Modern Treasury

Exploratory data analysis at Modern Treasury

Digital transformation has not been an easy path for industries that haven’t historically operated in the age of credit cards or digital transactions – think real estate, utilities, healthcare. Those companies rely on bank rails such as ACH and wire to move more than $750 trillion per year.

Modern Treasury was created to change that. The payment operations fintech provides software and APIs to help companies move money rapidly and track product monetization across the global, digital banking system. They’ve implemented Hex as their analytics workspace, and as a result, data scientists save hours on every project, collaborate with stakeholders across the business, and surface insights that inform key business decisions.

Why Hex?

Allison Schwartz and Chang Sun were hired as Modern Treasury’s inaugural Data Team members in December 2021. A month later, they implemented Hex as their analytics workspace.

Collaborative data science for SQL + Python + no-code notebooks

“The key selling point was its interoperability between SQL and Python,” said Chang.

“We needed a place to do our work, and when I came across Hex, I was impressed by its creative approach to data science workflows,” added Allison. “Hex allows us to flex up or down – we can save exploratory analysis work in SQL, or go deep with machine learning models in Python.”

A powerful complement to BI

“We had Looker already for dashboards and reports, and wanted something to complement that without duplicating functionality,” said Allison. “We sought something that could also be a powerful notebook-style data platform where we could string queries together.”

Hex puts security first

Operating in the financial sector, the security of their data and analytics workspace couldn’t be taken lightly. “Hex being SOC 2 compliant was a big part of why Hex was appealing,” said Chang.

Use cases

Allison and Chang use Hex as their analytics workbench to support exploratory analysis projects that help understand the shape of Modern Treasury’s business in a few key areas:

• Business volume and health: How much money is moving through its APIs? How much is getting reconciled? How much are customers relying on Modern Treasury vs. using old methods to move money?

• Billing and finance: How much of that volume does Modern Treasury bill on? Do their billing tiers make sense?

• Product R&D and Go-to-market: Are new products being used by clients? Should they change the way they’re bringing new products to market? How are marketing campaigns performing?

For example, the team wanted to investigate whether to invest in building a product around recurring payments. “To do this, we needed to understand which customers were processing lots of recurring payments on their own,” explained Allison. “We had to figure out how to define a recurring payment. We created an app in Hex for this analysis, and ultimately decided building the product would not be worth the investment.”

“In another case, we were trying to figure out how many customers were fully activated by our internal definition,” said Chang. “But our internal definition was not easily queryable or clearly defined in Looker. In Hex, we could look at customer activity in 5 of the last 7 days to easily scope out their engagement level. We ended up setting this up as an app in Hex to monitor over time because it turned out to be a really valuable metric.”

Workflows in Hex usually fall into one of two camps:

1. Longer form exploratory analysis that can be shared, saved, and looked back on. “You can get messy in Hex in scratchpad fashion and then build beautiful apps. These apps have been surprisingly useful for stakeholders because of their flexibility, interactivity, and nice presentation style,” said Allison.

2. A shared workspace to collaborate and publish data apps. “In one case, we had to get crisp on what query we were billing on. I worked closely with an engineer to go through this, and even though the engineer had no prior experience working in Hex, he got started with Hex super quickly and immediately saw its value,” said Chang. “Working together in Hex saved us many hours that we would have otherwise spent sending files and screenshots back and forth, etc. And now we have a saved repository – in the form of live Hex data apps – that can be shared with others too.”

Why it matters

On any particular workflow, Hex is saving Modern Treasury’s Data team hours of time that would otherwise be spent working in silos and sending screenshots or other attachments across email, Slack, or other systems for feedback.

“Hex gives us velocity of analysis. It makes it easy and fun to do quick investigative work, which was a pain before. The low floor/high ceiling approach is really true and facilitates fast experimentation,” said Allison.

These hours of efficiency gains can be spent on projects that add value to the business. For instance, the team is exploring paths to deliver insights to customers based on their learnings across accounts.

“Payments are directly part of customers’ critical systems -- we get lots of information about how companies work, and we’re now in a place where we can help customers understand their own businesses. We can assess which companies are performing well and which aren't,” said Allison. “We are starting to scope this out as one possible path to deliver new insights to clients.” Delivering these kinds of aggregate insights could potentially open a new revenue stream for Modern Treasury too.