Also FYI, CEO of Bolt here. Ask me anything!
"Mouse movements, keystrokes, capitalization, clipboard usage, and more make sense, because we understand all the elements of the DOM" ...
Such data is super valuable for fraud detection.
That is - if ability to model mouse movements as well as similar behavior features over time is indeed part of the offering and included into modeling [1]. Pretty much no one to speak of is productizing it today.
[1]
https://www.splunk.com/blog/2017/04/18/deep-learning-with-sp...
Every good customer is precious, and not only has order value, but also has customer lifetime value, referral value, and more. Falsely rejected customers will also go to competitors, handing that value on a silver platter to them.
Our unparalleled data visibility, fraud models, and review team results in a white-glove service that maximizes order approval rates and minimizes false positives. We’ve had countless customers switch from Signifyd and related vendors to Bolt, all of which realized these improvements.
Finally, our checkout drives 10%-50% newfound revenue. In summary, we can offer an all-in-one solution and massive revenue upside for making the switch. Let me know if I can help :-) rb [at] bolt.com
I've we have already approved the order, it is indemnified for fraud and we'll take the loss if it's a stolen good.
All future orders from that CC will be detected and flagged as having previously had a chargeback as well!