Brandon Carson

Left Stripe Sessions last week with two takeaways: 1. There's going to be a new Bloomberg - meaning proprietary (likely private) APIs and reliable data for transactions, and it won't be a SaaS or seats-based model. 2. No one knows how to price anything right now. We're all making it up. Starting with the Bloomberg: there's a gap in the market for an API-native, agent-first data layer that surfaces real-time information to AI systems making decisions. Know Perplexity occasionally gets clowned by Finance Twitter for underestimating Bloomberg (and not saying that's all Bloomberg is!) - but just in terms of protocol, that the real-time & agent-first aspect is critical to AI developer experience. At Sessions, the agentic commerce story was nearly all consumer and AI shopping. Makes sense as e-commerce volume is massive, so infrastructure will move there first, but don't think it's as large of a change to core value proposition as B2B. When AI agents start autonomously executing workflows - qualifying leads, enriching contacts, booking meetings, prioritizing pipeline - they need data that's: 1. Real-time (not quarterly exports) 2. Structured for machine consumption (not dashboards) 3. Comprehensive and reliable enough to make decisions on Bloomberg built a $15b ARR business surfacing financial data to humans making decisions. The parallel for B2B operational data - contact intelligence, company signals, intent data - is likely Zoominfo (one-stop shop), which is also historically seat-based. GTM is just one use case. The actual opportunity is every domain where AI is making consequential decisions and needs ground truth in real-time.

17

4 Comments

A lot of STR operators think: “Contractors are expensive. If I bring ops in-house, I’ll save money.” On paper, it looks smart. In practice, it usually backfires: higher costs, more headaches, and a second business to manage.

7

As a Stripe customer, we received an annual letter that is basically a snapshot of where the internet economy is heading - and it’s accelerating fast. A few numbers that jumped out at me: / Businesses on Stripe processed $1.9T in 2025 total volume (+34% YoY) - about 1.6% of global GDP. / Stripe now supports 5M+ businesses, and 200M+ people use Link. / They shipped 350+ product updates, acquired Privy (110M+ wallets) and Metronome (usage-based billing powering OpenAI/Anthropic/NVIDIA), and the Revenue suite is trending toward a $1B run rate. / Macro trend Stripe is calling out: the “sorting machine” is spinning faster - profit and growth are concentrating, and software + data centers drove ~46% of US GDP growth in 2025. / Startup trend: building is getting easier, so the pace is spiking. Stripe saw a new cohort growing ~50% faster than in 2024, with double the number of companies hitting $10M ARR within 3 months. The part I’d bet on: “The five levels of agentic commerce” Stripe frames agentic commerce as a ladder and says we’re hovering around levels 1–2 today: 1. Agents fill forms (you decide; it executes) 2. Descriptive search (“I need back-to-school supplies for…” not keyword search) 3. Persistence (it remembers your preferences) 4. Delegation (it buys for you under constraints) 5. Anticipation (no prompt; it knows and acts) And Stripe is clearly preparing rails for this: ACP with OpenAI, Shared Payment Tokens, an Agentic Commerce Suite, plus “machine payments” for agents paying for APIs/tools. Why this matters for founders, startups, and product teams: / “Global-by-default” is now table stakes: localised checkout, local payment methods, taxes - launched to 100+ countries “at once,” with examples like Gamma’s +22% India revenue after enabling UPI. / Stablecoins are quietly becoming a B2B payment rail: Stripe cites stablecoin payments volume of around $400B in 2025 (majority B2B) and is building for scale with Tempo (a payments-first blockchain). / Payments are still one of the highest ROI growth levers (auth rates, fraud, conversion) #Stripe #AI #AgenticCommerce #Startups #Payments #Product

19

1 Comment

I spent some time at Colorado Startup Week this week and here's the best thing I learned for growing companies: The 5 Conditions Framework, developed by TrueSpace. It's framework for startups-to-scale-ups and a practical checklist founders can use to figure out if they’re ready to graduate from starting to scaling. Condition #1 on the list: Alignment. Are you and your team aligned around three elements? 1️⃣ Direction — Where are you trying to go, and why? 2️⃣ Focus — What’s your market, who are your customers, and what’s your point of view? 3️⃣ Structure — Do you have the leadership and capital to get there? As a marketer, I love seeing alignment lead the list. Mel Bell, Garrett Jestice and others have clearly articulated lately the huge upside GTM alignment brings to teams, with speed the most obvious benefit. Driving alignment is a superpower every marketer should develop. We intuitively know the value of getting teams aligned and the fuel it adds to GTM execution. And it's great affirmation to see the deep datasets and analysis that Charles Fred and his team have conducted back up what many marketers know by intuition/experience/scar tissue. Also, a bit of reality check for marketers: There's five conditions on the list ... and the other five are probably harder than alignment. You can download the complete 5 Conditions Framework at TrueSpace's website.

8

6 Comments

Having coffee with YC founders and indie hackers in San Francisco taught me one thing. Nobody is obsessed with features. They’re obsessed with removing pain. Every conversation comes back to the same themes: On-call burnout. Repetitive operational work. Being paged at 2 AM for problems that follow patterns. What stood out most was the mindset. No ego. No flexing. Just curiosity. The best builders aren’t asking, “How do I build more?” They’re asking, “What can we delete?” Delete boring work. Delete constant alerts. Delete the need for humans to babysit systems. That’s where real leverage lives. SF doesn’t reward noise. It rewards founders who listen deeply and build what shouldn’t exist anymore.

24

11 Comments

It's not just VCs seeing lots of people aspiring to be like Palantir. I'm also seeing plenty of demand for Forward Deployed Engineers, the role Palantir created (or least carved out a name for). Everyone wants their own version of FDEs. It's a common feature with the hiring conversations I'm having with AI founders, and FDE job postings are up over 1000% according to the Wall Street Journal. But of course hiring someone with that job title won't magically give you Palantir's outcomes. Because Palantir's success isn't about a job title. It's about systems, culture, and context that took years to build. The a16z article explains brilliantly what you might be able to steal from Palantir, and what you might need to work out for yourself. https://lnkd.in/ecqwyE7D As ever, job titles are convenient shortcuts but terrible substitutes for clear thinking about what work actually needs doing. Before you hire a "forward deployed engineer," ask yourself: What problem are we solving? What does success look like? What's different about our context vs Palantir's? Those answers matter more than the job title ever will.

4

2 Comments

Blueprint Equity led Develo's $14M Series A, and I couldn't be more excited. I spend my days evaluating vertical software companies, and "AI-native" is the most overused phrase in our inboxes. Develo is the real thing. Aaron Sin and Han Ke didn't bolt AI onto an EMR. They rebuilt the operating system itself for pediatrics, unifying clinical, practice management, billing, and family engagement with an embedded AI scribe, automated charge capture, and chart intelligence native to an end-to-end core product. It's the first and only AI-native operating system purpose-built for pediatrics, and it's already running for hundreds of pediatricians across more than two dozen states. Proud to back this round alongside Villain Capital, z21 Ventures, Bienville Capital, and operators who know healthcare cold: Chris Sullens (CentralReach), Chris Barton (Community CareLink), and Paul Black (former Allscripts CEO). Congrats Aaron Sin, Han Ke, and the entire Develo team. Let's go build.

239

13 Comments

A lot of e-commerce teams are doing good work, they’re just buried under bad data and clunky integrations. We built Syndic8 from day one to deal with that mess directly, instead of bolting syndication onto a system that wasn’t designed for it. Syndic8 is designed to push product data across retailer systems and channels, avoiding the legacy code, costly rework, and slow integrations that plague many competitors. Today we syndicate to 220+ destinations. We started native, so our clients see value faster: cleaner integrations, fewer errors, and less lift to get products live everywhere they sell. Our open architecture makes add-ons simple and low-cost, and our AI agents find and fix data gaps before they block sales. If your syndication partner still looks like it's built from scotch tape, it’s time to ask who’s really doing the heavy lifting, and who’s just charging you for it. DM me for a 10-minute demo or request our 30-day pilot playbook — see localization time drop from weeks to minutes.

7

Hectocorns: Crossing $100B makes headlines. (Staying there is the next challenge) Databricks just joined the “hectocorn” club with a $100B+ valuation. It’s an elite list, which includes OpenAI, SpaceX, Anthropic, ByteDance, and Ant Group. Only 30 since 1990. But here’s what the headlines don’t mention: 👉 Legacy giants like Cargill, Koch, and Mars solved liquidity decades ago with phantom stock and profit-sharing. Employees didn’t need an IPO or a secondary to realize value. Today’s top-tier startups? They’ll need to get creative to attract and retain talent. Without pathways to liquidity, talent will drift toward competitors, a risk magnified in the Magnificent Seven talent wars and today’s AI poaching battles. * Secondaries: OpenAI’s $10B Microsoft-anchored deal in 2023 showed how urgently employees needed a path to cash. * IPO or bust: SpaceX and Stripe face constant speculation about when they’ll “finally” go public. * Creative financing: From structured buybacks to hybrid instruments, leaders are scrambling for ways to incentivize and retain people in the absence of public markets. * Stock as M&A Currency: Another wrinkle is the use of stock as currency for acquisitions. In public markets, it’s second nature for companies to lean on their equity to buy growth — Facebook’s $19B WhatsApp deal being the classic example. Private hectocorns, even at $100B valuations, don’t enjoy the same fungibility. Get the full scoop in my article below: https://lnkd.in/gATaYHrY Savvy’s Take: Valuation gets you into the $100B club. Liquidity keeps you there. © 2025. The Savvy Startup Advisor, LLC. All rights reserved. #Startups #VentureCapital #Entrepreneurship #Leadership #TalentRetention #TalentAcquisition #Hectocorn #Valuation #Liquidity #PrivateMarkets #AIStartups #VCHorizons #TalentWars #EmployeeOwnership #CapitalAllocation #ESOP

4

Did you know switching industries doesn’t have to mean sacrificing salary? My student Jask recently shared his story on my podcast: 👉 He transitioned into fintech, an industry he’s truly passionate about. 👉 Instead of a pay cut, his salary jumped by 40% when he accepted an offer from Capital One. What surprised him most? That pursuing a role in an industry he loved actually led to a big salary increase. Sometimes the right career move doesn’t just align with your passion—it also accelerates your growth. Watch the full episode now: https://lnkd.in/eNVTJzz2 Want to know what salary you should be aiming for?  I created a Product Manager Salary Table that gives you customized salary strategies based on your years of experience. 📥 Download it here: https://lnkd.in/e-hZkFwB #PMAccelerator

10

1 Comment

Excited to share Wizehire’s new partnership with Paycor! It’s been a real pleasure working alongside Judy Dolton and Peter Canter throughout this process; their collaboration made all the difference. Looking forward to what we’ll accomplish together and the value this partnership will bring to our customers. Check out Wizehire's new presence in Paycor marketplace: https://lnkd.in/exYCpqtX

7

Zillow scaled AI to 25,000+ homes. Then shut it down. The failure did not start at shutdown. It started when small overrides became normal. Adoption strong. Expansion accelerating. Revenue still looked healthy. But markets moved faster than the model. So teams stepped in. Manual fixes. Relaxed criteria. Quiet adjustments. Nothing dramatic.   Just drift. AI failure is rarely technical. It starts when leaders miss early shifts. I see this before numbers fall. The dashboard says “on track.” Fewer questions get asked. Protecting the system means clarity is slipping. You are no longer scaling strategy. You are compensating for it. Strong leaders watch behaviour. Where are outputs being corrected? Where is trust eroding? Overrides are not technical noise. They are leadership signals.  Scaling AI? Audit the overrides. If this feels familiar, message me “CLARITY.”

37

32 Comments

Last week, Ramp raised $500M, pushing its valuation from $16B in June to $22.5B. That’s a 41% jump in weeks, and a 73% increase since March. Fintech funding is rebounding after the 2022-2024 “winter.” Before that, global VC funding hit $127.7B in 2021, capturing 20% of all VC dollars. After the end of zero interest rate policies (ZIRP), which fueled easy capital, fintech funding tanked starting in late 2021. Global fintech VC funding collapsed to $30B in 2023, while the share of VC dollars dropped from 1 in 5 to just 1 in 12 by 2024. Since late 2024, fintech has been bouncing back, though it’s not at 2020-2021 levels yet. Big rounds are happening across stages, especially in enterprise fintech, with players like Sardine ($70M), Imprint ($75M), Airwallex ($301M), Rippling ($460M), and Plaid ($575M) leading the charge. A few interesting trends reported by PitchBook: • Median deal sizes hit $6.5M in Q2 2025, up 29% year-over-year • Enterprise fintech is dominating, capturing 74.6% of 2025 funding so far • AI is a game-changer (no surprises there), with AI-native fintechs generating 4x value per dollar invested and commanding 242% valuation premiums Public markets are heating up too, as I’ve written before. Circle had a massive IPO and has climbed over 200% YTD. Robinhood surged 157% YTD, making it one of 2025’s top-performing stocks (regrettably, I sold my shares last year). With more fintech companies likely going public soon, the rebound might just be getting started.

32

The best part about working at Ramp (other than the people and the Cometeer coffee) is the context and visibility we have into how companies spend money. September's Top SaaS Vendors shows what businesses are actually buying, based on billions of $ across anonymized card + bill pay transactions. A few takeaways we're paying attention to: 🤖 "Agent Stack" momentum: fastest growers included Augment Code and Braintrust. Companies are investing dollars into agentic evals + observability. 🤝 Next gen GTM enablement/data companies: spend on DataLane and People Data Labs shows that companies want fresher contacts and richer context in their GTM funnel tools. 📦 Pipe17 (order management system) cracked both our top AND fastest growing vendors this month. Picking the right vendor is critical. If you’re a finance, data, or procurement leader, Ramp data is the best way to benchmark your stack against where the market is moving.

82

4 Comments

SF AI-Native Operator Takeaway #2: In AI-native PLG, the hard part isn’t conversion... it’s discovery. Many AI-native teams are still talking about PLG using a classic SaaS mental model, but based on operator conversations in SF, that model is starting to break down in fairly obvious ways. The biggest bottleneck right now isn’t conversion. It’s discovery. In traditional PLG, users generally understood the category before they ever signed up. The problem was obvious, the product’s value was legible from the homepage, and the “aha” moment tended to show up quickly in first use. In that world, PLG meant optimizing onboarding, reducing friction, and improving free-to-paid conversion because user intent already existed. AI changes that assumption. In AI-native products, users are often curious but unclear. They don’t yet know what’s possible, value depends heavily on workflow, context, data, and role, and the product can feel abstract until it’s applied directly to their job. As a result, many users stall not because the product isn’t valuable, but because they haven’t discovered how it fits into their world and how they can't live without it. This is the real distinction people kept coming back to. PLG conversion answers, “Is this worth paying for?” PLG discovery answers, “What problem does this solve for me, right now?” What’s working best in practice is less about funnel polish and more about clarity up front: role- or workflow-specific entry points, guided examples instead of blank states, and opinionated first actions that show users a concrete outcome before asking them to explore. This also explains a broader pattern across AI-native companies. Forward-deployed teams and services-heavy delivery aren’t just implementation tools; they’re discovery mechanisms. They translate abstract AI capability into concrete workflow value, observe real use cases users wouldn’t self-discover, and feed those learnings back into what eventually becomes productized. PLG isn’t going away, but in AI-native companies it’s being redefined. Self-serve no longer means self-explanatory. Education becomes part of the product, and discovery has to come before optimization. The teams making progress aren’t obsessing over conversion rates yet. They’re focused on whether users see themselves in the product, how quickly they reach a meaningful outcome, and whether the product helps users get to a meaningful outcome for themselves quickly, without too much guesswork. Bottom line: in AI, PLG is less about removing conversion friction early and much more about creating understanding first. Once they understand, they may be hooked. Tomorrow is my last SF AI operator takeaway focusing on everyone's favorite topic du jour: 996 work schedules.

15

1 Comment

This is very true. The valuation trap is real. But I don’t think this is a YC story. It’s a market story. We’re seeing the exact same thing here in Israel. Seed rounds at 30–40M for companies with no product, no real validation, sometimes barely a prototype. Not second-time founders. First-timers. I’ve personally seen more than 20 rounds like this just last year. This isn’t about YC “pushing” companies. The system as a whole hasn’t adapted. The cost of building collapsed. The speed of iteration exploded. But pricing logic stayed anchored in 2021. When capital doesn’t recalibrate to reality, it doesn’t just create paper risk. It creates structural fragility. Good companies get boxed into impossible expectations. They are forced to grow into valuations instead of growing into product-market fit. That’s how promising businesses die early. The tragedy isn’t overvaluation. The tragedy is killing companies that could have become durable, profitable, meaningful businesses. This isn’t a YC issue. It’s a venture industry inertia issue. And the longer it takes to adjust, the more collateral damage we’ll keep seeing.

5

I started 2025 trying to pivot so LavaReach could find stronger product-market fit. I’m ending the year as Head of Finance at Landbase, having helped raise a $30M Series A. In between those two points was a lot of ambiguity, "founder-mode" grind, and one lesson: executing beats planning every time. Here is what the last 12 months taught me about luck, leverage, and scaling: 1. Luck is just volume plus action My journey to Landbase started with a cold LinkedIn DM to Daniel Saks. That message turned into an intro call, which snowballed into LavaReach getting acquired. Aside from just reaching out to people in your industry.Building relevancy and proofs will attract the right sphere of network. 2. You have to earn the right to model the data When I joined Landbase in February, I didn't immediately bury myself in the financial model. I got on the phone, booked the first few outbound meetings, and stood up the SDR function from scratch in my first month. You can’t build a real P&L or revenue model until you understand how a lead actually becomes a customer. Knowing how someone turn from never hearing about your company to a retained customer is key to building model that scales. 3. Founder Mode means filling the gaps Helping close our $30M Series A led by Sound Ventures and Picus Capital earlier in the year required more than just showcasing proof points and models. Especially in such a competitive industry where everyone is launching similar value props daily. It meant understanding the space deeply and having the context at the details level as well as knowing the bigger picture. It meant acting as the "adult in the room" while still being willing to do the unglamorous work. This was easy since I built LavaReach for almost 2 years, I knew where we had been and where it's going. 4. Growth is uncomfortable We’ve had to make tough calls this year to restructure the team. That led us to realize you need "founder density", a team of people who treat the business like it’s their own. Real execution isn't about maintaining the status quo; it’s about constantly raising the bar. 2025 was laying the foundation. 2026 is when we’ll scale 🚀. What was your biggest lesson in execution this year?

69

7 Comments

Finance firms got fined $1.23 billion in the first 6 months of 2025. 417% jump from last year. I've been talking to a lot of lending and fraud teams lately. Their eng stack is solid. Datadog, LangSmith, whatever they're running. Observability is there. But when compliance needs something audit-ready it falls apart fast. Because audit-ready means something specific. Why did your agent decline this application, on this date, for this customer, with this model version. That answer doesn't live in Datadog. It lives nowhere. So someone files a ticket. Engineering fits it in between product work. Compliance manually exports what they can. Audit gets through barely. Every quarter. EU AI Act fines go up to €35M or 7% of global revenue. SEC handed out $600M in recordkeeping penalties last year. Average compliance breach costs $14.8M. The problem is not the observability. Nobody built the compliance layer on top of it. Running early pilots with US lending teams. Good infra, but compliance still doing manual prep before every audit. Going well so far. Opening 3 more spots before we go wider. One agent, one audit cycle, see what your trail actually looks like. DM me or drop a comment if relevant.

11

3 Comments

Product-market fit genuinely feels different...and it'll show up in the data Demo to close rates are the easiest thing to look at for product-market fit right now. Vertical AI startups in our portfolio are seeing 50%+ demo to close rates, even with brand new account executives selling the product. It's happening in every market from AI takeoffs to AI CSR/Voice software to AI procurement solutions. When you see a demo to close rate over 50% chances are you have a jaw dropping customer experience and you should put your foot on the gas. If you're seeing something dramatically lower in mid-market or SMB ACV categories, it's worth revisiting your product and/or your pitch.

60

14 Comments