FEI IT, Seattle March 4, 2010 Gaming the Forecast Leveraging Collective Intelligence for Real-Time, Accurate Forecasts Mat Fogarty, CEO Crowdcast © Crowdcast, Inc. 2010 Speaker Background • • • • • • London School of Economics, Management Science Unilever, Finance Manager, CIMA UC Berkeley, MBA Electronic Arts, Director of FP&A Founded Crowdcast in 2007 San Francisco, Venture backed © Crowdcast, Inc. 2010 Crowd Forecast = Crowdcast • Forecasts are often subject to bias, stickiness and gaming • Crowdcast leverages wisdom of employee and partner crowd • Based on a prediction market • Produces unbiased, real time forecasts of key metrics • Valuable for new products, key projects, and external metrics © Crowdcast, Inc. 2010 Costs of Poor Forecasts © Crowdcast, Inc. 2010 What is a Prediction Market? • Based on stock market or betting market • Aggregates knowledge of diverse group • Players weight forecasts • Rewarded for early, accurate predictions • Anonymous and meritocratic • Accurate players given more weight over time © Crowdcast, Inc. 2010 Actionable Intelligence Intelligence Insight Crowdcast Platform © Crowdcast, Inc. 2010 Action Actionable Intelligence Intelligence Insight Crowdcast Platform When will new product ship? © Crowdcast, Inc. 2010 Action Actionable Intelligence Intelligence Insight Crowdcast Platform When will new product ship? © Crowdcast, Inc. 2010 50% chance late “Supplier problems” Action Actionable Intelligence Intelligence Insight Action Crowdcast Platform When will new product ship? © Crowdcast, Inc. 2010 50% chance late “Supplier problems” Change Supplier Actionable Intelligence Intelligence Insight Action Crowdcast Platform When will new product ship? © Crowdcast, Inc. 2010 10% chance late “New supplier rocks” Change Supplier Examples of Forecasts • • • • When will the product launch? What will be the product’s quality score? Sales units in first 90 days? Sales at $39 vs Sales at $49? Project Management • • • When will the next milestone be complete? How much will the project cost? Will the project achieve its stated goal? Industry Intelligence • • • Will a regulation pass? What will our market share be? Will our competitor launch a competing product? New Products © Crowdcast, Inc. 2010 Use cases • 400 employees forecasting success of game projects, 32% more accurate than official • Panel of scientists and doctors predicting FDA decisions & sales, more accurate 86% of time • Diverse team forecasting tech trends, demand, competitor actions and regulatory changes • Employees and customer community forecasting success of new products © Crowdcast, Inc. 2010 SCREENSHOTS © Crowdcast, Inc. 2010 Player – Select Question • Employee scans available forecasts and chooses one where he has special knowledge • When will the Beta of the Tablet Computer launch? © Crowdcast, Inc. 2010 Player – Enter Bet • Employee chooses a range of likely outcomes • Weights forecasts with virtual money • If correct, the bet will payoff • All input is anonymous © Crowdcast, Inc. 2010 Player – Add Comments • Why does the employee think this? • What is the underlying reason and assumptions? © Crowdcast, Inc. 2010 Player – Leaderboard • Leaderboard shows how accurate the players are, also which department is more accurate • Competition, bonuses and other incentives encourage participation and accuracy © Crowdcast, Inc. 2010 Manager – Forecast Analytics • Early warning of issues © Crowdcast, Inc. 2010 • How does the Crowdcast compare to the target? Manager – Forecast Analytics • Understand most likely outcome • Spread indicates range of outcomes for analysis of risk • Which teams are more optimistic / pessimistic about the metric © Crowdcast, Inc. 2010 • How has the forecast changed through time © Crowdcast, Inc. 2010 Complete solution Software © Crowdcast, Inc. 2010 Services • Secure hosing • Market design • Training, deployment • User engagement strategy • Custom report and analytics