Breaking pattern One startup test · 150 retirement homes · 149 noes · one spare room Rightfoot reported 6× recurring-revenue growth in 2025 Founder · operator · former IBM banking consultant

Person / Fintech / Field Notes

Danielle Pensack Learned to Build Fintech by Calling 150 Retirement Homes

Before Rightfoot built data infrastructure for lenders, its co-founder moved into a retirement community to test an entirely different idea. The experiment failed cleanly - and became the operating system for everything that followed.

The first useful fact about Danielle Pensack’s company is that it began with dinner at five o’clock. In the summer between her two years at Stanford’s business school, Pensack and classmate Deirdre Clute moved into a retirement community in Castro Valley, California. Their classmates dispersed toward more conventional internships. Pensack and Clute shared the one clean bed available, ate each meal at different tables with residents, dealt blackjack, improvised classes and tried to learn whether a peculiar exchange could work.

Their idea paired teenagers who knew their way around technology with older adults who wanted help using it. The students would earn money, which would flow into college-savings accounts. The arrangement would give two generations something each needed. There was only one practical problem before the experiment could begin: two MBA candidates did not qualify to live in senior housing.

Pensack phoned roughly 150 communities around the Bay Area. Most declined. Some explained that admitting people under 60 was not merely unconventional but against the rules. One administrator had an extra room and a willingness to try. A lopsided tally - 149 noes and one yes - became the price of admission to the field.

150Retirement communities called
1Community willing to host the test
$200K+Combined student loans held by the founders

The clean failure

The pilot produced something founders claim to want and often resist: a clear reason to stop. The exchange between students and residents had genuine warmth, but the earnings could not make a meaningful dent in the scale of tuition and debt. Pensack, Clute and technical co-founder Will Schmitt held more than $200,000 in student loans among them. The little marketplace was aiming at a mountain with a teaspoon.

They kept the problem and discarded the product. Rightfoot turned toward debt repayment itself, building APIs that could let fintechs, benefit providers and financial institutions put loan payments inside products people already used. The work was deeply unglamorous. Loan accounts do not route money like checking accounts. Servicers relied on old systems. Benefits companies mailed paper checks and sometimes learned much later that money had gone to the wrong place. Rightfoot worked on the instructions, reconciliation and status updates beneath the visible transaction.

“Focus is always the number one lesson to come away with.”Danielle Pensack

The company’s later changes make more sense when read as one long argument with friction. In 2023, Rightfoot introduced Connect Magic, a product designed to retrieve consumer-permissioned banking and credit information without making a borrower hand over online-banking credentials. For a lender, the result can be current balances, deposits, withdrawals and account activity. For a customer, it can mean one fewer login ritual at precisely the moment when access to credit or a repayment decision is at stake.

Rightfoot’s product path / the mechanism changed, the target stayed
College savings
through teen work
Embedded debt
repayment APIs
Permissioned data
for lenders
Three products, one recurring question: where does financial infrastructure force people to do work the system should do for them?

Training for the hidden layer

Pensack arrived at this work with a useful combination of impatience and institutional memory. After Tulane University, she spent four years in Southeast Asia with IBM. Based in Singapore, she worked on projects in the Philippines and Thailand, including API strategy and digital transformation for banking clients. It taught her how software moves through large financial organizations - and how slowly an old interface yields to a new one.

Her IBM record also carries an early clue about temperament. On a six-month assignment at a commercial bank in the Philippines, conventional emails were failing to reach employees. Pensack organized a flash mob as a communication device. IBM gave her a Manager’s Choice award for original ideas. Elsewhere at the company, she won $30,000 in an internal competition to develop a digital interviewing app and was selected for a Corporate Service Corps project in Ecuador.

Long before enterprise consulting, she was a sprinter. At Newton South High School in Massachusetts, Pensack captained the track team, earned All-American honors and helped set a state-meet record in a shuttle relay. She competed for Tulane as a freshman, running 400 and 600 meters. A sprint is a compact feedback system: the clock is blunt, the distance fixed, the excuses mostly irrelevant. It is tempting to draw too neat a line from the oval to a startup dashboard. Still, her later preference for explicit targets has the familiar cadence of someone who likes a visible finish line.

The founder move worth stealingSet the threshold before the interviews. A target chosen after hearing the answers is a story; a target chosen before them can become a decision rule.

Numbers that are allowed to say no

At Rightfoot, intentionality became a system, not a mood. When the team explored selling student-loan repayment as an employee benefit, it decided the model would proceed only if seven of ten prospective customers said yes. The result fell just short. A few maybes could have been promoted into assent. The founders declined the invitation to grade their own homework and moved on.

The same instinct appeared in fundraising. For Rightfoot’s $5 million seed round in 2021, the team evaluated investors partly on the composition of their partnerships and limited partners. The term sheet included a diversity rider that reserved 7 percent of the allocation outside the lead investor for underrepresented check-writers. Pensack later described the company’s cap table as majority female. These were constraints with consequences: the founders broke ties with investors who failed their criteria.

Metrics do not rescue judgment; they expose it. A founder still chooses what to count, where to draw the line and when changed conditions deserve a new test. But a number selected before the outcome has one important virtue. It can disagree with the person who selected it.

Two disclosed operating signals
Seed rider
7%
2025 revenue
The first bar is the share of non-lead seed allocation reserved by Rightfoot’s diversity rider. The second marks Pensack’s reported multiple of recurring-revenue growth in 2025; the scales are intentionally separate.

The regret was width

Pensack’s clearest criticism of her own early leadership is not that Rightfoot changed too often. It is that the company tried to serve too many groups at once. Its first customers stretched across lenders, retirement businesses, benefit firms, fintechs and financial institutions. The APIs could look similar while the sales motion, language and demands differed. A small team accumulated organizational surface area faster than it accumulated focus.

Remote work magnified the cost. Pensack has said she would have installed stronger documentation and communication infrastructure earlier. There is an appealing lack of theater in that answer. Startup retrospectives prefer the cinematic error - the missed market, the disastrous launch, the villainous competitor. Sometimes the expensive mistake is five customer types and a document nobody wrote.

Rightfoot’s public update about 2025 sounds like an answer to that regret. Pensack said recurring revenue increased sixfold as the company shipped infrastructure for lenders, introduced self-serve onboarding through a PCI-compliant portal and clarified workflows around when customers should act. The claim is the company’s own, but its shape matters: fewer engineering bottlenecks, shorter time to value, a product embedded in core routines. Focus, translated from slogan into workflow.

A faster loop

In March 2026, Pensack turned her attention to the changing mechanics of software creation. She argued that the important shift was larger than artificial intelligence writing code. Agents could implement a change, run a system, inspect its logs, test behavior and iterate. The feedback loop itself was collapsing. As execution accelerated, the constraint moved upstream toward specifications and system design.

It is a fitting observation from someone whose company began as a feedback loop with an address. Move into the place where the hypothesis lives. Watch what people do. Let the test answer. Adjust the system. Rightfoot’s path from college savings to debt payments to lender data can look like three businesses when viewed as a product catalog. Viewed as an operating history, it is one habit repeated at increasing scale.

There is a practical tension inside that habit. Faster loops can reward restlessness as easily as learning. A team can mistake motion for information and produce a new direction every week. Pensack’s story supplies the missing constraint: the purpose moved more slowly than the product. College access became debt relief; debt relief became better data for decisions. Each mechanism changed, yet each remained attached to a view that financial systems should create a fairer, less burdensome experience. The stable purpose gave the experiments somewhere to return.

Pensack has said people and travel keep her outside the Silicon Valley bubble. She has described homestays, a farm in Laos, journeys across Southeast Asia and a budget trip around Maui in a rented van. The details are light, but the pattern is familiar: proximity over abstraction, participation over observation from a safe distance.

The retirement-home experiment did not discover Rightfoot’s final product. It discovered something more durable for its CEO: a way to learn. One open room was enough to test the idea. One clean failure was enough to change it. Seven years later, with lender infrastructure, new automation and a much faster software loop, the original discipline remains visible. Get close. Name the threshold. Keep the purpose. Let the product lose.