SPONSORED CONTENT 
Walker Lewis
Senior Director, Customer Success
Critical Mix
wlewis@criticalmix.com
415-203-7976
I need answers right now, how quickly can you turn a project around?
The on-demand economy, as epitomized by companies such as Instacart, TaskRabbit, Uber, Hotel Tonight and OpenTable, is revolutionizing commercial behavior precisely by focusing on the “right now.” They align with consumers’ growing demand for convenience, speed and simplicity.
On-demand shrinks the gap between what we have already and what we need right now. I own a car but sometimes I order an Uber because I’m downtown and I need a ride right now. Maybe you already run a monthly tracking study and you’ve been running that study for three years. Sometimes you might need to gather instant insights to new questions because you need those answers to make new decisions right now.
At Critical Mix, we often see companies turning to instant insights to make decisions at the speed of business. We developed a specific product, called kNOW Instant Insights, precisely for that reason: to help clients quickly generate insights and make fast decisions – within 24 hours.
Two days before the Super Bowl, we helped a client determine whether people would watch the game; whether they would watch from their home, a friend or relative’s home or a restaurant or bar; whether people predicted New England or Philadelphia would win; and whether people found commercials to be an important part of the Super Bowl experience. We delivered instant insights from a census-balanced, nationally-representative sample of U.S. adults exactly when our client needed those answers.
Use cases for micro-surveys and quick insights have always existed but it was difficult to support those “instant” needs. That’s no longer the case. When you need answers to make important business decisions right now, statistically significant results are available within 24 hours.
Steven Struhl
Owner, Converge Analytic
847-624-2268
Convergeanalytic.com
When will useful artificial intelligence get here?
The surprising answer is that it has been here for many years, but not in the form of generalized intelligence like a robot that does your job for you. Rather, it is more like a very bright point of light that zeros in on solving a specific problem. For instance, classification trees – CHAID, CART, etc. – make many sophisticated judgements in dividing a sample into subgroups and testing these groups for significant differences. This method dates back to 1990. Similarly, Bayes nets learn complex patterns in data and make sensitive readings of effects – and have for over 20 years. You can find dozens of problem-solving methods presented at conferences on AI, even in free programs like Weka and R. You just need to know what to look for and where to look.