Concentric Market v7.0: A Breakthrough in Simulation
Think for a moment about what it takes to simulate a market from the bottom up. You need to represent thousands or millions of people each with their own individual preferences, behaviors, and connections.
You need to position the simulated people in a social fabric: one or more networks of friends, family, and acquaintances. You need to layer the actions of companies, influencers, and organizations by creating billions of synthetic events: impressions of ads, conversations, online ratings, product launches, and price changes among others. And then you need to account for the uncertainty and noise inherent in each one of those components by running the model thousands of times and find the distribution of likely outcomes.
There are two difficulties in this endeavor: One is the difficulty of developing a theory that will account properly for all of these components. The other is to engineer an algorithm that will run in a reasonable time given the processing power available to us today. The latest release of Concentric Market v7.0, the seventh large release since we founded the company in 2010, is a major step forward on both accounts.
We enhanced the Concentric Market model to incorporate a new type of an attribute, to add new competitive dynamics, and to enable a flexible timeline for simulations. The new type of attribute, the subjective attribute, enables more detailed exploration of categories like entertainment, consumables, B2B, and politics. Competitive positioning is now enabled, so that you can evaluate the impact of competitive messaging, negative campaigns, and review sites. The timeline is now flexible. That means you can run simulations for a few weeks or for a few years.
The model changes are important yet incremental in nature. The redesign of the algorithm to optimize its performance is a disruptive leap. We sped up the execution of simulations to make them 10 to 20 times faster. The greatest gains in speed happen for the simulations that took the longest to run. For example, simulations that previously took an hour to run, now complete in three minutes. Over the course of the past few years, the major barrier to continuous usage of our technology was that it took too long for certain analyses to complete. Now, speed is no longer an issue.
It is in the DNA of Concentric to continue pushing the envelope on both the theory and the performance of the algorithms. Our latest release is a big step forward on each, and we are really only gathering speed in our strides. Everyone at Concentric is already working on what is coming next: more seamless workflows for collaborative analytics and distributed computation, both of which we hope would further lower the organizational barriers to usage of simulation.
We hope you join our group of users, if you are not one yet, and experience the efficiency and effectiveness gains of simulation yourself.
The Concentric Team