Next Generation of Prescriptive Analytics: Market Simulation

This quote from an ancient Greek philosopher sums up the state of business, now more than ever. The pace of change is increasing for enterprises across a variety of industries, delivering an abundance of opportunity, but also driving more uncertainty than ever before.

Forward-thinking businesses try to combat uncertainty with data and analytics. Today’s analytics tools – business intelligence (BI), statistics, and machine learning – combined with lots and lots of data deliver insights that help companies of all sizes react to change faster and more effectively.

Predictive analytics have the ability to transform how companies manage their operations, make decisions, and deliver forecasts. However, most analytics tools use traditional statistical methods and rely solely on large quantities of historical data derived from the operation of the business, including sales, demographics, and other known outcomes. These analytical methods are great at finding the correlations inside the data, but the complexities of the market are lost when you only look backwards.

How can you truly understand a market solely on data, without understanding the dynamics of the people – consumers, customers, patients, guests, users, and other populations – making up that market? Without information on consumer sentiment and an understanding of how consumers interact and influence each other, you can’t. This is where market simulation enters the picture.

Market simulation is ushering in a new generation of prescriptive analytics by using a software platform to recreate how a consumer market behaves in a simulated environment. This includes the people, their sometimes-irrational decision-making, and how they interact by recommending (or bashing) the products they consider. In addition, a market simulation can include competitors, different product choices, marketing touchpoints, and how macroeconomic conditions affect market dynamics.

The end result is a new breed of prescriptive analytics that allows enterprises to make better-informed decisions by asking what-if questions in a realistic simulated environment.

But first, let’s get a baseline understanding of simulation technology.

What is Simulation?

At its simplest, simulation recreates a system or a process in software, allowing you to test different scenarios and forecast outcomes in this simulated environment.

In use for decades, simulation has benefited a number of different industries and applications. Meteorologists use it to forecast the weather. Pilots use it to safely practice flying. And companies of all sizes use it to optimize operations like an assembly line or quality processes.

Today, simulation has emerged as the next generation in prescriptive analytics by answering “what-if” questions at the speed of business. In fact, Gartner Research recommends simulation for strategic business analysis to drive growth, reduce risk, and improve business efficiency.

Why Now?

Market Simulation is gaining traction for advanced business analysis for two reasons – the maturation of the analytics market and advances in technology.

For the past decade, companies have been organizing their data, building dashboards, and proving the value of analytics for driving insights. User-friendly business intelligence (BI) and machine learning tools help more users deliver insights from historical data, driving value deeper into the enterprise. But now, many business leaders are asking: WHAT-IF I take this action based on these insights?

At the same time, advances in machine learning and cloud computing have sped up these analytics solutions to the point where insights are delivered at the speed of business, allowing both analytics teams and business users to make decisions much faster than before.

All analytics are aimed at reducing uncertainty in order to make better decisions. But unlike traditional analytics methods, market simulation incorporates the dynamics and complexities of the environment being simulated. With the advances in cloud computing and the maturity of how companies collect and manage data, simulation technology can now be applied to business analytics, when statistical and machine learning methods reach their limitations.

Analytics teams are consistently faced with trade-offs – no one model works best for every analytics challenge and different metrics are required to answer different questions. For example, should I focus on click-through rate (CTR) or conversion for my marketing analysis? In choosing my model, is accuracy, speed, or interpretability more important? For the best results, and deepest insights, the analytics team needs to use the right methodology to answer the question at hand.

Market Simulation vs. Other Analytics Methods

Traditional analytics methodologies like statistics, regression, and machine learning reveal correlations, or patterns, in data. Machine learning, today’s go-to method for prescriptive analytics, starts with pushing a lot of data through an algorithm, which learns the patterns and delivers a model you can use to make predictions about future data. Machine learning is ideal when you need to answer questions like “Is this person likely to default on a new loan?” or “Which subscribers are likely to churn?” In general, the more data, the better the accuracy of the predictions.

With machine learning, the insights developed allow specialized analytics professionals, such as data scientists and statisticians, to interpret the predictions and develop theories on how their market works in relation to the questions they ask, and ultimately to make decisions that affect business outcomes.

But market simulation is different. It begins with domain expertise to develop a model of the market in which you operate. Data is used to calibrate or tune, the model so that you can develop the most accurate representation of the market you are trying to analyze. The market simulation can then be used to answer many strategic go-to-market questions. The key piece of the puzzle is the inclusion of causal factors of human behavior, including influence, perception, and network effects like word-of-mouth. With market simulation, it’s the level of domain expertise that improves the model, not just the amount of data.

Market simulation works well when you have complex questions, limited data, and high uncertainty or volatility. Questions like “what happens to revenue if I change my marketing mix?” or “how do I launch a new product in the most effective way?” or “what if a competitor enters the market with a new product that undercuts our pricing?” or “how do I improve my customer experience?” and many other market-centric questions are commonly addressed through market simulation.

The Science Behind Market Simulation

Today, Concentric offers the only market simulation platform for delivering trustworthy answers to business questions across almost any market. This broad capability is enabled by the integration of five different sciences – Agent-Based Modeling, Behavioral Economics, Network Science, Marketing Analytics, and Machine Learning – into Concentric Market^®.

Agent-based modeling (ABM) is a methodology that represents people as diverse individuals. With ABM, you model the properties and behaviors of the individual participants in an environment – consumers, customers, visitors, patients, voters, drivers, etc. – to forecast aggregate outcomes.

These outcomes can be market share, revenue, technology adoptions, audiences, votes, or traffic jams. ABMs are common in astronomy and physics as a way to analyze galaxies, stars, planets, atoms, and elementary particles. Recently, ABM has moved into mainstream economics and social science applications, and now Concentric is making ABM accessible in the business world.

How To Simulate a Market

Define Your Market

The first step in populating a market simulation is to define the players in the market, including the competition. For every decision a person makes, they choose between two or more alternatives. These may include competing products, services, brands, or the choice to do nothing at all. For example, a new TV streaming service would identify companies like Netflix, Amazon Prime Video, and Hulu as direct competitors along with cable, satellite, and premium channels. However, it’s important to also consider that a consumer may opt out of watching any programming.

Create Simulated Consumers

Next, you need to recreate how your simulated consumers make decisions by accounting for their current preferences or perceptions in the market. You may even account for how different segments of the population have different preferences. Many companies use market research, segmentation studies, or other business intelligence to weight the factors people may care about, such as price or quality, convenience or speed.

Incorporate Factors That Influence Decisions

Next, you incorporate all the other factors that influence individuals’ choices. Oftentimes, these factors include the marketing activity of a particular company, competitive actions, and in some cases, word-of-mouth.

Train the Simulation

The goal of creating a market simulation is to mirror the real-world as accurately as possible. In order to do this, the simulation must be calibrated by adjusting settings until there is enough confidence that the simulation can reliably predict future outcomes. This is an on-going process since there will always be new data and learnings that need to be added to represent an evolving market. People aren’t static, so a simulation can’t be either. Over time, this allows the simulation to become even more reliable in predicting outcomes.

Run Scenarios

With a calibrated simulation model, you can begin experimenting to see how various product updates, pricing strategies, marketing campaigns, and other factors can influence consumer decision making. Ultimately, the goal is to identify which strategic “levers” to pull to increase the likelihood that consumers select your product over other offerings.

For example, a car manufacturer may seek to understand the impact of adding TV monitors to its line of minivans. Through market simulation, that manufacturer can clearly see the impact on sales and may decide to invest in TV monitors or pull a different lever altogether, such as safety features or other amenities. Companies are also able to leverage market simulation to better understand the impact of scenarios they don’t have control over, such as a new competitor entering the market or the rise of gas prices.

Market Simulation in Action

Want to know how marketing simulation is used in the real world? Here are a few deeper dives into some of the use cases for marketing simulation.

Marketing Mix Optimization

“Half the money I spend on advertising is wasted; the trouble is I don’t know which half.” John Wanamaker coined this famous phrase at the turn of the last century, but it still rings true. Companies have a difficult time determining which marketing programs produce results because of how marketing programs affect and influence buyers. With market simulation software, companies can build their own marketing mix and attribution models in-house, as well as forecast the impact of their optimized media plan – instead of waiting months for a report.

Customer Experience

Understanding how customers perceive a business, and how this perception affects revenue, is difficult or impossible to achieve with traditional analytics solution. Word-of-mouth, recommendations, social influence, and other very human factors make it difficult to capture and analyze data with machine learning. Market simulation is built to include these factors, offering an ideal platform for asking questions around how perception will increase visitors, drive sales, and improve perception.

Pricing

How do companies know what price is best for reaching their revenue goals? Many use a cost-plus model where they take the cost to produce and sell the item and add some percentage for profit. But this could lead to pricing miscues where they are leaving profit on the table or they price themselves out of the market. With market simulation, you incorporate the dynamics of the market – competition, consumer perception, and bigger industry trends like the closing of brick-and-mortar stores – to create pricing strategies that maximize revenue and perception.

MARKET SIMULATION

The Next Generation of Prescriptive Analytics.