Move Fast, Own the Details: What Binance, Agoda and N26 Taught Mayur Kamat About Product

Kamat describes Binance as a company of almost unprecedented scale. It launched in 2017 as a crypto exchange, became number one within six months, and by his account reached a peak valuation of around $400 billion with about 2,000 employees. As head of product, his top-line KPI was trading volume, and his monthly figure was measured in trillions of dollars. Teams of 10 to 20 people ran multi-billion-dollar profit businesses.

That pace was only possible with a very flat organization. Founder and CEO CZ had around 55 direct reports, and for much of the company's history there was only one level between individual employees and him. One-on-ones were impossible, so the culture leaned on one-to-many communication. The leadership team met every day at 11 PM Singapore time, weekends and holidays included, which meant no decision waited longer than 24 hours.

Kamat took a working method from this that he calls running products via daily meeting. At the nightly call the team picked an owner for something urgent, and that owner went all-in until it was solved, reporting back every day. Some of the problems were huge, such as getting 15 financial licenses in 15 countries in three months. His conclusion is that smart people given hard problems and no constraints except time can move mountains quickly.

Asked what being in the details actually looks like, Kamat points to Binance's onboarding problem. Crypto had gone from sign-up with an email address to a full bank-style KYC flow, and conversion fell from roughly 100% to 2%. The team tracked a spreadsheet of about 500 cells, covering the top 50 countries and the top 10 document types, with the conversion rate in each. They asked pointed questions, like why a passport from Kazakhstan converted so poorly and whether a new vendor, better imaging or a new SDK would help.

CZ's rule was that no user gets left behind, since even one user in Congo represented financial inclusion. The company could also throw resources at the problem. A 20-person KYC team was once told it would have 500 people for three months. Kamat admits this has downsides, but says that if you care purely about execution, the power is surprising.

He had felt the same thing once before, on his first day at Google. An iOS 4 bug made every Gmail user re-sync their whole inbox, causing 1000x load on the servers. Google's head of infrastructure flipped a switch and thousands of times more servers appeared for two weeks. For Kamat the lesson is less about headcount and more about how quickly a large organization can maneuver when it decides something is urgent.

Lenny asks whether Kamat would carry the 50-reports, no-user-left-behind approach to other companies. He says it depends on preconditions. The company must be growing so fast that employees learn simply from seeing problems at scale. People must also be well compensated, and Binance's bonus structure ranged from 0% to 500%, so staff cared about KPIs more than about promotion ladders.

The third condition is a mission people actually believe in. At Binance, Kamat says, many employees came from countries where people lack access to financial tools or even a currency they can trust. For them, a dollar stablecoin was not speculation but protection from a currency losing value. That belief that they were empowering billions of people is why a lot of staff were willing to move mountains.

Kamat's first piece of career advice is to work at fast-growing companies, because learning compounds. He compares it to compound interest: even a low rate gives very different results if it compounds daily rather than yearly. At Microsoft, some products from his internship had not shipped when he left three or four years later, while at Binance something shipped every day or even every minute.

The second piece is to optimize for what you are great at. Strengths show up early, he says, and early in a career you cannot easily work around weaknesses the way a senior leader can by building a complementary team. A creative, experiment-minded PM belongs on a growth team, not on a compliance project, while a structured, high-EQ stakeholder manager would struggle in a fast growth team. Third, he says not to optimize for compensation in the first 15 years, because he expects 90% of an executive's pay to arrive in the last five.

His last point is the one he finds most interesting: decide early whether you truly want to be a C-level executive. Ambitious people drift into product management assuming that is the destination, but a great career can end as a director with a rich life outside work. He says this is not about sacrifice. If the work fills your cup, long hours will not feel like a sacrifice, and once someone is talking about work-life balance they are probably not on that track.

Asked about other paths, Kamat lists three. The first is founding a company, which suits PMs who feel frustrated by everything they cannot control, though he warns that much of running a company has nothing to do with building product. Founders he has worked for have told him they wish they had his job. The second is to deepen your expertise in a domain such as KYC onboarding, fraud, card schemes or loyalty and retention, and the third is to decide work is only one part of your identity.

On why N26 appears among the top companies for producing strong PMs, he points to three things. N26 was a category-defining first mobile bank, so each PM solved problems for the first time. Early shared success also creates a dense alumni network, like the PayPal mafia, and the founders are mission-driven people whose ambition raises everyone's definition of success. He adds that fintech PMs have two customers, users and the regulator, and nearly every trade-off is existential.

Lenny asks how people discover their strengths. Kamat mentions the book Now, Discover Your Strengths, later renamed Strengths Finder 2.0. He also describes a six-hour psychological assessment that Agoda paid about $5,000 per PM candidate to run, which measured aptitudes such as pattern recognition and numerical ability along with dominance and warmth. He took it reluctantly and found the results very accurate.

A cheaper option came from an executive offsite where Ray Dalio's approach was presented, a roughly $50 assessment in which you rate yourself and your leadership team rates you too. The overlap between the two is where your true strengths lie. Kamat also offers a simpler method: PM work breaks into a few repeated activities such as roadmaps, managing engineers, stakeholders, launches and data, so notice which ones you enjoy and which drain you. Then check whether a prospective employer is hiring you for the parts you dislike.

Lenny notes that new PMs often want to think about strategy, though they must earn that right. Kamat then offers a self-described controversial view: in product, if you can test a hypothesis, you do not need a strategy. If a feature will add a given number of users at a given conversion rate and acquisition cost, you test it, and if it works you do more of it. He says strategy slides often end in packaged intuition where the loudest voice wins, and he recalls a former Google product chief who told PMs to come with data, because otherwise his ideas win.

At N26 he operationalizes this with an experimentation platform, and he shouts out Statsig. He looks at which metrics experiments move, the p-value and how fast they reach significance. Where tests are impractical, such as compliance, legal matters or pricing in Europe, he asks for deeper pre-work. He also warns against reading aggregate dashboards: when Bitcoin crashed, sign-ups fell even though the product was doing well, and only a variant versus control comparison showed that.

He sees a cultural benefit too. Everyone thinks they can do a product manager's job, from the CFO to any user, whereas nobody tells the accountant how to cook the books. Building experimentation gives PMs a discipline and a science, so a bad idea can be answered with evidence from past tests. Getting a team there takes a long time, he says, but then every experiment gives a small dopamine hit.

Back on what N26 does well, Kamat says that in Europe, with far fewer unicorns than the Bay Area, being one of the first gives a hiring advantage, so the input is very smart people. The problems they face are harder, and the recent push toward experimentation gives them a skill set few companies have. He argues that companies world-class at experimentation are rare, and that nothing compounds wins faster than experiments.

He also likes banking as a business. Everyone needs a bank, so the addressable market never runs out, and the product is naturally self-hedged. When interest rates are high, spending falls but deposits earn money, and when rates fall spending rises and interchange income grows. N26 invested heavily in savings while rates were high and is now investing in lending as rates drop, so the portfolio complements the macro environment.

Lenny adds an observation from his research. Many of the companies that produce the best PMs, such as Intercom, Revolut and N26, are successful outside the US and so attract the top talent in their region. Kamat agrees the formula is simple: be the big fish in a small pond, draw the best talent, and work on hard problems.

For the trade-offs between regions, Kamat recommends starting a tech career on the US West Coast, which has the densest concentration of talent, growth companies and networks. He names exceptions: Dubai for crypto, Bangalore for Indian talent, and London and Singapore for finance. He notes immigration policy may make the US option harder now. A strong early brand from Microsoft, Google or OpenAI travels with you when you later move.

His own move to Thailand in 2018 was driven by family. With young kids in Seattle, a daughter with asthma and two working parents, Bangkok promised better weather and household help. That freed time for work, and Agoda became a bootcamp in experimentation-driven growth. He notes that Booking.com grew from about $10 billion to about $170 billion in ten years while still selling flights, hotels and cars, showing growth can come from optimizing small things.

Living abroad has pros: you learn what a truly global product needs, understand different laws and become a better partner to legal and compliance teams. The cons are lower pay, especially in Europe, and a ceiling. In Seattle or the Bay Area you can change jobs without moving, but moving from Bangkok to a bigger role meant relocating the family to Singapore, back to Thailand and then to Spain. He says job and location usually come as a package deal.

Kamat summarizes Shreyas Doshi's leverage idea: you always have more problems than time, so work on those with a 10x positive or negative impact. The multiple may be 3, 5 or 100 depending on the business. In most fintech companies, he says, that means growth or compliance. The principle holds at any level, but it gets harder as a CPO because every problem that reaches you is important.

He illustrates this with Alex Allgood, who ran White Pages when Kamat joined. Allgood had a movable desk and placed it in whichever department had the highest-leverage problem until it was solved. When Kamat asked for a PM focused purely on SEO for their caller ID app, Allgood instead became the product manager on that scrum for three months, since he had built White Pages on SEO. Kamat saw the same pattern with CZ, who sat on certain products himself, and whose lead PM there said no to him all the time out of mutual respect.

He pairs this with humility about the details, since many high-leverage problems are not grand strategy but questions about why something is not working. For his own schedule, he insists on a calendar with open space and says a full calendar is a badge of shame. Lenny compares the approach to what people call founder mode and recalls Brian Chesky doing something similar at Airbnb.

Kamat says he has not found a personal AI game changer yet, though N26 uses Gemini for meeting notes and Writer for copy across several languages. At the company level he sees three areas where AI changes things: coding, where he cites a rough 18 to 25% productivity boost for developers, customer support, where AI can automate 60 to 70% of basic questions, and fraud detection, where models can pick up on the language patterns of fraudsters. He jokes that tools which expand a few words into an essay while others compress it back are a reverse zip.

His most spectacular failure was being the first PM on Hangouts. Google had thousands of people, its full power and Larry sitting with the team, and still failed to build a great messaging product. After 15 years of reflection he points to solving for the wrong audience and the DNA of the company, noting that certain companies struggle in certain categories, like Google with social. The team did invent WebRTC, which he says now underpins video and voice calling across the world, and he later wrote the spec for an enterprise version that became Google Meet.

At Agoda he repeated the lesson. The team bought $50 Android devices at a mall to let hotels scan and charge cards by camera, and thousands of hotels used them. But licensing took six to nine months, and COVID hit just as it launched. His takeaway is not to take on projects that run six months to a year, and to shorten the time to validate a hypothesis. He closes by urging listeners to look for overlapping superpowers, between you and a company, and even between life partners.