In the industry, most criminal record database vendors use the venerable but trustworthy FUD tactic. Standing for Fear, Uncertainty and Doubt, they make you fear they might have data other providers don’t. They make you uncertain of how you can reliably ascertain this, and they try to make you doubt your existing provider.
They also like to play the statistics game. As Mark Twain was fond of saying, “There are three kinds of lies: lies, damned lies and statistics.” With a twinkle in their eye, they ask, “So, how many records are in THEIR data?” Seems like a good question, right? He with the most records wins? Maybe, if all records were created equal. But one common buyer’s remorse complaint in criminal records is, “Why, oh why, are there so many duplicates? Can’t you clean up your data?” LOL. If they cleaned up their data, they wouldn’t have the largest record count.
What’s the next one? “How many jurisdictions?” seems like a good question. But again, what does it mean? One person lists “NC” as a jurisdiction. Another lists every county in NC. And then someone has this county they count as a jurisdiction from which they have 73 records …from 15 years ago. Where did they get them? Maybe from a court runner? How complete is that information?
What’s next? Here’s something we tried to do at Accio Data: You find people who found records at one vendor but not at another. But guess what we learned about the records that the “other vendor” didn’t have? Several interesting things:
- Turns out the other vendor DID have them. It was differences in profiles/how filtering happened.
- Turns out the other vendor DID have them. But the other vendor’s search engine didn’t look at AKAs. In this case, a search was run for the fictional Adam Badboy. Adam was a registered sex offender. In the sex offender registry, he was listed as having a primary name of Andy Badboy. And an alias of Adam Badboy. You sure would think that when you sent the vendor a primary name of Adam, they’d check all of the names. They didn’t. Takeaway: The search engine is at least as important as the data.
- The other vendor didn’t have them. Turns out the record was expunged, but one of the vendors wanted record count.
- The other vendor didn’t have them. But they also had records that the first didn’t. Because no two databases are alike.
So back to the question of how do you compare vendors? Run searches that have hundreds of records from both, put them in spreadsheets, and try to compare them? You’ll spend a LOT of hours accomplishing very little. Believe me. We tried it. It wasn’t fun.
I was talking to a client about the problem, and they said the most prophetic thing: We vet our vendors by SALTING them. Just like we do county criminal searches. If we get back the same answer on our final report from both vendors, they’re equivalent. If we don’t, we investigate. What does the same answer mean? The same answer means you added the same jurisdictions when processing in either database. Or if you added more jurisdictions from one of the databases, the result of the county search was the same (which would likely mean the criminal database had stale or expunged information).
What if you do this and they’re not the same? Then you must figure out why. Could be data deficiencies, different filter settings, different vendor software capabilities, or any number of other things.
So that’s how I recommend you do it. When evaluating a new MJD vendor, salt them. In Accio Data, it’s as easy as listing your potential new vendor as a salting vendor and putting in the percentage of searches that you want to salt with them.
About the Author
Barry Boes, CEO of Accio Data, has more than 30 years of experience providing solutions to consumer reporting agencies. Since childhood, he’s been fascinated by high performance, easy-to-use software and the productivity improvements they bring to their users.
Barry Boes is the Founder and CEO of Accio Data.