How to identify risky businesses with alternative data

Information about How to identify risky businesses with alternative data

Published on October 17, 2020

Author: ExperianBIS

Source: slideshare.net

Content

1. 1 © Experian Sip and Solve How to identify risky businesses with alternative data Featuring: Stephen Weber Innovation Manager Experian Business Information Services

2. 2 © Experian What makes a business risky? Let’s flip that questions and ask: what are the top reasons businesses fail? Leadership Didn’t have the right team running the business Outcompeted Product/Market Fit Pricing and Cost Issues Access To Capital Ran out of cash Lacked Business Model Poor Product Offering 42% https://www.fundera.com/blog/what-percentage-of-small-businesses-fail 29% There’s no market need for their services or products. 23% 19% 18% 17% 17%

3. 3 © Experian What makes a business risky? Let’s flip that questions and ask: what are the top reasons businesses fail? Leadership Didn’t have the right team running the business Outcompeted Product/Market Fit Pricing and Cost Issues Access To Capital Ran out of cash Lacked Business Model Poor Product Offering 42% 29% There’s no market need for their services or products. 23% 19% 18% 17% 17%

4. 4 © Experian Knowing the type of business helps address a bigger problem Compliance Risk Policy Adjust Policy Knowing a business’s industry

5. 5 © Experian Industry Classifications SIC codes were developed in the 1930’s and haven’t been updated since 1987 Developed in the 1997 with Canada and Mexico. Removes inconsistencies of SIC codes and gives more specificity. NAICS codes are the industry standard to identify businesses today NAICS codes don’t cover every type of business and has trouble keeping up with market changes SIC NAIC 4 digit 6 digit

6. 6 © Experian Keywords A way to go above and beyond industry codes to quickly understand complex entities that models can utilize Burger & Bowl Bowling Venues Entertainment American Cras arcu ipsum, vestibulum eu lectus a, gravida rhoncus elit Cras arcu ipsum, vestibulum eu lectus a, gravida rhoncus elit Cras arcu ipsum, vestibulum eu lectus a, gravida rhoncus elit

7. 7 © Experian Keywords A way to go above and beyond industry codes to quickly understand complex entities that models can utilize Burger & Bowl Sports Bar Beer Wine Spirits Bowling Venues Entertainment American Cras arcu ipsum, vestibulum eu lectus a, gravida rhoncus elit Cras arcu ipsum, vestibulum eu lectus a, gravida rhoncus elit Cras arcu ipsum, vestibulum eu lectus a, gravida rhoncus elit Play Area

8. 8 © Experian 48% 78% 45% 84% Website Data 48% of businesses have a website Employment Data 78% of people and their employers in the US Social Media 84% of businesses use social media Keyword Sources Keyword data sources that move the needle

9. 9 © Experian Website Descriptive Search Engine Data Website Keywords from Content Employment Data B2B industry category of business Social Media Keywords from reviews Social Media Type of business to consumers Employment Data Previous industries of owner 220%Increase in industry classification data using keywords Individual keyword data attributes

10. 10 © Experian Top-Time Wine Cellar Traditional credit bureau industry classification code data SIC Codes: 1. Liquor Stores 2. Cigars 3. Food Stores, Misc. NAICs Codes: 1. Beer, Wine and Liquor Stores 2. Tobacco Manufacturing 3. All Other Specialty Food Stores

11. 11 © Experian Example SEO Data: A family-owned retail store founded in 1957. Our 24,000 square feet of retail space showcases a huge selection of rare wines, hard to find spirits, beers, cigars and much more. Content Keywords: Wine tasting, Pipes, Lighters, Knives, Keg Equipment Phone Numbers: (800) 555-3005 Email Addresses: [email protected] [email protected] Physical Addresses: 120 Sierra St. Costa Mesa, CA 92708 Website Data Social Media Employment Data Business Categories: Beer, Wine & Spirits, Tobacco Shops 5 Star reviews on avg every 4 days Keywords from reviews: stogies, whiskey, imported, tastings, vape Owner Name: Jack Tyler Employment History: Jack Tyler owned another liquor store between 1989- 1999 Age: 46 Gender: Male Employees: 16 Top-Time Wine Cellar Phone, website, hours, age of business, historical performance

12. 12 © Experian Example SEO Data: A family-owned retail store founded in 1957. Our 24,000 square feet of retail space showcases a huge selection of rare wines, hard to find spirits, beers, cigars and much more. Content Keywords: Wine tasting, Pipes, Lighters, Knives, Keg Equipment Phone Numbers: (800) 555-3005 Email Addresses: [email protected] [email protected] Physical Addresses: 120 Sierra St. Costa Mesa, CA 92708 Website Data Social Media Employment Data Business Categories: Beer, Wine & Spirits, Tobacco Shops 5 Star reviews on avg every 4 days Keywords from reviews: stogies, whiskey, imported, tastings, vape Owner Name: Jack Tyler Employment History: Jack Tyler owned another liquor store between 1989-1999 Age: 46 Gender: Male Employees: 16 Top-Time Wine Cellar Phone, website, hours, age of business, historical performance Leaks/Dark Web Mentions: 2 mentions on pastebin 1 mention on raidforums

13. 13 © Experian ModelingModelingRisk IndexRisk IndexKnockoutKnockout Data Evaluation Data Evaluation Keyword Source Strategy Best Practices Its best to start with a data evaluation either using a portfolio or book of business with the new data appended to evaluate hit rate with each data source Some decisioning can be done off just the raw data, but its likely you will get better results by creating some of your own aggregates Use filtering techniques to make sure you get the data that is useful. Common method is to start off with a rules- based approach to this and then move to an ML approach that decides what should be filtered Create a model by training on specific examples and using batch data to create your predictive model. We have seen with keywords that many time ML does a great job in this area.

14. 14 © Experian Moving the needle How does all this data and techniques move the needle for your business Friction-less All the data that is needed for these data points is frictionless – just an email, website, and business address. No need for documents here. Speed All this data can be provided through API for full automation of any process and batch for any evaluation or model training Corroborating Data Any decision is only as good as the data it’s using – so using various data sources that corroborate each other is a crucial part of any decision

15. 15 © Experian Thank you! Have questions or want to learn more about using alternative data to identify risky businesses? • Send us a message in the “Chat” function or via: bit.ly/sip-solve-chat Upcoming Sip and Solves - • Join the list to stay in the loop: bit.ly/sip-and-solve

Related presentations


Other presentations created by ExperianBIS

Special Report
29. 05. 2014
0 views

Special Report