Unreliable data is the biggest challenge facing go-to-market and operations teams. According to SiriusDecisions, account-based marketing (ABM) departments cited ‘lack of account and contact insights’, and ‘data usability issues’ as their top hurdles second only to ‘insufficient budget’.
Most notably, data matching causes significant business challenges when records must be merged together to avoid duplicates and guarantee accuracy and coverage. In data science, a ‘match’ is only as good as the reference set being used to render it and there is a new innovation powered by machine learning and natural language processing technology that is completely revolutionizing the reference set used to update B2B data worldwide.
We sat down with EverString’s Chief Operating Officer, Amit Rai, to learn more about the ground-breaking new innovation in data management that is transforming how businesses leverage data.
Q: What is EverString’s source of data?
A: Most data providers are reselling information from other flawed sources, carrying the faulty information onward and perpetuating the problem. These traditional data resellers use surveys, web forms, call centers and other time-consuming, error-prone human labor techniques that do not scale. The result is a trade-off between coverage and accuracy.
EverString uses advanced, business-specialized machine-learning algorithms, trained to mimic industry professionals, to crawl cached data available from the entire World Wide Web. EverString owns and maintains this database, which is continuously and automatically kept up-to-date.
Q: What if a company does not have a website? How does information get collected?
A: Companies that do not have a website still have some digital presence available, whether it’s through sites like YellowPages.com, Yelp, Google Business Listings, Google Maps, LinkedIn and others. These digital traces are indexed on the World Wide Web.
Q: Does EverString crawl directory pages like Yellow Pages, LinkedIn, and Facebook?
A: No. EverString extracts information about businesses indexed by the World Wide Web. EverString uses proprietary business natural language processing techniques, a process that mimics how a business professional would think if tasked with finding information about a company or a solution. This includes but is not limited to executive-level business logic, reasoning, context, nuance, jargon, and other relevant semantics.
Q: Does EverString include international data?
A: Yes, EverString serves all major English speaking countries, including the U.S., Europe, Canada, and Australia.
Q: How many companies are in EverString’s database? What is the coverage area?
A: With the world wide web as the reference set, EverString is able to provide data on as many companies that can be searched on the internet. Trillions and trillions of digital footprints are available in multiple locations, sites, pages, and properties across the World Wide Web.
Q: What is EverString’s data accuracy rate?
A: Since EverString uses the broadest and most up-to-date reference set, our data provides the following accuracy metrics:
• Nearly 100% match-rate (3x better than traditional vendors)
• 3x better match quality than traditional vendors
• 100% fill-rate (3x better than traditional vendors)
• 70-90% fill accuracy on NAICS or SIC (2x better than traditional vendors)
Q: How does EverString measure data accuracy?
A: Data accuracy is measured in 4 parts: Match-rate, match accuracy, fill-rate, fill accuracy. Periodically and regularly, EverString Data Teams take a random, stratified sample and then vet that sample against known golden datasets.
Q: Does EverString use human labor to clean data?
A: No. EverString creates pristine data sets and uses those to train machines to mimic the logic, thought process, valuation measures and other nuances of top business professionals.
Q: How frequently is EverString’s data refreshed or updated?
A: EverString systems are monitoring the web 24×7, and updating data monthly. Extraction happens in real-time. Behavioral data is refreshed weekly.
Q: How does EverString achieve a 99.8% match-rate?
A: EverString’s smart matching technology produces a 99.8% match-rate because the World Wide Web is used as the reference set, ensuring the most comprehensive list of accounts are matched and further scored.
Machine algorithms trained to think like business leaders, continuously synthesize the information with expert semantics, nuances, logic, reasoning, education, and experience. Other data providers base a match on a static database that requires constant, manual scrubbing and limited machine resources, resulting in up to 70% of potential addressable accounts being excluded.
Q: How does EverString’s matching process differ from others?
A: EverString is the only data source that uses business-oriented natural language processing (NLP) technology to mimic the knowledge, logic, reasoning, and analysis of industry experts.
When assessing the coverage of a data vendor, the traditional approach is to ask that vendor how many companies they can match in a given zip code area. Then, the vendor tries to produce a list of companies it has in the database, that can match the criteria. But this is a flawed method for one critically-important reason: The matching process is referencing a potentially outdated database as the master file.
Instead of zip codes, smart matching synthesizes trillions of digital signatures together, to develop a comprehensive view of an organization including the logic-based calculation of useful information such as a company’s position in the marketplace compared to competitors, the propensity to purchase certain software, and many other attributes.
Consider the difference between fingerprint matching technology and DNA testing. The former method is useful at helping identify someone, but the latter provides an unsurpassed, submicroscopic level of granularity. Just like DNA contains trace bits of code that make each human unique, the internet has trace bits of data that tell a unique story about each business, each industry, and each market.
Explore EverString’s unique matching technology and the science behind this exciting industry innovation. Download the full white paper here >>
PLUS, stay tuned for another post about account-matching and the all-important reference set.