E-commerce Fraud Facts

I think there's some confusion -- we actually do have a global machine learning model, which is what customers will start with if we haven't received any labels (training examples to learn from). But, your mileage may vary with the global model. We think it's a starting point, but that fraud differs from site to site in subtle ways, and to be truly effective you need to have your own model. This is why training examples are so critical. Once we receive enough training examples, we're able to build a model specific to your site.

That said, I'll be the first to agree with you -- we haven't done a great job setting the right expectations and messaging, especially on the point above (that your mileage may vary with the global model). We've been working hard to improve this, and there are now several reminders in the console, and an introductory tour, that emphasize the importance of a thorough integration with labels to achieve best results. With machine learning especially, bad data in = bad results out. Also, our first UI/UX designer started a month ago, and improving this experience is a top priority.

With regards to making fraud facts more functional and actionable -- we actually do provide reasons we think a user is suspicious in our web console and API (see https://siftscience.com/docs/getting-scores). But, there isn't an "aggregate learnings" page like what's presented in the blog post -- that's on the 3-month roadmap.

Again, sorry for the subpar experience, and I'll follow up with you separately. We have a very technical product -- we're working hard to abstract the complexity, while making sure we set customers up for success. We've also been overwhelmed with customer demand, and have been scaling the team to keep up (7 FT employees two months ago, 15 now). That said, no excuses. Just know that we're not happy to hear about experiences like yours, and are working hard to make it right.

Thanks, Jason

(EDIT) Aaron and I just had a good phone call -- we're going to make things right.

One thing I forgot to mention is that in edge cases what you do is send off an auto email to the registrant with a challenge question or requesting additional information from them. Or you say that the bank has declined the charge.

So for example if the person says they are "Bob Wagner" a business owner living in the US but replies to the email using broken english or doesn't understand the question you know something is wrong. Of course you try and do it in a way that doesn't tip them off that you suspect anything (other than in the case where you tell them that the credit card has been declined (when it hasn't) to see how they respond). Someone who has committed fraud will of course be paranoid. In fact a kid using his mothers credit card will be paranoid and respond in a unique way that says "shit" in some way shape or form.

Edit: Auto email must be formatted and appear to have been personally written by a person just to the recipient. Otherwise many cases won't be responded to at all.

On your site this phrase:

"We used to have a problem with fraudulent orders."

My suggestion is that the following is much clearer:

"We no longer have a problem with fraudulent orders."

Also the other info I would present in a less negative way. People tend to respond better (imho and experience) to a message presented in a positive way rather than negatively.

So for example rather than:

"Fraud can kill your business" I would say:

"Reduce Fraud and Make more money".

Instead of "Stop the cycle of worry" I would say:

"Sleep at night and make more money"

etc.

Obv. there are many twists to this that's just two examples.