What this covers

Patrick O'Shaughnessy talks with Eric Glyman, co-founder and CEO of Ramp, about how a corporate card company founded in 2019 reached an $8 billion valuation. Glyman begins with industry history. For roughly 30 years only dominant, FDIC-insured banks could issue cards, so competition became a race to give away points while quietly devaluing them. Ramp began instead by asking what costs businesses money and time, and the card was a byproduct. He describes talking with about 100 finance teams before shipping, asking for advice rather than pitching. He also covers the receipt problem, where Ramp texts cardholders at purchase and matches receipts by OCR. He invokes Henry Ford's faster-horse line, and describes a Waze-style benchmark of what top finance teams do.

The second half turns to running a company at extreme speed. Glyman says his share of headcount fell from 33 percent to about 0.3 percent. He argues that people who improve 5 to 10 percent a month stop scaling mathematically, and that Ramp's worst problems came from knowing about a problem and avoiding it. His model of leadership is coaching, citing John Wooden and Bill Walsh.

Several of his positions are contested. He says founders who chase the highest price each round make a local optimization error, and that Ramp sold only 15 to 20 percent per round. He treats secondary liquidity for employees as a management obligation, criticizing Airbnb-style restrictions. He says marketing should create desire rather than list features. Finally, he explains why Ramp holds part of its balance sheet in stablecoin securities, which he says yield far more than bank deposits.

Sharpest takeaway

Glyman argues Ramp won by aligning with business customers' desire to spend less money and time rather than competing on card rewards, and that its interchange economics can fund software that builds trust and compounding growth.

  • Incumbent card issuers optimize for spend and points, while business owners want control, profitability, and time back
  • Ramp built trust by co-building with finance teams before launch and monetizes through interchange as customers grow reliant on the software
  • Separating slow, regulated functions from fast product teams let Ramp grow far faster than incumbents

The claims · ranked59 claims · weighted by value

This asset isn't compiled yet

You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.

0.80

Products should solve the underlying job to be done rather than the feature customers ask for, since customers often gravitate to familiar solutions, so Ramp applied the Henry Ford faster-horse principle of building a car rather than a faster horse.

normativehigh valueestablishednovelty 2/4durability 4/4· Eric

“we went back to this Henry Ford ism of you want to build a car, not a faster horse. Trying to really understand what's the problem, what's the job to be done”

0.76

Business strategy requires choosing a few areas of excellence, since a plan to be excellent at everything is fundamentally crazy, and standardizing the software stack for finance, HR, and operations is a useful trend.

normativehigh valueestablishednovelty 2/4durability 4/4· Eric

“you cannot be good at that many things at once. Even the greatest company in the world can pick a couple of things that they are extraordinary at”

0.73

Rewards and rebates became the biggest cost of doing business for card issuers, and because only dominant, FDIC-insured banks could enter the industry for about 30 years, competition shifted into a race to give away points while quietly devaluing them.

causalhigh valuecontestednovelty 3/4durability 3/4· Eric

“rewards and rebates, which for most of the industry has become this number one big thing, of the biggest cost of doing business was actually rebating it back to the customer.”

0.73

Founders overfocus on product features and ROI, but organizations are not fully rational, so marketing should create desire by identifying what people want and what sticks in their heads rather than informing them about features.

normativehigh valuecontestednovelty 3/4durability 3/4· Eric

“think less about marketing the product and the futures, and getting more into how do you create desire? What are people looking for?”

0.73

Founders who optimize for the highest price each round make a local optimization error; Ramp instead aimed to maximize long-term total enterprise value by assembling a collective of investors who trust them and by selling less equity, around 15 to 20 percent per round.

normativehigh valuecontestednovelty 3/4durability 3/4· Eric

“I think that people optimize locally of the highest price and least solution surely must be the thing. Argues that is wrong optimization. It's about maximizing long-term total enterprise value.”

0.73

Letting employees sell some stock as they become vested helps keep them focused long term, and rules that bar secondary sales for current employees, like Airbnb's, create a negative incentive to leave the company just to get liquidity.

normativehigh valuecontestednovelty 3/4durability 3/4· Eric

“if you're at the company, you can't sell secondary and it created this negative incentive for people to say, if you needed liquidity, leave”

0.68

Providing employees with secondary liquidity is a management obligation, typically becoming relevant around years three to six, because talent is a key input and allowing people to take some value off the table lets them focus on long-term work.

normativehigh valuecontestednovelty 2/4durability 3/4· Eric

“I actually do think it's an obligation of management teams to get to. I think it tends to happen in probably the third to sixth year”

0.64

A bank may pay only 1 to 2 basis points on corporate cash, whereas asset-backed, one-for-one dollar-collateralized stablecoin securities can yield high single digits, ranging from 5 to 9 percent monthly, letting a small allocation generate most of the yield on the balance sheet.

factualhigh valuecontestednovelty 3/4durability 1/4· Eric

“your alternatives are earn one to two basis points your bank, which is fine. ... the yield on it can be high single digit percentage points varying month to month, but can go anywhere from 5%-9%”

0.61

Avoiding an unnecessary dollar of spend is worth 50 to 100 times more to a business or consumer than earning 1 to 2 percent back in rewards, so the real value lies in preventing waste rather than rewarding it.

causalhigh valuespeaker onlynovelty 4/4durability 3/4· Eric

“1% back, 2% back, whatever it could be, sounds great, but not spending that dollar in that first place for the end consumer is 50 to 100 times as powerful.”

0.60

As a company grows, a founder's direct impact shrinks: Eric went from being 33 percent of headcount to about 0.3 percent, so the founder's role shifts to clarifying vision and goals.

causalhigh valuespeaker onlynovelty 3/4durability 4/4· Eric

“there was a day when I was quite literally 33% of the headcount. Today. I'm 0.3%.”

0.57

Ramp's premise is to use interchange not just to pay out rewards but as a pool of funds to build great software that makes customers want to use the card more than others in their wallet, creating a value flywheel that increases volume and improves economics.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Eric

“could you actually use interchange, not just as a means of divvying up rewards, but as a fundamental pool to go and create great software that creates a flywheel of value”

0.57

Before shipping any card or product, Ramp talked with about 100 finance teams and founders by asking for advice on helping them spend less money and time instead of pitching a product, which shifted the relationship from opposing sides of the table to collaborative problem solving.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Eric

“before we shipped a single card or a single product, we talked with a hundred finance teams and founders. And rather than going and saying, Would you buy our product”

0.57

A card differs from traditional SaaS because it is invisible and does not require a buying decision; monetization grows with customers' trust in the business as they use the product more, rather than through seat-based contracts.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Eric

“It's not, here's a contract which you buy, and if you have seats, you go with it. It's invisible. It almost pays you. Our monetization effectively grows with people's growing trust in the business.”

0.57

Card issuers never collected receipts, so employees forgot them and receipts arrived a month late, taking up to 24 hours to transfer from Amex to Concur. Ramp instead texts cardholders at the moment of purchase and uses OCR to match receipts, cutting average receipt turnaround to about 30 minutes.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Eric

“Let's text a card holder, when they have a receipt in their hand. Let's build some OCR to be able to match it to the right transaction. And from then to today, the average receipt is turning within 30 minutes on Ramp.”

0.57

Ramp deliberately designed its organization with a methodical, deliberate function for regulatory, legal, and financial-services risks that must never fail, and a fast product and engineering organization for software that does not directly move money.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Eric

“how could we design parts of our own organization that dealt with regulatory, potential legal, or even financial services, that could not, and could never go down, to be extremely methodical and deliberate?”

0.57

At very high growth, people who are proficient at getting 5 to 10 percent better each month do not scale mathematically, so teams must be reassigned or people with proven scaling experience must be hired above them.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Eric

“At that scale, people who were really proficient in great getting 10% better, 5% better every month, mathematically, we're not scaling.”

0.57

Aggregating data from many companies to show what the highest performing finance teams do, the way Waze uses other drivers' positions, lets a payments platform make thousands of companies more efficient than one-to-one card relationships could.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Eric

“Just as Waze, in many ways, revolutionized the navigation and traffic. It wasn't about having a map and knowing where things were, the most interesting data was other cars in the network”

0.57

Small and midsize businesses increasingly use card and flex payment products with 30 to 60 day terms to match cash inflows from e-commerce sales against inventory timing, shifting from income statement thinking to lifetime cash flow matching, which Eric links to Amazon's cash-flow efficiency philosophy.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Eric

“more efficiently map eCommerce inflows and sales versus your inventory timeline of getting things shipped.”

0.56

Most of Ramp's serious problems came from avoidance, where people worked around each other or did not confront a fundamental breakdown, and the company's job is to ask whether to take the problem forward and face it directly.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Eric

“the failure modes are when people work around each other and don't actually address something here is broken.”

0.56

A leader's main job is coaching people to live up to their potential, using the model of John Wooden and Bill Walsh, rather than solving functional problems themselves, since failure happens when the leader solves the function instead of playing to a person's strengths.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Eric

“my favorite business thinkers ever, were coaches. So whether it's like John Wooden, UCLA, Bill Walsh, The Score Takes Care of Itself, I think a lot of my job is not to do.”

0.56

Ramp's most serious troubles came from knowing about problems, such as people not scaling or broken team dynamics, and failing to act quickly, because at extreme growth rates, marginal 5 to 10 percent monthly improvements no longer keep up.

causalhigh valuespeaker onlynovelty 2/4durability 4/4· Eric

“Generally, the moments when we found ourselves in the most trouble is when we knew there were problems along the way, or there was something fundamentally broken, we were getting signals, we didn't act on it quickly enough.”

0.54

Ramp started by asking what costs businesses money they should not be spending and what costs them time, then built processes or software to fix those, with the card emerging as a byproduct rather than the goal.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Eric

“The question was not, again, how do we design something that'll get people to use a card that was one of the byproducts, it started with, what's costing business money they shouldn't be spending?”

0.54

Patrick cites his friend Jeremiah's principle that for any company, the gap between what it makes and what it sells is often the story of its margins.

definitionhigh valuespeaker onlynovelty 3/4durability 3/4· Patrick

“for any company, you should understand what they make versus what they sell. And the gap between those two things is often the story of their margins.”

0.53

Ramp's data shows travel and entertainment spend has suddenly returned and grown rapidly, but it is now distributed from many cities to secondary destinations, with spoke-to-spoke travel and frequent hybrid offsites replacing hub-and-spoke patterns.

factualhigh valuespeaker onlynovelty 3/4durability 2/4· Eric

“a sudden return and rapid growth of travel and entertainment spend”

0.53

Ramp delayed its founding and put debt on the company so early employees could buy founder shares, and it considered tax optimization for employees as part of attracting people who take big risks.

factualhigh valuespeaker onlynovelty 3/4durability 2/4· Eric

“even actually delayed the founding of the company, put debt on it originally, so people, the first time effectively, could buy founder shares.”

0.53

Ramp never made a fundraising pitch deck; instead it met investors to ask for advice and show progress, building a pool of potential collaborators who could help with customers, hiring, and business problems.

normativehigh valuespeaker onlynovelty 3/4durability 2/4· Eric

“But we've never actually made a fundraising pitch deck. Whether it was a seed, to even the most to recent rounds, we've never gone out to race.”

0.53

Taking a lower price in a round benefits employees through excess value perceived on the table, gives investors a hot deal that appears under-priced, and keeps future raise options open, creating a sense of inevitability modeled after SpaceX.

causalhigh valuespeaker onlynovelty 3/4durability 2/4· Eric

“it starts to create, I think what it adds up to is the sense of inevitability.”

0.53

Ramp's idealized end state is software so smart that a dollar spent through it buys a dollar and five cents of value, with time-consuming tasks automated and crowd-level data from many companies used to make every customer's spending more efficient.

forecasthigh valuespeaker onlynovelty 3/4durability 2/4· Eric

“if you were spending through the software, a dollar elsewhere counted for a dollar five?”

0.52

Ramp treated talent density as an explicit goal, reasoning that people are drawn to organizations like McKinsey and Goldman because other good people are already there.

causalhigh valuespeaker onlynovelty 2/4durability 3/4· Eric

“Why do so many people go to McKinsey? Why do so many people go to Goldman? Because other good people are there.”

0.52

Parsing messy card merchant descriptors and clustering them by merchant revealed overlapping SaaS subscriptions a company was paying for, and this manual pattern-finding became the basis for automated software features.

causalhigh valuespeaker onlynovelty 2/4durability 3/4· Eric

“If you actually understood what the merchants were, you clustered them, it turned out for them, they were paying for, I think in order of five sets of software that did the same thing in project management.”

0.52

Incumbent banks and card issuers like Amex and Chase are slow because compliance functions built for real regulatory purposes have expanded into bottlenecks, with some changes such as a webpage edit or button color taking three months, and the card form factor has barely changed in 30 years.

causalhigh valuespeaker onlynovelty 2/4durability 3/4· Eric

“Compliance functions that were set up for real and important, I think, regulatory review purposes were extended to become, effectively, a bat that made development take three months, for even changing a webpage, or the color of a button”

0.52

Ramp's best ads were usually the tenth or fifteenth iteration, and the team tested drafts by showing variants to its large base of existing customers and watching their facial reactions, since first impressions are critical.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Eric Glyman

“your best ads are never the first, it's usually the 10th or 15th, and you need to cycle to get real honest reactions.”

0.52

Large credit card issuers focused on getting people to spend more, earn points, and show off the brand, while business owners wanted control, profitability, and to get home earlier, creating a fundamental misalignment that Ramp was built to exploit.

causalhigh valuespeaker onlynovelty 2/4durability 3/4· Eric

“So it felt there was a large and fundamental misalignment, where the large scale credit card issuers were thinking a lot about how to get people to spend more money, earn more points, show off the brand”

0.50

The best marketing lesson is to show ideas to real people before launch, using three versions and reading their reactions, because the best ad is usually the 10th or 15th iteration rather than the first.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Eric

“your best ads are never the first, it's usually the 10th or 15th, and you need to cycle to get real honest reactions.”

0.49

A year-and-a-half pilot with a large public company failed because of an endless list of requests, leading to mutual trust loss and team exhaustion, which Eric argues stemmed from people avoiding the fundamental breakdown rather than addressing it.

causalhigh valuespeaker onlynovelty 2/4durability 2/4· Eric

“we were in a pilot with a large-scale public company for a year-and-a-half. And there was a never-ending list of asks”

0.49

Because Ramp is close to the movement of funds, it can learn what best-in-class looks like and recommend more efficient processes, and it moved beyond a 30-day fixed card product to over a billion dollars a year in bill payments within months of launch.

forecasthigh valuespeaker onlynovelty 2/4durability 2/4· Eric

“if you are deeply embedded close to the metal in the movement of funds, you have insights about what best in class looks like, could you make recommendations”

0.49

Ramp asked early customer Candid to share 90 days of transactions, then cleaned messy card descriptors and found redundant project management and design software, including five overlapping tools and grandfathered legacy plans, yielding about a quarter million dollars in savings and a card switch.

factualhigh valuespeaker onlynovelty 2/4durability 2/4· Eric

“We found a quarter million dollars in savings. Let's talk through it.”

0.49

Loans secured against private stock are a growing alternative, and Eric argues management teams have an obligation to address employee liquidity, which he thinks typically becomes necessary in years three to six.

forecasthigh valuespeaker onlynovelty 2/4durability 2/4· Eric

“I also think too, if you're sort of asking what are people's desires ... it's an obligation of management teams to get to. I think it tends to happen in probably the third to sixth year”

0.43

Interchange on a card transaction is paid by the merchant and flows mostly to the issuer, who keeps roughly 2 to 2.6 percent on credit cards, while corporate or premium cards can carry fees up to about 3 percent and debit about 30 cents.

factualestablishednovelty 1/4durability 3/4· Eric

“Depending on the type of card used, that can vary anywhere from debit, call it 30 cents, to corporate or high end premium consumer cards could be fees of up to call it 3%.”

0.40

Businesses cannot be excellent at many things at once; even the greatest companies pick a few areas of excellence, so a plan to be excellent at everything is fundamentally flawed.

normativeestablishednovelty 1/4durability 4/4· Eric

“you cannot be good at that many things at once. Even the greatest company in the world can pick a couple of things that they are extraordinary at”

0.40

Common employee liquidity designs include tender offers lasting at least three months and broadly marketed, caps such as selling no more than X percent of vested stock or Y percent per year, and regular processes such as Anduril's roughly semiannual window.

factualcontestednovelty 2/4durability 2/4· Eric

“can't sell more than X percentage of your stock of long term vested, or no more than Y percent per year. Trying to do that and then run regular processes.”

0.39

Leaders should clarify the vision and quantitative goals down to fine detail so people understand the goal and can propose their own routes, which creates alignment across large organizations.

normativeestablishednovelty 1/4durability 3/4· Eric

“making clear, what is that we're trying to solve for? Can you go down to the nth degree of more concreteness?”

0.35

Issuers keep most of interchange because they take on the risk of the transaction, including clearing, payer default, and underwriting, and after financing costs, credit losses, fraud losses, and rewards, the remaining margin is thin.

causalestablishednovelty 1/4durability 3/4· Eric

“If there's credit involved, which is really the dominant product, most cards high end like that, they're taking the underwriting risk involved, and for that, they're able to keep the majority.”

0.34

First-time founders often operate with a scarcity mindset of not letting a dollar go, and the mature shift is an abundance mindset in which leaders focus on clarifying goals so teams can propose paths.

normativespeaker onlynovelty 2/4durability 3/4· Eric

“people's first company it's, Don't go out of business. Be terrified to let a dollar go. It's often a scarcity mindset.”

0.33

The pandemic turned the savings premise from interesting into fundamentally important, and remote work made Ramp's digital product first in importance.

causalestablishednovelty 1/4durability 1/4· Eric

“And then the pandemic was a huge accelerator for us. That just went from, the premise of savings went from interesting to fundamentally important.”

0.31

Eric's consumer background made product sales simple, but enterprise relationships are complex because they involve buying trust, which requires a different approach than selling a product.

normativespeaker onlynovelty 2/4durability 2/4· Eric

“I come from a consumer background, where it's generally simple. We want to have a product, let's do this. I think in some enterprise relationships, it's really complex. It's buying trust, it's whatever.”

0.31

Ramp's first product was not a simple corporate card but a set of tools that streamlined closing the books, including collecting receipts for transactions over the 75 dollar IRS threshold, syncing them to accounting software, and surfacing ways to spend less.

factualspeaker onlynovelty 2/4durability 2/4· Eric

“If you look through it, it actually was a set of products, streamlined all around helping people close their books faster, going from card, which would trigger IRS requirements if you spent over $75 to collect receipts”

0.31

Giving early team members significant responsibility while accepting that some will leave to start companies is a deliberate approach, and Ramp has supported its roughly 22 early people who later founded companies.

normativespeaker onlynovelty 2/4durability 2/4· Eric

“when I look at some of the first 22 have gone on to do that, and we've been there to support them.”

0.31

Holding stablecoins as an actual market participant gives Ramp credibility with fast-growing crypto companies, because it understands their problems firsthand, even though Ramp does not yet offer crypto-related products.

normativespeaker onlynovelty 2/4durability 2/4· Eric

“what better way to have credibility in it than to actually be a real market participant?”

0.31

Sales operations and revenue operations are very bespoke and somewhat chaotic in most companies today, while CRMs like Salesforce handle contacts but not calendars, productivity, or per-person calculations, making this an area ripe for a product that companies would buy.

factualspeaker onlynovelty 2/4durability 2/4· Eric

“sales ops, revenue operations. It's very bespoke I think in a lot of companies right now and a little bit chaotic.”

0.30

Loans secured against employee stock are a growing trend, moving from banks like Silicon Valley Bank and First Republic to specialized companies built to provide them.

factualestablishednovelty 1/4durability 2/4· Eric

“the other trend as well has been loans that people can take against it, sort of going from only Silicon Valley Bank, First Republic to even companies being built specifically to provide this.”

0.27

Ramp holds a single-digit percentage of its balance sheet in stablecoins, amounting to eight figures in funds, and began this only after accumulating hundreds of millions on the balance sheet.

factualspeaker onlynovelty 2/4durability 1/4· Eric

“single digit percentage, that can say more, but it is substantial. It's meaningful eight figures funds that are in it.”

0.25

About 6 percent of Ramp's companies use all its integrations, including Gmail, Amazon Business, and Lyft, and for them 90 percent of receipts are fully automated with no human touch.

factualspeaker onlynovelty 2/4durability 1/4· Eric

“6% of our companies that enable all of our integrations with Gmail, Amazon for business, Lyft and 90% of all their receipts are fully automated, no human touch at all.”

0.24

The kindest thing someone can do is take a chance on a founder early by trusting them and giving them a shot, because founders face constant doubt and people get only one shot.

normativespeaker onlynovelty 1/4durability 2/4· Eric

“it's believing in me and checking in. ... people only have one shot”

0.22

Ramp's revenue grew 65 times year over year in 2020 and about seven and a half times in 2021.

factualspeaker onlynovelty 2/4durability 1/4· Eric

“stumbling for the real revenue 65 times over year-over-year in 2020, seven-and-a-half and 2021.”

0.21

Ramp's first 10 employees included seven engineers, one design team member, and one talent person, chosen for their ability to build things of structure and find outlier talent.

factualspeaker onlynovelty 1/4durability 2/4· Eric

“The first 10 people, seven were engineers. One was on design team from IDO. One was on talent.”

0.21

Early adoption spread through word of mouth as finance teams and founders told peers in their networks that Ramp was different, aligned, and saving money, even though the category was one people wanted to stop hearing about.

factualspeaker onlynovelty 1/4durability 2/4· Eric

“when you make a decision, they would call up other finance teams, founders, people in their network, and for the first time people were hearing actually this product is different.”

0.18

Ramp reached nine figures in revenue within less than two years of launching, and Eric sees data products built on its embedded position in fund flows as a major next opportunity.

factualspeaker onlynovelty 1/4durability 1/4· Eric

“to do nine figures in revenue in less than two years from launches is a little bit wild.”

0.18

Ramp had over 100,000 cardholders across about 5,000 businesses, making it a large-scale relationship business where founders can talk with many customers regularly.

factualspeaker onlynovelty 1/4durability 1/4· Eric

“We have over a 100,000 cardholders, but these are 5,000 businesses.”