Ramp just announced it has crossed a stunning $1B ARR — growing 100% (!) and just closed its latest VC round at a $32B valuation. Co-founder CTO Karim Atiyah came to SaaStr Annual + AI Summit to share how they scaled to the first $100m ARR. And see everyone at SaaStr 2026 from May 11-13!!
Ramp’s Toughest 5 Learnings Getting to The First $100M+:
- Asymmetric thinking requires capital cushion to execute. You can’t make $100,000/$10,000 bets if you can’t afford to lose $10,000. Being second-time founders let them raise an outsized seed round, which enabled going after hard problems with high barriers to entry. First-time founders often can’t access this playbook.
- Exceptional people are spiky, not well-rounded. The traditional hiring checklist optimizes for minimizing bad hires but is terrible at identifying 10x performers. Hiring Stanford freshmen with zero fintech experience—but who’d sold companies to Apple or were extraordinarily fast builders—required throwing out conventional wisdom.
- The cost of credit mistakes is 100x the reward. With a business model generating ~1% profit on deployed capital, every underwriting error requires 100 successful customers to break even. Some areas have negative asymmetry—where downside vastly exceeds upside—and require extreme conservatism even while taking huge risks elsewhere.
- Fundraising terms compound more than valuation. Clean terms from reputable investors in your first institutional round make every subsequent raise easier. Bad early terms (aggressive liquidation preferences, protective covenants, unfavorable board composition) become anchors that narrow your path to a large outcome.
- Product velocity is a compounding advantage. Choosing vendors like Linear over Jira because they enabled faster development cycles—even without feature parity—meant that speed advantage compounded as the team grew. Small velocity differences early become massive advantages at scale.
The Setup: From Consumer Savings to Business Spend
Before Ramp, the founding team had different but complementary paths. Eric Glyman founded Paribus in 2014, a consumer fintech that automatically secured refunds when product prices dropped after purchase. At its peak, Paribus saved 10 million customers millions annually on online purchases. Capital One acquired the company three years later.
Karim Atiyah’s journey started differently. Born and raised in Lebanon, he came to the US in 2007 to study, first landing in New York for consulting work before quickly returning to his first love: writing code and building software. After co-founding and selling Paribus to Capital One, he spent two years in the card industry learning the infrastructure that would later power Ramp.
But here’s what’s remarkable: In the same timeframe it took to build and sell Paribus, Ramp achieved an outcome 100 times larger. By 2019, when Eric and Karim started Ramp together—Eric as CEO, Karim as CTO—that acceleration wasn’t luck. It was systematic thinking about risk.
The Asymmetric Risk Framework
Most startup advice centers on de-risking. Find product-market fit. Validate your market. Minimize downside. Eric and Karim took the opposite approach: maximize upside potential, even if it means accepting higher barriers to entry.
Karim’s framework is simple: Look for bets where success delivers 10x returns while failure costs 1x. Don’t minimize downside—maximize upside.
“Think about it like poker,” Karim explains. “You’re dealt a mediocre hand, but the pot is enormous. The expected value calculation changes completely.” Or consider a coin flip: heads wins $100,000, tails loses $10,000. That’s asymmetric. The catch? You need to be able to lose $10,000 to play.
They call this concept internally “asymmetric outcomes”—and it applies across every aspect of building the company: problem space selection, hiring, vendor partnerships, product development, and fundraising.
Choosing Problem Spaces: Going After Hard Problems
Ramp’s decision to tackle business spend management wasn’t obvious. Financial services has massive barriers to entry. Most banks are over 100 years old. American Express, Ramp’s incumbent competitor, is 161 years old. You need banks to work with you. Card manufacturers must produce physical cards for you. The switching costs and regulatory complexity are real.
But that’s precisely why they chose this space.
As second-time founders with a successful exit, Eric and Karim raised an outsized seed round—giving them the ability to “lose $10,000” in the metaphor above. That capital cushion let Ramp pursue a problem space where barriers protect you if you succeed. The harder the problem, the more valuable the moat once you solve it.
“There’s a lot of debt in financial services,” Karim notes. “It’s not random that most banks are more than 100 years old. Amex is 161 years old at this point. In financial services there are high barriers to entry—you need to convince banks to work with you, card manufacturers to make cards for you. The fact that we were second-time founders allowed us to raise an outsized seed fund, and as a result we decided to go after a bigger space where it’s harder to start.”
The implicit lesson: If you can only raise $1 million in seed funding, don’t try to launch rockets into space. But if you can raise more, the opportunity set expands dramatically.
Hiring: Slope Over Experience
The average hiring advice follows a checklist approach: list required competencies, interview against each dimension, hire candidates who check every box. This methodology optimizes for minimizing bad hires. It’s also terrible at identifying 10x performers.
“Most people who are exceptional at what they do tend to be very spiky,” Karim explains. “They don’t have an evenly distributed skill set, and that first approach makes it very easy to find the flaws and disqualify candidates, but it makes it very hard to find the people who have 10x potential and can grow with you.”
Ramp’s early technical hires illustrate this. Karim brought on two Stanford freshmen he’d worked with as interns:
- Veral: Sold a company to Apple as a freshman. Exceptional iOS developer. Zero fintech experience.
- Calvin: Extraordinarily fast at building software. Zero fintech experience. Didn’t have the communication polish of senior engineers.
Neither candidate would pass a traditional hiring rubric for an early-stage fintech startup. Both became foundational to Ramp’s engineering culture.
“These are the types of people that we chose to hire early on, not because they had the experience and they were low-risk bets, but because they had the slope that we were looking for,” Karim says. “When you hire these people they scale with you, they scale with the company, and it makes it much easier to keep growing.”
Karim jokes that they try to “hire for a slope that matches our logo”—referring to Ramp’s upward-sloping brand mark. The rate of improvement matters more than current position. These engineers scaled with the company rather than becoming bottlenecks as complexity increased.
Vendor Selection: The Jira Trap
“No one ever got fired for choosing IBM” became “no one ever got fired for choosing Jira.” Both statements reflect the same risk-averse thinking that caps upside potential.
Jira does everything adequately. If things go wrong, you have defensive cover—everyone uses Jira. But making obvious, safe vendor choices means missing opportunities with up-and-coming companies that could alter your trajectory.
“The problem with making these really obvious safe choices is you miss out on potential opportunities of working with an up-and-coming vendor that can really alter the trajectory of your company if it works out,” Karim notes.
Shopify and Lyft bet on Stripe extremely early, when it was a non-obvious choice. Stripe offered something differentiated: it made onboarding trivial for developers. That single advantage compounded over time.
Ramp identified speed as a core competitive advantage. When evaluating issue tracking software, they chose Linear over Jira. Linear didn’t have feature parity with Jira, but it enabled much faster development cycles. That velocity advantage compounded as the team grew from 10 to 170 engineers.
The switching costs when you’re early-stage are minimal. “You have two or three accounts to replace, you only have a few issues to move over, so why not try it and make this bet?” Karim asks.
When it works, you build deep partnerships. The vendor invests in your success. You shape their product roadmap. Your interests align. Ramp still has that cake their friends at SwagUp sent them—a literal example of having your cake and eating it too.
“When you’re working with these scrappy early-stage vendors, you’re able to build relationships, help shape their product vision so it can match your product vision. You have skin in the game in each other’s success, so it makes things a lot smoother and allows you to move a lot faster.”
Product: Solving Problems, Not Adding Features
Customers famously wanted a faster horse, not a car. If you build by asking customers for their next desired feature, you end up with the TV remote on the left—dozens of buttons because you kept asking “what else do you need?”
The remote on the right exists because someone asked about the underlying problem, not the solution.
“There’s a big difference between the remote on the left and the remote on the right,” Karim explains, showing two contrasting images. “On the first one maybe you just went and asked your customers ‘which other button do you need, what’s the next feature that you want’ and you end up with something that looks like the one on the left. But if you make big bold bets on product and listen to your customers’ problems—not the solutions they’re giving you—you’re able to actually achieve big leaps and let your product actually help you in the market.”
This doesn’t mean ignoring customers. It means distinguishing between what they say they want and what would actually change their behavior. The 10x product comes from the latter.
Negative Asymmetry: When to Be Conservative
Not every bet should maximize upside. Some have negative asymmetry—where downside vastly exceeds upside potential. These areas require extreme conservatism.
At Ramp, credit underwriting falls into this category. Ramp extends credit to businesses who pay back at month-end. The business model generates approximately 1% profit on deployed capital.
“For every mistake that we make, we need 100 customers to pay us back,” Karim explains. “So this is an area where we’ve decided to not take any risk, because the cost of making a mistake is 100 times bigger than the reward if we get it right.”
Every company has similar negative asymmetry zones. A security breach. A GDPR violation. A customer satisfaction collapse. “Try to figure out the areas where you can take a lot of risks and make these big bold bets, and isolate them from the ones where you shouldn’t be,” Karim advises.
The discipline isn’t just taking big risks—it’s knowing where NOT to take them.
Fundraising: Terms Over Price
Headlines celebrate high valuations and large rounds. Price dominates the narrative. But price is almost never a good indicator of fundraising success.
“It’s easy to get drawn by the headlines—big prices, large rounds—and think that price is really everything,” Karim says. “But in reality, the price is almost never a good indicator of how well the fundraising went. There’s a lot that goes into fundraising, and the terms that you raise at are extremely important, particularly when you’re raising your first institutional round.”
Terms matter enormously, especially in your first institutional round. Raising from reputable investors typically means clean terms: no aggressive liquidation preferences, minimal protective covenants, reasonable board composition. These terms compound positively through every subsequent round.
“If you’re able to attract a reputable investor, generally what reputable investors come with is really clean terms. You avoid liquid things like liquidation preferences or restrictive covenants or giving up too many board seats early on. And as a result, if you’re able to raise from reputable investors on very clean terms early on, it makes every other fundraising conversation much easier down the line.”
Bad early terms become anchors. Every future investor inherits the precedent. Every conversation gets harder. The path to a large outcome narrows.
“You don’t want to optimize for that first fundraise, but every time you fundraise you want to be thinking about all the other rounds that you’re going to have to do after that, especially if you’re trying to go after a big outcome.”
Beyond terms, reputable early investors unlock asymmetric advantages for B2B companies:
- Customer Access: Potential customers take calls because of your investors
- Angel Investor Attraction: “If you’re able to attract a really reputable investor even if it’s not at the highest price, and you’re a B2B company, you can call up your potential customers, you can call up angel investors. Today a lot of founders are angel investors and they get asymmetric upside from investing alongside these reputable investors because they don’t have to pay the 2% and 20% that the VCs charge.” Founders get to invest alongside Founders Fund, Sequoia, and Andreessen without management fees—a massive arbitrage. Many become customers and advisors.
- Future Round Facilitation: Later investors underwrite your business partly based on who backed you early
The second-order effects of investor selection dwarf the first-order effect of valuation.
The Framework in Practice
Eric and Karim’s framework reduces to a simple heuristic: In every major decision, ask whether the outcome distribution is asymmetric. Can you win 10x while losing 1x?
Applied systematically:
- Problem space: Choose hard problems with massive TAM over easy problems with obvious paths
- Hiring: Optimize for slope and spikiness over well-rounded competence
- Vendors: Partner with emerging companies that excel in your priority dimensions
- Product: Make bold bets on problems, not incremental feature additions
- Fundraising: Optimize for term quality and investor value-add, not valuation
- Risk management: Identify negative asymmetry zones and build absolute protection
The meta-lesson from Karim: “If you think about startups generally, people tell you most startups fail, 90% of companies don’t make it. It’s like a plane is crashing and you’re trying to rebuild that plane before gravity just pulls you down and crashes you. So you only have a limited amount of time to make a couple of bets, and you want to be focusing on the ones that have a potential to give you that escape velocity and really take off. Instead of asking people for advice and looking for the average of all the advice you get out there, you should probably be taking bigger, bolder bets—because if the average is failure, you don’t want to be taking average advice.”
“Bigger bets don’t have to be scarier,” Karim concludes. “They tend to be less risky actually, especially if you look for opportunities to have an asymmetric outcome in the case of success.”
Their Top 5 Mistakes Getting to The First $100M+:
- Optimizing for valuation over terms in early rounds. They see founders celebrate high-priced seed or Series A rounds while accepting aggressive liquidation preferences, restrictive covenants, or giving up multiple board seats. These terms become anchors in every subsequent round. A lower valuation with clean terms from a reputable investor is almost always better—the second-order effects (customer introductions, angel investor attraction, easier future raises) massively outweigh the first-order effect of price.
- Choosing safe, established vendors to minimize perceived risk. The Jira trap is real. When you’re small, switching costs are minimal—you have 2-3 accounts, a handful of issues to migrate. That’s precisely when you should make asymmetric bets on emerging vendors that excel in capabilities you care about. Optimizing for “no one will blame me if this fails” means missing vendors that could 10x a core capability. The cost of vendor mistakes at 10 employees is tiny; the benefit of finding a Linear or Stripe is enormous.
- Using hiring checklists that optimize for minimizing bad hires instead of finding 10x performers. The competency checklist approach makes it easy to disqualify spiky, exceptional people who don’t fit the mold. Hiring Stanford freshmen with zero fintech experience seems crazy until you realize they had slope—they’d sold companies to Apple, built at exceptional speed, and could scale with the company. Looking for well-rounded candidates means missing the people who can actually alter your trajectory.
- Listening to customer solutions instead of customer problems. Building by asking “what feature do you want next” gets you the TV remote with 47 buttons. The 10x product comes from understanding the underlying problem and making bold bets on solutions customers can’t articulate. There’s a reason people wanted faster horses, not cars—customers are better at describing problems than prescribing solutions.
- Taking uniform risk across all business areas instead of isolating negative asymmetry zones. The asymmetric framework isn’t about taking maximum risk everywhere—it’s about identifying where upside is 10x downside versus where downside is 100x upside. At Ramp, credit underwriting has massive negative asymmetry (1% profit margin means every mistake requires 100 successful customers to break even), so they take zero risk there. That conservatism frees them to take huge risks in product, hiring, and vendor selection. Most founders either take too little risk everywhere or fail to protect the critical failure modes that could sink the company.
Most founders are playing the $1,000/$1,000 coin flip game, grinding out incremental progress. Eric and Karim built Ramp by consistently finding and making $100,000/$10,000 bets instead.
That’s how you build the first $100 million—and the next billion after that.
