Jonathan Chin, Co-Founder and Head of Growth & Data Strategy at Facteus, discusses how his company evolved from an analytics firm serving banks to becoming one of the largest providers of consumer transaction data in the US.
In this interview, he talks about scaling a data-as-a-service (DaaS) company and why DaaS is still an important term for the industry. Chin also shares insights on modernizing data delivery and what DaaS companies should – and shouldn’t – borrow from software when building products.
A few topics covered in this interview:
- Building a DaaS business
- Comparing DaaS vs. SaaS
- Managing data product life cycles
- Modernizing data delivery
Companies have sold data for decades, if not centuries. How is data-as-a-service different?
The biggest change in the data business is that teams don’t just want a dashboard or report anymore; they want access to the underlying data itself. That’s happened because the technology to analyze data is just much, much cheaper than it’s ever been. Not only is it cheaper to actually run the machines, but you can go open-source a good statistics model that’s backed up by research that costs you nothing. We’re transitioning to a place where we are actually selling data – and that comes with a different set of requirements and challenges.
The shift toward data feeds must have pretty deep implications for the way you deliver your products. Talk a bit about how Facteus is thinking about modernizing its delivery infrastructure?
The single biggest change has been the amount of data we’re moving on a daily basis. What used to be a few bytes in a dashboard is now terabytes transferred via data feeds. The challenge is that the cloud storage and data platforms through which we are moving the data not only don’t play nice with one another, they often do not place nice with themselves. What most people do not understand is that two regions in the same platform are basically two different worlds. Bobsled has effectively solved this problem for us, which is huge.
“Data-as-a-service” draws on an analogy to what happened in software over the past decade. What’s different when you think about building data versus software products?
The big difference is that software is nothing until it’s something. When we’re building a data product, we’re molding clay that exists – maybe a data set we licensed – into the shape we need for our customer. In software, you can technically build anything; in data you have very real limitations. If you don’t have data about Europe, no amount of intelligence or engineering resources will get you to have that information.
How does that translate into actually running a data business?
A big difference from software is that the value proposition for a data product can be extremely dynamic. With a data product, it’s often hard to predict how a customer will end up using your data. Often, the thing folks start using our data for is not the thing they end up using for months or years down the road. That dynamism is a good thing because you can add value for a customer in multiple places; but it’s a bad thing because your sales process is a lot less targeted because you’re always trying to answer: how can I help?
There’s a lot of conversation today about bringing a product mentality to data. What makes a data product successful in your eyes?
One of the important considerations for a “data product” is the use case – something that we’re not always disciplined about in the data world. Once you identify the use case, it makes it a lot easier to know what part of the product you should invest in. For instance, when we deliver products to our banking customers we invest heavily in merchant tagging because they care about what individual customers are doing. But for investors, they care about trends: so in some cases, we actually provide them a statistically representative sample.
How do you think about developing “features” for a data product?
A ton of the product features we build are focused on a simple goal: how do we make it more convenient for our customers to consume our data. That might mean the ability for us to deliver our data in the format of their choice and increasingly that means delivering directly to data warehouses like Snowflake. Our product lifecycle is all about unlocking more features so it’s more convenient for our customers to consume our data.
Facteus did not start as a data company. Talk a bit about how Facteus made the journey from an analytics company to a data-as-a-service business?
Benchmarking was our first foray into building an external data product, so to speak. At this point, we worked with several large banks and our clients started asking what our other clients were doing around non-competitive parts of their business like fraud rules. So we started to build benchmarks across our customers, and in doing this, we had to build a deep knowledge of data security, privacy, and compliance, particularly around transaction data. As stewards of this data, we had to understand things like PCI level one compliance, GLBA, and privacy regulations. Slowly, we started to realize that this was actually a monetization engine and led to the next iteration of the company.
How did the actual transition happen? How did your team come to the decision to build a data business?
For years, we relied on third-parties to monetize the data, but over time we realized that the data industry was growing so fast and we would be stupid not to be a part of it. We also thought we could add a ton more value by expanding our data rights and putting our domain knowledge to work, so we started going back to our partners to negotiate bigger agreements.
When we were looking at the competition, we saw that there were a few companies doing something similar. But we also saw a huge opportunity for a new player like us to enter the market. To replicate what we've done, you would need a deep understanding of the payments industry, strong relationships with trusted partners, and expertise in data security and stewardship. Plus, you'd have to know how to handle and productize the data, which is where we really excel.
Find out why Facteus and others leading data-as-a-services business rely on Bobsled to power their delivery programs. Schedule a demo today.
By clicking download you're confirming that you agree with our Terms and Conditions.