Marketers are facing new challenges every day in how they reach and understand consumers. In recent years, newly developed technologies have changed the way consumers interact and make decisions by adding new devices, channels and platforms for media consumption in addition to new forms of content generated at higher levels and greater quantities. Consumers divide their time and attention not only among different media channels, platforms and devices but also among sub-channels within the same media.
This complexity means that consumers now can switch seamlessly between an array of (digital) channels, connected devices and platforms, creating new, more complex, dynamic, variable and volatile purchasing paths. In addition to this, we’re also faced with changing consumer-purchasing decisions, behaviors and choices from what they do, how they do it, when and why, etc. Consumers are becoming increasing difficult to understand and predict with precision.
The speed, direction and magnitude of the changes in marketing and consumer behavior is creating a daunting reality for marketers. In order for marketers to thrive in this new environment, they will need to understand and embrace the key dimension of modern marketing—complexity. This complexity deals with interacting elements, their interdependence and connectedness, and the diversity of the elements. With so much going on, we need new and updated approaches, new ways of thinking and new perspectives on marketing, brand and business strategy in order to successfully understand and predict consumer decisions.
Traditional marketing science techniques (which involve modeling, simulation, and optimization) do not sufficiently address the increasing pace of change in this fractured and diverse marketing environment and the dynamic and complex nature of consumer behavior. Therefore, more advanced approaches are needed.
One of the best solutions to understanding the new way consumers interact and make decisions comes in the form of complexity science.
Complexity science is a combination of various frameworks and concepts—including behavioral economics, consumer theory, demand theory, complexity theory, chaos theory, systems theory, game theory, network science, interaction theory, cybernetics, nonlinear dynamics, emergence theory and more—from a variety of disciplines (physics, chemistry, biology, anthropology, psychology ethnography, economy, sociology, ecology, social science, engineering, neuroscience, cybernetics, management, etc.).
In simplest terms, complexity science strives to uncover the underlying principles and emergent behavior of complex and dynamic systems (like consumer markets), which are made up of many individual elements or agents (such as consumers).
One emerging field in complexity science that may help marketers better understand consumers is Agent Based Modeling (ABM). ABM is a more sophisticated, discrete simulation technique that integrates the areas of economics, marketing, artificial intelligence and engineering to study complex systems and consumer behavior with greater comprehensiveness, depth and detail over other possible modeling/simulation/optimization alternatives such as marketing mix modeling (MMM), choice-based conjoint modeling, discrete-choice modeling, Markov chain Monte Carlo and more.
ABM studies the interactions between people, things, places, and time using both computational and mathematical models. ABM techniques combine individual person decision and network rules to model behavior. ABM provides the ability to simulate the individual (micro) behavior of heterogeneous, autonomous and diverse agents (like consumers) in an environment, based on predetermined decision-making rules affected by external influences as well as defined properties of that agent. It then analyzes the overall (macro) dynamics and outcomes of the system (like a brand’s marketplace) that emerge, the aggregate behavior is simply the sum of all of the individual agents.
Structure of an Agent-Based Model
The three ideas central to ABM are objects, emergence, and complexity. A typical agent-based model has the following elements:
- • Agents, their attributes and behaviors
- • Agent relationships and methods of interaction (an underlying topology of connectedness defines how and with whom agents interact)
- • Agents’ environment (agents live in and interact with their environment in addition to other agents)
- • Numerous agents specified at various scales, typically referred to as agent-granularity
- • Decision-making rules or heuristics
- • Learning rules or adaptive processes
- • An environment
Furthermore, the fundamental building blocks in every ABM can be organized according to the PARTE framework: Properties, Actions, Rules, Time, and Environment. The first three elements (Properties, Actions, Rules) define the agents, while the next two (Time, Environment) define the context. However, there can be vast variation in the form that each element (P, A, R, T, E) takes from model to model. This flexibility is part of the power and potential of ABM.
Key Advantages of Agent-Based Models
In the highly dynamic, competitive and complex market environments of telecom, insurance, auto, health, etc., the consumer’s choice depends on a number of individual characteristics, inherent dynamics of the consumer, network of contacts and external influences that are best captured and understood within the ABM paradigm.
Advantages of Agent-Based Modeling and Simulation:
- • Incorporate substantial heterogeneity (among individuals) and complexity into the decision-making on the modeled consumers
- • Model the interdependent behavior of consumers with sufficient level of detail
- • Account for differences between segments based on the behavior of individual consumers
- • Describe how consumers sense and interact in the marketplace
- • Can effectively reproduce emergent behaviors
- • Understand the aggregate characteristics of the system that emerge from the actions, decisions and interactions of the individual agents
- • Simulates the actions and interactions of autonomous agents with a view to assessing their effects on the system as a whole
- • Show how changes in individual behaviors will affect the system’s emerging overall behavior
- • Explain the ways that complex systems adapt to internal and external pressures so as to maintain their functionalities
- • Predict, emergent marketplace behaviors (decisions and outcomes) that cannot be foreseen
- • Does not report a single future outcome of a given set of marketing decisions – rather, identifies a range of possible decision outcomes and gauges their relative probabilities of happening
- • Can directly incorporate sophisticated spatial elements, to effectively model dynamics that result from exposures across space and time (such as advertising), patterns of contact between individuals (influence through social networks), the impact of context on decision making, and geographic constraints on choice set
- • Can include structurally rich, dynamic, and heterogeneous representations of detailed social network structures or environmental exposures and influences
- • Model the coevolution of environment and individual behavior, on potentially divergent time
- • Ever-changing marketplace
- • Growing number of platforms, devices, channels and applications
- • Volume, variety, velocity, veracity and variability of data
- • Fragmentation of the advertising/marketing tech stack
- • Continuously evolving consumer behaviors and needs which, depend on individual characteristics, inherent consumer dynamics, network of contacts and external influences
- • Unpredictable nature of the buyer’s journey
…is turning marketing into an extremely complex discipline, making it critical to address this challenge through a modeling approach that is dynamic, flexible, and adaptable.
We now have the ability to combine and use new, rich data sets, processing power, device-graphs, powerful computer-based models, whole-system multi-scale models, data-intensive science, machine learning, and artificial intelligence to describe and simulate the most realistic (real-world) consumer and marketplace complex dynamics to gain insight into and predict consumer behavior (with increased accuracy and confidence).
This makes ABM a more viable solution for improving decision-making effectiveness, and ultimately the basis of a Marketing Decision Support System (MKDSS) to help marketers:
- • Make smarter decisions about their audience through a segment targeting strategy, channels, content, and messaging (type, timing) to create a compelling, connected, meaningful consumer experience
- • Reach certain consumers, at certain times, in certain ways (in the right context), on certain channels, devices and platforms, with just the right message for their stage in their decision journey to drive cognitive, emotional, behavioral and transactional impact
- • Develop more effective and efficient marketing strategies that will have an enormous impact on advertising, brand and business growth
- • Increasing the predictive value (and reduce uncertainty) of their decision-making
Marketers should not let the model learnings completely overrule their human instincts and experience. They should use the model insights to inform their strategic decision-making. The human factor should remain one of the most critical parts. Modeling should not eliminate the need for strategic thinking; rather, it should fuel that thinking. Modeling should not negate the need for sound design and implementation of marketing strategy – on the contrary, it should elevate the importance of it. In other words, marketing leaders need to strike the right balance between and art and science of marketing (which is a journey of balance that requires continuous practice, continuous study) to make sound and rational decisions, which in turn leads to sustainable improvements in marketing performance.