Possible Solutions Because of the importance of ROI, of making marketing investments more predictable, and of integrating marketing into a company's financial model, over the past two decades most companies have engaged in econometric modeling to try to solve this problem. All companies agree that such modeling lacks granularity and has numerous validity gaps, leading to little impact on finding ways to increase ROI. Over the same time frame there have been several important projects which, although ultimately not scalable solutions, have pointed the right direction for getting to true marketing ROI measurement and optimization. IRI's BehaviorScan, Arbitron's ScanAmerica, and Project Apollo are all examples of singlesource systems which utilized hard (electronically gathered) media and purchase data gathered from the same households. These systems have shown that singlesource fills the gaps left by modeling, and repeatedly show that it is possible with singlesource to select media (TV programs for example) that directly skew to the types of purchasers who are currently most sales-responsive to a brand's TV campaign. Studies using singlesource have shown that a 40% improvement is possible in the number of impressions delivered to such responsive purchaser types per dollar of media spend. This obviously is capable of making a significant improvement in ROI as contrasted with picking shows that skew to sex/age groups which have some weak overlap with the true purchaser targets. The problem has always been scalability. Meaning, how does the industry get a singlesource system that can charge reasonable prices for the research, and yet deliver the huge sample sizes that are needed so that all brands (and not just the few largest brands) can be served with statistically significant insights? This is TRA's breakthrough. TRA is based on scalable singlesource, using a proprietary privacy solution which makes possible passive media and purchase singlesource data collection from massive near-census sample sizes, while maintaining consumer anonymity. Set top box data and other digital media clickstream data, frequent shopper card data and advertiser's own household level purchase data, audited print circulation by household, advertiser CRM records, and other databases can be leveraged through TRA . Each data stream is managed in the most sophisticated manner with considerable accumulated expertise, rather than simply taking the data that comes out of e.g. set top boxes without regard for the inherent problems in specific data streams