Innovaitors - Sift: Digital Trust and Safety with Machine Learning

Innovaitors - Sift: Digital Trust and Safety with Machine Learning

Sift uses machine learning to help businesses build out their trust and safety efforts by preventing fraud before it happens and providing frictionless and delightful experiences to trustworthy customers.

Listen to learn how Sift protects clients like Airbnb, Yelp, and Twitter from bad actors, fake accounts, spammy content, and more using AI, machine learning, and a customer-first approach.

What is Sift?

In 2011 when Sift was founded, they were essentially a solution in search of a problem, says CEO and co-founder Jason Tan. This was still the early days of machine learning. They knew that ML was going to disrupt many different industries, but they wanted to find a problem domain where the value proposition would really stick.

One of the exercises they did was ask their friends in the tech community about the challenges their businesses were dealing with, and the issue of fraud emerged as a big problem with which people wanted help.

The fact that they were outsiders in the fraud space helped them approach this domain with first principles and a more modern perspective on how to solve this problem effectively. That is how Sift was born.

Trust and Safety on the Internet

Legacy approaches to security on the internet operate much like airport security, not really thinking about user experience, according to Jason. What it really needs is more of a TSA pre-check type model, which allows a frictionless experience to trustworthy customers who are not risky.

“For us, we see the opportunity as much bigger than fraud,” he says. “We use the words trust and safety very intentionally...it’s really that conjunction ‘and.’ It’s not trust or safety, it’s trust and safety.”

When you’re optimizing for both of these variables, it’s really about looking at the big picture. Sift is concerned with removing both friction and risk from every interaction with the end user.

Potential and Adoption

If you look at Sift’s client list, it’s interesting to see the adoption of their technology platform across a diverse set of customers. One of the big opportunities for their services is in the ecommerce world.

Currently, e-commerce accounts for less than 15% of retail sales in the U.S., and it’s continuing to grow rapidly. There is definitely the potential for that equation to be flipped on its head, and the bad actors in the fraud industry will follow.

But there are many other applications for this technology. Sift also solves for account takeovers, fake accounts, and spammy content.

Facebook has recently reported that over the last 10 years they took down over 216 million fake accounts. YouTube had to hire a few thousand people to moderate content. Smaller companies likely won’t have the resources to spend on these problems, and that’s the big opportunity they see.


Key takeaways:

  • Digital fraud is a massive and growing category with increasingly sophisticated actors;
  • Sift is a category leading provider of digital trust and safety;
  • Sift relies on both ingesting large amounts of customer data and applying machine learning to these data sets to improve its fraud models and ultimately protect its customers.

Sift uses machine learning to help businesses build out their trust and safety efforts by preventing fraud before it happens and providing frictionless and delightful experiences to trustworthy customers. Listen to learn how Sift protects clients like Airbnb, Yelp, and Twitter from bad actors, fake accounts, spammy content, and more using AI, machine learning, and a customer-first approach. What ...

Show

Episodes