How to address the multiple challenges of traditional modeling (see MODELING SOLUTIONS: WHEN and WHY TYPICAL MARKETING MODELS DO NOT WORK.)? There is not one silver bullet for all of them; however, we may offer better solutions for each of those and other problems than the traditional MMM modeler uses. The only way to do that is to accumulate a certain number of experts on one team, which allow AMModels to use the widest variety of different techniques. This is why we call the company Advanced Marketing Models, where Advanced means our ability to customize our modes depending on the client’s needs in a uniquely wide range of existing methods and approaches.
This is why we have gathered a group of professionals (see OUR CORE TALENT) who can critically analyze the nature of the client’s problem and find the optimal ways to model it. Personal skills, expertise, and participation of our experts in corporate projects allow them to be on the cutting edge of the latest achievements in their field. The regular screening of published books, articles, reports, and presentations is part of our professional life.
Our professional ethics includes the following:
- We are proud to state that our models are correct as much as could be achieved in a concrete project. We guarantee that all assumptions used in modeling were quantitatively evaluated and deemed acceptable. Our client gets all required estimates and calculations proving our assumptions. We work together on it.
- We do not use automatic black-box software solutions or hidden “proprietary know-how algorithms”, as some modeling companies claim. We use only open-published results, preferring works and accomplishments of our team members and the latest available versions of software.
- We guarantee that we apply maximum professional efforts to make models as close to reality as possible
- not more but not less than that.
- We follow up our models as long as a client needs it without any additional payments.
The types of problems and relevant types of models we can work on are the following:
- Marketing Mix Models
- the model of sales as a function of many exogenous and endogenous data such as syndicated scanner and audited data including trade, distribution, all advertising and consumer promotion activities, economic, socio-economic, weather, holidays, and other data taken on a weekly basis;
- the further calculations of different kinds of marketing elements such as incremental volume, general Return on Marketing Investment (ROMI) and ROMI by trade tactics, and others;
- the optimization of marketing spending allocation for the next fiscal year;
- the comparison analysis of marketing effectiveness for the brand/portfolio and competitors;
- the models may be built in Volumetric, Dollars, or Market Share forms. Models provide the required level of disaggregating (markets, size, flavor, etc.);
- the set of models for the client portfolio allows simulating a variety of marketing scenarios and conditions, including competitors’ activity and portfolio/brand pricing analysis.
- Direct Marketing Models
- the model formulating the rules for strategic segmentation, loyalty and behavior analysis and helping to create the marketing plans for the different groups of population;
this class of models allows improvement in customer relationships and loyalty profitability programs.
- Data Mining and Data Transformation
- a wide class of problems, related to classification and analysis of large data arrays;
- Choice Models - New Product Marketing Analysis
- a class of models, including Conjoint Analysis and Discrete Choice Models, which describes the situation of launching new products or multicriteria estimation, when many attributes are associated with human preferences and new product advantages;
- the model shows, before the real sale will take place, how price, package size, or flavor of the new product affects the sales.
- Marketing Optimization and Simulation Models
- the model allows finding the best combination of efforts and resources to obtain some goal with constraints;
- the model indicates how to minimize expenses or maximize sales, investing in the set of advertising channels knowing budget limitations in advance and using results of MMM.
- Agent Based Models (ABM)
- this class of models represents a very new and promising area for marketing;
- instead of statistical analysis, ABM creates “agents” (like customers, brands, channels, etc.) who can play the endless variety of roles and who evolve into complicated system themselves, without centralized management;
- ABM requires setting up certain rules for agents’ interactions, reflecting the essence of the business, and then lets the agents work until some stability emerges;
- ABM uses all available actual statistical data to calibrate and validate the built models to get a good fit on micro- and macro levels. In that sense ABM is similar to statistical models, but in a specific fashion.
- ABMs are especially good for simulating different marketing scenarios; in fact, no statistical model may compete with them in that virtue.