Fintech’s Next Innovation Unlock? More Data Portability
In the era of AI, financial data portability will only keep accelerating
KEY TAKEAWAYS
- Historically, increasing financial data portability has helped drive a new wave of generational fintech companies.For example, Plaid made it easier to share banking data, which was a key enabler for companies like Chime, Robinhood, and Brex.
- Agentic AI can now access and process data at near-infinite scale. Data that typically required an army of humans to access and enter manually is now unlocked with AI.
- Agentic AI has the potential to bring data portability to almost any dataset. This could unleash a new wave of infrastructure and application layer companies across categories including accounting, procurement, and risk operations. Companies like Adaptive in construction account, and Anon in agentic data extraction, have already taken advantage of this.
Financial data is more accessible – and portable – than ever before. Consumers can easily share their transaction and balance data to apply for a Buy Now Pay Later loan on sites like Affirm or Klarna. Businesses can easily share their ACH and balance information to fund their new Ramp account. Anyone who has linked one account to another using Plaid, Yodlee, or one of their competitors, knows how relatively seamless this process has become.
In the era of AI, financial data portability will only keep accelerating. Improved OCR and agentic-data extraction products are enabling unstructured data to be parsed and analyzed easily; luckily, there is no shortage of this in financial services.
Before the proliferation of financial data aggregators like Plaid, physical document sharing and data entry were the primary ways people shared financial data. When someone applied for a loan from a new institution, they needed to go to their bank to print their bank statement. This lack of financial data portability caused less competition and ultimately worse customer experiences.
Plaid and other digital banking data aggregators like Finicity significantly enhanced financial data portability, especially for consumers. Data portability was one of a few major inputs leading to a new generation of consumer-focused financial services companies. Consumers were tired of their sub-par experiences with traditional banks and brokerages, so they opened accounts at neobanks like Chime or neo-investment platforms like Robinhood. Others wanted or needed to be underwritten with data outside of just their credit score, so they applied for a loan with Upstart.
Meanwhile, the Consumer Financial Protection Bureau’s recent 1033 ruling is a major policy win for the cause of financial data portability. Most financial institutions by law will need to enable consumers to seamlessly share their financial data with the tools they choose to use. Banks that were previously hesitant or lacked the technical ability to support financial data portability will have to get on board. As a result, consumers will be able to more easily explore financial offerings and will become more accustomed to sharing their data.
More companies are applying the “user-permissioned data access” concept to other financial data sources. Companies like Codat and Rutter have made accounting data available via API. Finch has done the same for employment data, and Method has accomplished this in the field of debt data.
AI now has the ability to pull data from almost any source through agentically parsing unstructured data. These “AI-enabled data integrations” like Anon will be able to unlock any user-permissioned data source, even for sites without APIs.
As financial data portability increases, two categories of opportunities are emerging:
- Existing categories that will face a replacement cycle due to lower data gravity as users can more easily migrate their data to a new solution, reducing incumbent data stickiness.
- New solutions that previously weren’t possible emerge
Let’s unpack the opportunities in each of these categories below:
01. Existing categories that face a replacement cycle due to lower data gravity
Risk operations tools. Risk teams across onboarding and underwriting have more data sources at their disposal than ever before. As these datasets increase, risk operations tools like Federato in Insurance and Taktile, Greenlite, and Accend in Financial Services will likely become more critical for managing the growing complexity of decision logic across these sources.
Loan servicing. Traditionally, lenders and servicers have a point in time snapshot of a borrower’s financial health during underwriting. The primary data points available after this are if the borrower pays their debt on time. With continuously integrated cash flow, debt, and accounting (if B2B) data, risk teams can get a real-time view into account activity to better understand their portfolio risk and more proactively help borrowers with repayment plans and financial health. Method, Plaid, and Basis already serve user-permissioned data to lenders to help increase this visibility.
Accounting services: Companies have traditionally used legacy ERPs like Quickbooks and NetSuite to manage their finances. Through accounting and bank data integrations, companies can easily sync their data with third-party tools in real time. This makes it easier for a company to use an external vendor to manage certain accounting workflows. For example, Adaptive is using Quickbooks and Plaid integrations to help GCs and construction accounting firms more easily manage their AP/AR and bookkeeping, while Basis uses GL and Plaid integrations to help large accounting firms automate bookkeeping workflows for their clients.
Tax services: Similar to accounting, most of the datasets used for filing taxes are from data sources like ERPs and GLs that traditionally had low financial data portability. Consequently, tax firms would spend lots of manual hours ingesting this data. AI can now parse this data, which creates opportunities for new, tech-enabled firms like Haven or platforms like Black Ore that enable CPAs to process more tax clients.
02. New, previously impossible solutions emerge
Continuous accounting close - Today, companies close their books every month partially because of the manual, high-latency work required to pull financial data. Now that it’s becoming more feasible for them to pull this information in real time, companies can have an accurate view of their books in real time, achieving “continuous close.” Companies like Numeric, Stacks, and others have taken advantage of these improved connections.
Synthesis of alternative data for underwriting. Traditional credit scores do not always capture your full financial picture. This causes lenders to reject applicants that they could accept if they had additional data. As mentioned above, some lenders have implemented raw alternative data like cashflow data into their underwriting processes. However, most lenders want an out-of-the-box, standardized solution. Now that new sources like banking data have become available and are increasingly relied on, there’s an opportunity for vendors like Prism Data, Heron Data, and Nova Credit to help lenders more easily make sense of these new datasets.
Real-time vendor aggregation and pricing visibility for procurement - Now that AI can parse invoices, contracts, and different online sites, it could be possible to get a real time view into all vendors and their associated pricing. This may make it easier to aggregate the supply side of a marketplace and help overcome cold-start problems. KeyChain has employed this AI-enabled approach to help retailers identify the best possible manufacturing partners, and Parspec has done this in lighting and electrical construction materials.
Agentic data extraction and integrations - AI agents can already replicate many human tasks, and this will only accelerate as inference improves. Soon, agents will be able to access any dataset a human would at an infinite scale, even if it needs to be human-permissioned. There are likely many use cases of this data, especially in financial services, that would support a new company to be created. Companies like Deck, Extend, Integuru, Anon, and Blue Sierra are using AI to parse data that was hard to scalably access and process.
At Emergence, we’ve been spending a lot of time thinking about how increasing financial data portability will impact fintech and B2B software and the infrastructure needed to support this.
With growing AI breakthroughs and regulatory tailwinds, financial data portability will only accelerate. Historically, an increase in data portability has helped drive a new wave of generational companies, and we’re confident that this will continue. We’re excited to see which companies of today will paint the future the way that Plaid, Chime, and Robinhood did when they started 10-15 years ago.
These rapid developments present some new questions we are working through:
- What other categories will be enabled due to increased data portability?
- How will data permissions work in a world of AI-enhanced data portability?
- How can regulators promote innovation while mitigating harm to consumers?
Are you a founder building in this space? Or are you considering doing so? We’d love to hear from you! Feel free to reach out to kyle@emcap.com
Enjoying this article?
Sign up to gain access to our thought leadership and have future articles delivered directly to your email.