Demyst Team

Demyst Team

5 Hurdles Facing New Financial Products

Digital customers have high expectations, abandoning transactions at the first sign of friction. At the same time, financial institutions need to comply with regulatory demands and control costs. These conflicting requirements are challenging to manage in an increasingly competitive landscape.

Todd Anderson, LendIt’s chief product officer, convened a panel of financial services experts to discuss how successful products balance these priorities. His conversation with David Hicks, head of fraud at Koalafi, and Brian Lanehart, president and co-founder of Momnt, explored their efforts to develop innovative products and the lessons they learned along the way.

“People are really going from zero to one very fast, and they're actually able to stand up products with great customer experiences,” Hicks said. New competitors in the financial space, who aren’t slowed down by the legacy technical debt, can quickly tailor experiences to meet specific customer demands. “It creates pressure for everybody.”

Lanehart agreed about the combination of competitive pressure and technological advancement, which have driven innovative ways to offer new products. One example was Momnt’s response to customer concerns about the global pandemic, which led them to develop a contactless application process. 

Both Lanehart and Hicks applied external data to launch their products successfully, and the conversation with Anderson discussed the obstacles that they encountered along the way. By working with a partner who specialized in external data, Lanehart and Hicks were able to overcome major challenges that have prevented new financial products from getting off the ground:

1. Credit Bureau Data Delays 

Although integration with credit bureaus has been described as long and tedious, Lanehart explained that it’s essential — the majority of lenders are still using traditional credit metrics to determine who can access funds. 

Some companies are exclusively using alternative data for credit decisioning, but Lanehart anticipates that it will be difficult for them to continue unless they integrate credit bureau data. “The lending institution is so familiar with traditional data, you have to use it,” Lanehart said. “At least, as a start.” 

Integration delays can compound quickly if lenders want to draw data from multiple bureaus that require separate integration processes. These development cycles can be shortened by working with partners who specialize in external data and who have already integrated their systems with credit bureaus. The partnerships enable better management, faster iteration, and more positive consumer outcomes.

“As a startup, there's no way we could have gotten to market with a beta product in six months without someone like Demyst,” Lanehart said. “It would take forever to go through every single credit bureau.”

2. Integrating Low-Value Data Sources

Hicks was enthusiastic about new products that merge data from multiple sources. “If you look at the spectrum of data out there, and you think about external data and internal data, you certainly want to have both,” he said. “The credit score that we're all used to seeing is designed to satisfy all kinds of different needs and purposes that are similar, but not exactly the same.”

However, It's important to understand which external data products will be useful. Lanehart said, “That was a big challenge: is the data valuable?” He stressed the importance of working with an external data partner who can recommend data sources based on a deep understanding of existing products. 

“If you go straight to a data provider,” Lanehart said, “they’ll do everything they can to sell you,” He added, “There's no way we could become experts on all the available data that's out there.”

Hicks also pointed out the opportunity costs involved in integration. “Tech capacity is a scarce resource,” he said. “I don’t want to do a thousand different integrations to actually figure out who I want to use.” 

3. Competing in Crowded Markets 

The data from established credit bureaus — like Experian, Equifax, and TransUnion — has created an environment where lenders compete for the same customers on the same terms because they have the same credit scores. By looking beyond traditional data, Koalafi and Momnt were able to identify new opportunities; their richer understanding of consumer risk profiles identified customers whose scores were likely to change. 

“Using alternative data — from Demyst, for example — we can see that a 640 is behaving like a 720,” Lanehart said, “even though their score is 640.” These insights allow lenders to adjust credit offers in ways that benefit consumers that are overlooked by traditional credit models. 

To identify these opportunities, lenders need to integrate multiple sources of data. “It's very tricky, when you're out there working with all those different vendors, to make sure that you're going to be able to get an apples to apples read,” Hicks said. “If you use a centralized party, it does create a little bit of scalability naturally.”

4. Ignoring Algorithmic Bias

Apple and Goldman Sachs encountered difficulties with their 2019 credit card program that served as a warning for lenders who develop new credit products: regulators are paying attention to algorithmic bias.

Lanehart explained that Momnt takes these concerns seriously. “We're very specific, and we're very intentional as we feed data into our credit policy: Is there anything in here that would ever be considered biased data? Whether it's from our perspective, or a regulator's perspective, or most importantly, a consumer’s perspective.”

Regulatory monitoring upholds standards and governance across financial products, and Lanehart welcomed it. “We’ve been through SOC 2, we’ve prepared hundreds of pages of documentation as we built the platform, we know what the bank regulators are gonna want.”

Ultimately, new financial products will need to explain how their models use data, and they will need to identify and avoid biased data. “Demyst is a good partner in that they’ll tell us — because they're seeing the data being used in so many different contexts,” Lanehart said. “Like us, they’re also watching the regulatory news, and they’ll say ‘there's a little bit of questionable data here.’”

5. Relying on Fragile Workflows

“Your customer is not forgiving,” Hicks said. “They don't care if Equifax is down, or if any credit bureau is down.” Instead of focusing on technical details, customers focus on the fact that they aren’t getting what they want. “That reflects on you as the actual lender.”

Lenders need to meet these heightened expectations in an environment where data problems will happen. “People go down all the time,” Hicks explained, “even if it's just little blips here and there.” Lenders can prepare for these setbacks with ongoing management of their data sources and an awareness of how those sources are performing.

“If you use a company that is in the middle, it can actually help streamline that entire process,” Hicks said. When lenders have a partner managing the orchestration layer for their financial products, it provides a single point of contact for identifying and correcting service outages. 

Balancing Opportunities and Challenges

During Anderson’s interview, Lanehart and Hicks explained how they successfully avoided these challenges and delivered innovative financial products to new markets. Anderson identified a common theme in their stories, noting that “It's really about the combination of the various data.” 

“Alternative data by itself is not going to get you there,” Anderson said. “In many cases, traditional data by itself is limited.” But by merging alternative data and traditional data, lenders can compensate for shortcomings with greater flexibility that allows them to deliver more value to customers. 

This places an emphasis on finding efficient ways to combine different sets of data. “External data, no matter what, it’s tricky,” Hicks said. He pointed out that various data sources require discovery work, testing, and integration, followed by the transfer, storage, and ongoing management of the data. 

“There’s a lot that’s gonna go into that,” Hicks said.”No matter how you’re going to tackle it, it is going to be a large undertaking.” Finding the right partner for external data can make that undertaking more manageable. 

The full video of Anderson’s discussion is available on the Demyst website. New users can also create accounts on the Demyst platform and explore ways to integrate new external data sources.

Don't settle for half the story

Demyst gives you access to all of the data you need. Evaluate thousands of data attributes from hundreds of possible data connectors all pulled into your own custom-built APIs for instant data deployment.