How Data is Gathered at Prospect Visual?

by Gerry

Have you ever wondered how Prospect Visual gathers information for its proprietary database? In today’s topic, we pry open the Chamber of Prospect Secrets at Prospect Visual.

Just in case you haven’t read my other articles about the history of Prospect Visual’s database, here’s a SportsCenter’s Top Ten Prospect Visual Plays of the Decade Breakdown:

1. Our Database was created and filled using a Natural Language Processor (NLP) that reads various articles on the internet to aggregate data.

2. We never add client proprietary data to our database.

3. Our database is enhanced and updated with additional clients. (We do external research into each contact given to us by the client.)

4. We use SEC filings, Businessweek, and Businesswire as primary sources for our information.

5. We’re working toward increasing the database to 165 million individuals.

6. We update our database quarterly to reflect more current positions held by our saved contacts.

7. We are mostly interested in dates of positions and position levels at companies.

8. We’re not interested in adding information from LinkedIn to our master database. (To clarify, we can add LinkedIn information to each account, just not to our master database. I will explain later why adding it to the master database might be a bad idea.)

9. Top secret algorithms are used to determine connection strength between contacts.

10. We’re generally not interested in addresses, phone numbers, and capacity. (We’re in the business of relationship mapping, not profiles.)

Regarding the LinkedIn information, we don’t update our master database based on user information because user information can be deliberately altered, while SEC information is written in a formal document. For example, here is the link to Alcoa’s DEF14A : http://www.sec.gov/Archives/edgar/data/4281/000119312512109136/d301718ddef14a.htm

If you scan the document, you can see that everything is described (in perfect grammar, might I add) in a factual form that the NLP can read.

On the other hand, we have social media websites such as LinkedIn, Facebook, Twitter, etc. While LinkedIn is supposed to be a professional website, it does contain misleading and sometimes outright ridiculous information on profiles. Here, I show an endorsement page on LinkedIn that is slightly misleading.

One of the most important aspects of the Prospect Visual database is maintaining high data integrity. One mistake on one profile propagates throughout our database and creates connections that do not exist.

The above sums up our strategy at Prospect Visual in terms of data collection, showing how we prioritize our methods. We hope you trust our infinite wisdom in the cryptic art of data collection.