This Small Business Innovation Research (SBIR) Phase I project addresses the challenge of offering reliable, actionable and real-time intelligence about private companies - a challenge acutely felt by financial institutions, investors, service providers, researchers and entrepreneurs. The company plans to develop a platform that will (1) collect information on private companies from millions of relevant sources; then (2) extract and structure this information and finally (3) algorithmically deliver an ongoing quantitative measure of company creditworthiness and health - thus developing a Private Company Strength & Sentiment Score. The Phase I research is comprised of three stages: (1) Set Up, which includes the identification of Strength and Sentiment Indicators (SSIs) that drive the score as well as identification of relevant data sources; (2) Data & Technology, which includes optimizing existing technology for collection of SSI data as well as collecting real financial and operating private company data for a test sample; and (3) Analysis, which includes preparing the data set, running statistical analyses and creating SSI scores. This final step of Phase I entails examining correlations between outputs and the empirical data collected for private companies so that predictive value can be assessed.
The primary commercial benefit of this effort will be the increased availability of competitively priced credit to US small businesses. If successful, the effort will reduce the information asymmetry surrounding private company information and thus allow financial institutions to better estimate the initial and ongoing risk associated with private small businesses who may be seeking credit. Broader societal impacts resulting from increased credit availability to small businesses include increased "economic growth, employment and payrolls at businesses of all sizes" according to the SBA Office of Advocacy. In addition to benefitting institutional lending, the effort has the potential to offer significant upside to those businesses providing trade credit or supplier financing of purchases as well as to the equity investment community, e.g., venture capital, private equity. Longer-term, the data collected to create and track companies' scores represents a rich repository of entrepreneurial & private company data that can be leveraged to create data-driven offerings that assist entrepreneurs, academics, researchers and public policy professionals their attempt to understand and support our nation's entrepreneurial ecosystem.
Assessing Private Company Health Using Advanced Language Computing Techniques This SBIR project sought to develop an innovative software system directed at providing actionable, real-time intelligence into the health of private companies. To date, the information such firms have had to rely on has been driven by highly imprecise and imperfect heuristics and often bad and outdated data. The software works by analyzing a diverse array of structured, semi-structured and unstructured information sources and identifying events and signals that have predictive value for assessing company health. To enable this evaluation, a variety of machine learning and advanced language computing techniques are used to both collect, transform and ultimately analyze these inputs. The disparate signals collected by the software are then algorithmically evaluated using statistical models to arrive at an overall evaluation of a company's health. The ability of this software to understand companies at a more nuanced level and on a more frequent immediate basis would have profound impacts and utility for firms who regularly work with small businesses. At present firms working with private businesses are hamstrung by a lack of credible, actionable, scalable and real-time information and hence are unable to distinguish with any degree of precision between healthy and unhealthy firms.
