Windfall.
Quietly pricing America.

The company that knows what your customer is worth

It is a Tuesday at 430 Pacific Avenue, in the Jackson Square block where the Gold Rush actually started. Inside Windfall's office, 120 people are doing something the Gold Rush would have recognized. They are figuring out where the money is.

Windfall is a B2B SaaS company built on a single, audacious claim: it maintains a net worth estimate for every household in the United States. That claim is not a marketing flourish. It is the product. Fundraisers at Make-A-Wish use it to decide who to call. Marketers at Inspirato use it to decide who to email. Analysts at the University of Michigan use it to plan capital campaigns. More than 1,500 organizations now pay for the same thing: a less embarrassing way to guess.

The company sits at the intersection of three words that have been worn smooth by overuse - data, AI, workflow - and somehow makes each of them mean something concrete. Data: a refreshed picture of household wealth and career signals. AI: predictive models that turn those signals into a likelihood. Workflow: an integration that drops both into Salesforce, HubSpot, or Outreach before the sales rep finishes their coffee.

Legacy consumer data was old, wrong, and confident about both

Long before Windfall, consumer data existed - in the same way that a 1998 Encarta CD existed. It was technically information. It was also out of date the moment it shipped. Vendors stitched together credit headers, magazine subscriptions, and decades-old census tables, then sold the result as "insight." Marketers paid for it. Fundraisers paid more for it. Everyone quietly suspected the data was bad. Nobody had time to prove it.

Arup Banerjee, Cory Tucker, and Dan Stevens had spent careers building data products at Box, GoodData, and Radius. They knew, with the unkind clarity of people who have shipped the sausage, exactly how the sausage was made. The trio kept noticing the same pattern. A company would buy a wealth-screening file, run a campaign, and act surprised when half the addresses bounced and the other half pointed at retirees in zip codes that gentrified a decade ago.

So they asked an annoying question. What if you rebuilt the entire dataset from scratch, using deterministic public signals - property records, equity filings, philanthropic giving, career trajectory - and refreshed it on a cadence that respected how fast human lives actually change?

Three product people, one stubborn thesis

In 2016, the trio set up shop in San Francisco. Banerjee took the CEO seat. Stevens took the COO seat. Tucker built the engineering core. None of them came from the traditional data-broker world, which turned out to be the point. They were not trying to optimize the old industry. They were trying to leapfrog it.

The thesis was uncomfortably simple: if you can model every American household reliably, you do not need to sell different products to retailers, fundraisers, and travel brands. You sell them the same picture of the same household, then let each vertical do its own thing with it. One dataset. Many doors.

Investors took some convincing. Bullpen Capital led a seed round in 2019, after Windfall posted what the company says was 500% year-over-year growth. En Pointe Capital followed with a Series A in 2021. By the time Morgan Stanley Expansion Capital led the $65M Series B in April 2025, the question was no longer whether the bet worked. The question was how big it could get.

One platform. Four verbs.

Windfall's marketing site lays out the product in four words a fifth-grader can defend: Identify, Understand, Engage, Measure. This is, mercifully, not just framing. It maps to actual modules.

The Windfall Application is the SaaS front door - a place to slice the audience, build segments, and push them out the door. Enrichment appends net worth, career, and household signals onto a customer file. The Windfall API does the same thing for engineers who would prefer not to log into a SaaS app, thank you. Wealth Screening is for fundraisers who need a tidy list of donors with capacity. Predictive AI wraps the whole thing in propensity models that estimate likelihood to buy, donate, or upgrade. New for the AI cycle: audiences generated automatically and pushed to paid media platforms, which is exactly as useful, and exactly as worrying, as it sounds.

The clever move is that everything sits on the same underlying graph. A retailer using Windfall to find high-LTV customers is, in effect, using the same data a nonprofit is using to find major-gift prospects. The vertical changes. The substrate does not.

What the customers actually do with it

Customer logos are the cheapest form of credibility, which is why most companies overpay for them. Windfall's list is interesting because of how disparate it is. Make-A-Wish, a wish-granting nonprofit. Inspirato, a luxury travel subscription. The University of Michigan, a Big Ten research institution. Columbia University Irving Medical Center, a hospital. Facet, a fintech. These organizations have almost nothing in common, except a need to figure out, very quickly, which humans on their lists are capable of writing a meaningful check.

Fundraisers use Windfall to pre-screen prospects so gift officers do not waste a flight. Marketers at hospitality and travel brands use it to suppress audiences that will never convert. Healthcare advancement teams use it to plan capital campaigns. The same dataset, in each case, becomes a different verb. That is the unsexy point.

The integrations matter too. Windfall plugs into Salesforce, HubSpot, Outreach, and the rest of the CRM and marketing stack. The data does not live in a vendor portal you log into once a quarter. It lives next to the rep, the gift officer, the campaign builder. Which, if you have ever tried to make a sales team adopt a new tool, is the only thing that determines whether it gets used.

Things that did not fit in any other section

Democratize the picture, not just the pitch

Windfall states its mission in plain language: change how organizations perceive and use people data. The vision goes further - democratize access, workflows, and insights on people data. Stated aloud, this can sound like a buzzword salad. Stated in practice, it means a regional hospital should be able to run a capital campaign with the same intelligence Bain runs on a Fortune 500 - without hiring Bain.

That ambition has obvious tension. A net worth dataset for every American is, by definition, an enormous lens on private life. Windfall publishes a privacy and security stance, and its data comes from public and modeled sources rather than tracking pixels. But the philosophical question that hovers over the category - how much should anyone know about anyone - is not one the company can answer alone. It has chosen, instead, to compete on usefulness while taking compliance seriously enough to keep its enterprise customers in their seats.

Predictive everything, fewer cold calls

The Series B was raised in the same quarter every B2B SaaS company suddenly remembered it sells AI. Windfall's version is less hand-wavy than most, because it has the raw input. Predictive models need data, and Windfall has spent a decade compiling exactly the kind of data that makes a model useful: stable signals about real households, refreshed on a cadence that respects how lives actually move.

The product roadmap, in broad strokes, points toward fewer manual lists and more automatic ones. Audiences that build themselves. Models that retrain themselves. Engagement workflows that triage themselves. If that future arrives, the sales reps and gift officers using Windfall will spend less time prospecting and more time, well, doing the part of the job that requires being a person. Which was probably the point all along.

Back to Pacific Avenue

It is still Tuesday at 430 Pacific Avenue. The 120 people are still working. The Gold Rush block is still the Gold Rush block. But the thing on the screens has changed. It is not a map of California. It is a map of capacity - a quietly drawn picture of who can give, who can buy, who can be helped, who can help. Windfall did not invent the urge to know that. It just decided to ship it as software.

The check from Morgan Stanley says the market agrees. The 1,500 customers say their teams agree. The skeptical part of any honest reader - the part that wonders what it means to model an entire country - is still allowed to have its doubts. That is the deal with a company like this. You do not have to love the premise. You only have to admit it works.