choice
  • By gathering data on perceptions of your brand and competitors on a set of key attributes, and also including a tradeoff task in the survey, we can obtain insights into:
  • The importance of your key brand attributes to customers
  • How your brand performs versus competitors
  • Areas your brand can own
  • How your brand can differentiate itself from competitors
  • The impact on your brand share of choice when key attributes are improved upon
Segmentation
  • We believe to be a true segmentation scheme, we must identify groups of customers or prospects that would respond differently to variations in the marketing mix based on varied behaviors or needs. The segments must:
  • Be real – ask yourself, “Have I seen this type of customers or prospects before?”
  • Be well-differentiated – ask yourself, “How could I approach each segment differently?”
  • Be identifiable and actionable – ask yourself, “Can the marketing team repeatedly and consistently find each segment and reach them through targeted media and strategic communication initiatives?”
  • The clustering approach we use for developing segments is a Bayesian mixed modeling method. It was developed in-house
Max Diff
  • Max-Diff is an indirect scale-free method for determining importance.
  • By selecting the most and least preferred, trade offs are forced and the data is used to model a relative rank ordering of preference. It is used to measure the preference for or the importance of multiple items such as:
  • Attributes
  • Brands
  • Advertising claims
  • Product features
  • Product benefits
  • Unmet needs
Craft a story
  • Story building is still based on choice methodology. Respondents are shown a subset of messages in each task but a major point of difference to get at “Story Building” is to question how messages best support each other to directly identify potential positive interactions in building a story flow. Also, respondents decide when they feel messages do not improve the story anymore.
Data Fusion
Data fusion is the merging of primary customized research with existing customer data and it allows to link choice models to internal Customer Relationship Management (CRM) systems to increase the power of “revealed data” (customer data) as well as understand the change between past and present behavior and predict future adoption for very powerful and effective target marketing.
Experimental Design
  • We have a perfect track record of mistake-free experimental designs for our research projects. This is a critical aspect because collecting the wrong data based on a poor experiment sullies the entire project--it is not something one recovers from.
  • We have developped our own Bayesian code over the last two decades to create those optimal experimental designs.

Toolbox

Some form of discrete choice modelling, DCM, is usually at the centre of everything; it's our main tool but we spin it in all directions. We build our own Bayesian code to derive every step, starting with data collection, the choice experiment (DCE), then the estimation of choice models, segmentation, Max Diff and finally the simulation of other possibilities with stand-alone software.
We claim to
  • come up with experiments that lead to more relevant models
  • estimate more appropriate models and
  • end up with simulation tools that are more user friendly and more powerful than anything else we came across
We make it all evolve every day by solving new problems.
At the core, choice experiments or DCM is about getting people to face different options and make a choice. The question is not necessarily "which would you buy?" and that is where creativity comes into play to solve different business problems or gather better data. Ultimately, we believe that collecting 'choices' lead to reliable data simply because in real life, people make choices every day.
This methodology is not just a hammer to a nail: with the right spin it is a complete toolbox.
  • Discrete Choice Experiments
  • Discrete Choice Modelling
  • Max Diff
  • Segmentation
  • Data fusion
  • Bayesian statistical modelling
  • WHAT IF Tool