Big @firstround news today! We’re launching Product-Market Fit Method (a free intensive 14-week experience for early founders building epic B2B SaaS companies) and publishing the first session on our internal framework for all to read (with benchmarks, Looker's real data, and tactical advice from iconic enterprise founders). Even though finding product-market fit is the single most important thing for a startup, it’s still underexplored and seen as more art than science. We wanted to change that. I’ve personally talked to hundreds of founders about this topic, digging into what they did in the first 6-9 months of company building. (We’ve published dozens of those interviews on The Review in our “Paths to PMF” series.) This video previews some of what we learned — thanks to @christinacaci, @zachperret, @lloydtabb, @jboehmig, & @jaltma for sharing their lessons! In addition to that research, we’ve also drawn from our own 20 years of data and 500+ pre-PMF investments. What emerged was a very consistent set of patterns for sales-led B2B companies — the basis for our new framework and PMF Method’s 8 tactical sessions. In the program, we help early founders discover what customers really want, build the right v1 product, and close their first enterprise sales. We ran a beta version late last year with a tight-knit group of founders (ex Stripe, Plaid, Airbnb, Twitter, Greenhouse, Grammarly) and the feedback was great — my personal favorite was: "I feel like I shaved 12 months off the time it would take us to get to PMF.” Here are a few key dates and details: - The Summer 2024 session of PMF Method runs 5/29 - 8/28. - Application deadline is 11:59 PDT May 7th. - Any early founder working on a new B2B SaaS company is welcome to apply. Bonus points if you’re technical, have a clear product idea but haven’t raised yet and are <12 months into working full time on your idea. - PMF Method is 100% free. It costs you $0 and we own 0% of your company. Like with The First Round Review and Angel Track, our mindset is to openly share knowledge that we’ve put hundreds of hours of work into curating with the broader startup community, and give it away for free. That’s why we’ve also published our framework, so every builder can use this resource, even if they don’t do the program (it’s linked in the next post). Check out the links below for more details. Can’t wait to read applications!

Very rare to add a new app to my main home screen but I’m excited about this one

1/ Excited to finally share some news I've been sitting on: As @phineasb and I announced on @twentyminutevc this morning, I'm officially joining @firstround as a Venture Partner. 😀 Thrilled to join the team that invested in me back when I was a founder. medium.com/@firstround/addin…

10 years ago, @paraga and I were working together to introduce ranking to Twitter’s timeline for hundreds of millions of consumers. Now he’s started @p0 to build infrastructure for a very different kind of user: AI agents. And in a full circle moment, we’re able to work together again, as @firstround invested in his seed round last year alongside @khoslaventures and @indexventures. Parag and I had a chance to sit down for a great conversation last week (his first interview since leaving Twitter!). It was a fascinating reflection on what he learned and what he’s doing differently. We dig into: - The transition from CTO → CEO → Founder - What he was *really* thinking when all the Twitter drama was going down - Why he chose to be a founder (vs. take another big role) - How he landed on the idea for Parallel - What fundamentally changes when you’re designing for AIs vs. humans - How he thought about fundraising and building his early team - How working with customers like Clay shaped Parallel’s earliest products - The case for building “slow” APIs, and the advice he got from @patrickc - His takes on the modern agent stack and how evals need to evolve And much more! Immensely proud to be able to partner with Parag and the entire Parallel team. They’ve made incredible progress in a short amount of time and I can’t wait for more folks to start playing around with the APIs they’ve built. The full conversation is below, as well as more details on what Parallel is up to.

How do you evaluate products as a seed investor? (part 1) 🧵👇

Seeing founders pitch straight from Figma recently (instead of using Keynote/Google Slides/PDF). Maybe since they're in Figma all the time already, they just use it when they need to make "slides"?

2014: Working with @paraga on scaling Twitter 2025: Investing in his $100M Series A for @p0 Ever since we invested in his seed 2 years ago, the @firstround team has had the privilege of partnering closely with Parag as he’s built a stacked team that’s created world-class web infrastructure specifically designed for AI agents. With a bet that LLMs wouldn’t make the web obsolete and that agents would instead use it way more than humans ever did, the Parallel team started with blank slate thinking about how AIs would search the web in an inherently different way. They’ve since created the only web APIs that are a native interface and tool for agents. And they’re all world-class in quality, latency, and cost. That’s due to the fact that they built their entire stack with crawling, indexing, and retrieval infrastructure that centers on AI as the primary user. Companies on the frontier of building with AI (like Clay, Sourcegraph, Owner, Starbridge, and Actively), are using Parallel’s APIs to power a wide range of use cases, from GTM workflows to helping lawyers find precedent and insurers underwrite claims. Today’s $100M Series A (led by @IndexVentures and @kleinerperkins) gives Parallel plenty of fuel for their mission to make agents more useful, today.

A microphone that records meetings and dumps the transcripts into Slack. Dropshipping for ecommerce merchants. A team collaboration tool. These were all ideas @christinacaci explored before landing on SOC 2 and @TrustVanta. As we were developing @firstround's PMF Method, we learned so much from her journey — grateful she took the time to share lessons for other builders. 🙏 Here were a few of my top takeaways: 🔍 Go problem first vs. technology first. "Vanta started not as Vanta but as a startup in search of a problem. There was a dark couple month period of wandering in the desert building a bunch of software that no one wanted. We eventually decided to stop building and go find a person with a problem and ask them a bunch of questions, and then just rinse and repeat. The fuzzy metric was to keep doing this until you had a conversation and you sort of already knew 80-90% of what the person said back to you, where very little of it was new.” 📈 Market > everything. "The market wins, the founder doesn't win. If you choose a market that doesn't exist it sort of doesn't matter what happens afterward. But if you did a market sizing exercise prior to Vanta, there were maybe 1000 SOC 2s done in the world, for call it $100K. There was a very real period where we thought it could be a bootstrapped small business. But when you talked to a company that just got one they would tell you how it just transformed their sales process, like it was a magical document. So a big part of the thesis was lowering the cost of getting these reports for a smaller company and then more smaller companies would get them.” 💵 You can't shy away from founder-led sales. “Prior to Vanta the last thing I sold was Girl Scout cookies. But I think founders have to do that early sales because you're putting yourself in front of potential customers saying, ‘I built this thing that is supposed to do X, Y, and Z for you. Do you like it? Does it work for you?’ Those conversations are so helpful to feed into what you're building and how you explain it to the next person. You also want to learn the sales skills over time, but you can actually get away with a lot by just being curious.” So much more of Christina's and other founders' tactical advice for every stage of product-market fit is in our new essay (link in the next post!)

Question: how do you successfully sell a product that provides extreme customer value...BUT that value isn't obvious to the customer until after the product is used...AND there’s upfront activation energy required? Open to any and all ideas/examples…thanks!

Rocketship is an overused term. But it’s hard to come up with another way to describe @FAL’s trajectory — going from $1M to $40M ARR in one year is more vertical line than hockey stick. Their crazy focus on performance and reliability have meant that they’ve built what developers and enterprises actually need to deploy AI-driven media creation at scale, handling over 100 million inference requests daily with 99.99% uptime. That's why companies like Quora, Canva, and Amazon Ads rely on their platform for image generation — and why @burkaygur and @gorkem are perfectly positioned for the next wave of generative video. Their recent AI video starter kit (an open-source, in-browser editor for AI video, linked below) was the first of many cool launches they have up their sleeve this year. Proud @firstround had the chance to back this team from the seed. This is a team that saw where the puck was going and furiously built ahead of the curve. This time next year, even these numbers will look small. 😉

There’s no single right way to find PMF. But I’ve met several successful founders with an approach I find interesting. They start by problem hunting — looking for things people hate doing, or tasks that are tedious and time-consuming...

Services as Software Seeing some great founders tackle this theme lately across different verticals: SEO, recruiting, growth marketing, trust & safety, legal, etc. Things that you'd traditionally hire an agency, BPO, or consultants to do for you. But well-designed AI systems can compete at a fraction of the cost, work 24x7, and produce results that are often just as good or better. Trend among these startups towards "full auto-pilot" to solve a customer's problem end to end, vs. the "co-pilot" model embraced by incumbents who already have the product/customers/seats (and don't want to cannibalize those seats). Very interested in meeting more founders taking this view.

👇My framework for what makes a great product manager. I think this works for about 70% of companies but may look a bit different depending on your industry, business model and size. Is there anything you would add or remove?

Over the past year, I’ve interviewed dozens of founders on one of the most important, yet elusive, parts of company building: finding product-market fit. I’m excited to share these fascinating stories in a new series launching on the @firstround Review today— details below ⬇️

Before founding Looker, @lloydtabb was CTO at several companies. At each one, he built internal data tools to help people understand what was happening in the business. After the third time, he realized that if you’re building the same thing over and over, that’s a signal that there’s a startup idea in there. While putting together @firstround's PMF Method, we spent tons of time digging into Looker’s data from over the years, in addition to asking Lloyd to reflect on their journey. Grateful that he’s always generously shared what he’s learned with other founders. 🙏 Here are a few of my favorite insights: 🚙 Experiment to find the right value vehicle. While Lloyd was sure about what value he wanted to deliver, he wasn't certain of the right vehicle — so his early experimentation was centered on figuring that out. “For our first four customers, I had different ways of delivering what I was building, from pure consulting to a standalone product. But the problem with just giving someone the software was that they didn't get as much value out of it. I realized I needed to be teaching how to do data as well as delivering a product.” What ended up working best was part product, part service. “We used the demo as a chance to build a proof of concept, so we didn’t have a dummy sales pitch version — we always asked the prospect for an actual dataset to play with. Then it was a very quick forward-deploy. We would come in and do a free trial where we would set up the software and teach them how to use it. And then we would watch for engagement. Only when there was engagement would we close the deal. We had almost no early churn because we only sold customers who got the value out of it. We had a better than 75% close rate on trials.” 🎓 If you’re in the SaaS business, you’re in the education business. “The thing that I realized very early was that Looker was an education company. We knew that the value was getting the customer to understand how to build tooling so that they can understand their data. We built chat into our product so that when they had a question, they could get to us immediately and get answers to their questions so that they could provide answers to their organization's questions. In SaaS, focusing on making your customer successful is a retention strategy, not a cost center. Complicated products require education. If you're selling a complicated product, you are in the education business — so the support you provide is everything.” 🏆 At-bats are precious — maximize learning from each one. “One of the things about early product-market fit that's really important is to look at your at bats. When somebody shows up and expresses an interest in the product that you’re offering, make sure that you do everything that you can to maintain close contact so that you can learn from them. If you're not driving yourself crazy thinking about how you can capture it, you're wasting opportunity.”

Excited for what’s next – joining @Dropbox to lead product and work alongside an incredible team!

Most builders don’t become investors. I’m looking for someone who breaks that pattern and wants to join me @firstround to support the next generation of brilliant founders with a unique insight that’s highly technical. I spent my career working on products like Gmail’s user interface, Facebook’s Newsfeed, Twitter’s timeline and Dropbox’s bottom-up revenue engine — all products that are extremely technical under the hood but simple and powerful to the user. I also founded a company back in 2013. I always loved working with investors who thought like builders — because they were builders. Many VCs pattern match on metrics, TAM, and “heat” in a sector. Few can spot technical breakthroughs before they're obvious. Even fewer can help founders navigate the journey from technical insight to compelling product. In this role, that would be your day job. I’m looking for someone who’s: - Been deep in the technical weeds — shipping code to production, scaling systems, building devtools or open source projects. - Naturally drawn to the cutting edge — reading research papers, testing new programming tools, and forming strong opinions on where technology is headed. - Thinking about business models as much as technical architecture — studying great infra and security companies, and caring equally about pricing models as the developer experience. - Already embedded in the startup ecosystem — going to AI hackathons on weekends, joining dev communities, or being the person founder friends turn to for technical advice. So if you’re merely interested in startups, this role isn’t for you. If you love the idea of investing but haven’t actually built anything, it’s also not for you. This opportunity is perfect for someone who wants to go seriously deep on learning the craft of early-stage investing and becoming an essential partner to pre-product-market fit founders. To be clear, this is more of an apprenticeship, not a shortcut to leading your own deals. You’ll work closely with me to develop a strong technical investing perspective, learning how to spot and support pre-PMF founders. This role is based in the Bay Area. If you’re interested, apply below. 👇

📣 I'm looking for a new Chief of Staff to work alongside me at @firstround! You can learn more about what I'm looking for in this critical role and how to apply here: linkedin.com/posts/toddj0_we…

Big congrats to @TaraViswanathan, @RosaHamalainen and the entire @Rupa_Health team for announcing their planned merger with @FullscriptHQ today. I could talk about Rupa’s impressive growth, from the number of practices and lab order volume to the 100s of thousands of patients served. But it’s impossible to see Rupa’s trajectory and not be blown away by one singular thing: how Tara operates as a founder. I first met Tara back in 2018 when I angel invested — in a very different idea. The big vision was the same: to bring root cause medicine to everyone on the planet (aka address why you’re feeling sick, rather than simply treating symptoms). But the shape of the business was quite different. From a “Zocdoc-style marketplace” for holistic health providers to a virtual practice platform, it took a few iterations before ordering lab work emerged as the “picks and shovels” wedge in a rapidly growing space. Tara was relentlessly focused on finding the right problem rather than getting attached to her solution. That fire has carried her through ever since. After joining @firstround, I led Rupa’s seed in 2020. From the chess-playing strategy in board meetings, to the more in-the-weeds product jam sessions, I’ve seen how she brings a crazy high rate of learning and depth of thinking to every challenge up close. And if you know her IRL (or just follow her on X 😆), you’ll know that her off-the-charts energy and passion is infectious. I did many calls to help close Rupa candidates over the years, and hyping Tara up was the easiest gig in the world (and even had one of my best chiefs of staff join the team to work for and learn from Tara cc @LindsayGo1dman!). So proud of how she channeled all of that into building a platform that’s beloved by doctors, patients, and lab providers. Backing whatever she does next will be the fastest first check in I’ve ever written.

Big full circle moment today. If you told me back when I was at Dropbox 10 years ago that I'd be leading my then-colleagues @jgreze’s & Tony Vincent’s $18M seed round, I would have been floored that my future self became an investor (not a VP Product or a founder again) — and I definitely would have been shocked by how big seed rounds had gotten. But I wouldn’t have been surprised that I'd jump at the chance to back Jean-Denis and Tony. Even back then, it was clear they were destined for big things. JDG went on to become Plaid's CTO, while Tony sold his next company to Google and became a big part of their AI initiatives (on the main stage for the I/O keynote last year). Today, they're launching @TownTaxCo — something special they've been quietly building to help SMBs tackle taxes. Small businesses are getting crushed here. They don’t have in-house teams working year-round to minimize what they owe, so they have to go it alone or rely on overwhelmed CPAs who drop balls, cause last minute scrambles, or even straight up ghost them mid-season. Automation has been hard to get right previously. It’s the kind of messy market most startups avoid. With a dusty industry and a hard to distribute product, it takes a team like this to get it right. JDG is amazing technically, but he’s an atypical CTO who’s also really strong commercially — he joined us for First Round’s PMF Method last cohort, and it was cool to see how advanced his thinking already was around distribution and partnerships. And Tony is an incredibly strong product founder with a design background and a great sense of craft — he just makes killer products. They’ve been super thoughtful about how to rebuild tax technology from the ground up with AI. With Town, your (human) tax advisor works year-round on strategy while their AI handles all the tedious compliance work and remembers every detail about your business, year after year. Only a few months in, they’ve already got impressive traction to show. There are 100+ businesses on board (from e-commerce shops and dental practices to startups) and they’re partnering with bookkeeping and fintech platforms to embed tax support right where owners already do their day-to-day work. Very honored that they chose to work with @firstround, and very excited for Jean-Denis, Tony, and the whole Town team — this is just the beginning of something special.

The founder stories I heard during interviews for @lennysan's newsletter have stuck in my mind in every founder meeting I've had since. The #1 signal of PMF that I look for now? Ideas that have been validated with real people. Here are gems I wish every founder could hear 👇

I don’t know how to comprehend these highlight reels of police brutality (in response to protests of police brutality) and the President sitting back and fanning the flames...

Reaching the scale to host your own conference is a massive milestone for a company. Since @fal’s inaugural Generative Media Conference is this Friday, I sat down with @gorkem to dig into the last 3 years of their journey. Their 10/24 conference has an insane speaker lineup with lots of real-world use cases of generative media in action. The preview video below is just one example (we used fal’s platform to create all the visuals). Aside from the fun we had turning me into George Clooney and making a giant 3-ton Roomba cat destroy our podcast studio, what stood out was how far all of these models have come in the last few years -- and how easy it is to get started with fal and play around so many different options. There’s a ton of interesting technical optimizations and developer-focused decisions that made that experience possible, so it was great to hear Gorkem unpack all of their unique company and product building choices. One of the more interesting ones that they hardly ever talk about is their pivot (when I invested back in 2022, it was a completely different product for data teams), so it was fun to get into that story as well. Full conversation below, hope you enjoy it!

Most intimidating app ever for introverts 😳

The past few days have been extremely challenging for everyone in LA. I'm deeply saddened for so many families in our community — and so grateful for the teams of firefighters who have been working heroically around the clock. We were able to visit our house on foot this morning, and see that our immediate neighborhood (Santa Monica Canyon) is ok and our home is still standing. But just blocks away there is devastation — families who lost their homes, their kids' schools, their entire neighborhoods. My heart goes out to them, and all of LA will be rallying to help them build back up over the coming months. ❤️‍🩹 Hoping for a contained fire and low winds over the next few days. 🤞

The flames in LA are largely out — thanks to our incredible firefighters. Now everyone here is learning about what comes next. For homes that survived, it’s masks, air filters, house inspections, industrial hygienists helping remediate. For homes that didn’t, I’ve heard stories from community members who survived previous fires — first comes a search for remains, then families sift for belongings accompanied by fire personnel, then they rake away the rubble for toxic disposal. Houses that are still standing are inspected, some will be deemed unsafe and condemned — even if they stood through the fire. Torn down to the studs. Probably 2 years (or more?) before the first people move back into reconstructed homes. What do people do during that time? Rents are already 2X what they were 2 weeks ago, and inventory is gone before it shows. Offers being made on lots at 10% of what they cost 2 weeks ago. Some families moving 20 miles away to Manhattan Beach, 120 miles to Palm Springs, or further. Looking for new schools and normal housing prices. I hope they come back. I’ve never lived through anything like this before, and I’m feeling extremely lucky that our neighborhood on the border of the Palisades and our kids’ school in Santa Monica made it through relatively unscathed. I grew up in LA before moving to the Bay Area for many years — and when we had kids, we moved back so that they could grow up near their grandparents. We’re committed to LA and will be part of this rebuild over the next several years. There are a lot of wild takes and negative stories on this site about what's going on. Here's my POV: it’s already bringing out the worst in a very small number of people, but we're seeing so much more of the best. Volunteers are showing up in huge numbers, organizations are overwhelmed with donations, and there's a real sense of people pulling together to help their neighbors. This is the real LA — and it's why we're lucky to call this place home.

Running out of new ways to describe just how outlier @FAL’s growth has been these past few years. The numbers paint a good picture: 2M+ plus developers. 300+ enterprise customers. Billions of AI-generated assets monthly. 60x revenue in the last year, just crossing $95M. $1.5B valuation. Raised A, B & C rounds in 18 months. But for me the real story is how @burkaygur and @gorkem have consistently been far ahead of the wave. Back when they first started, generative AI was just getting going, and was often clunky and slow. As they were trying to optimize Stable Diffusion, they saw this new category of generative media coming, and went all in. (We’ve been all in too here @firstround, investing in every round since their 2022 seed.) Generating media is a brand-new behavior, but there’s so many use cases in tons of different verticals. It’s the early innings still, but we’re already seeing super interesting applications of this technology’s future start to emerge, like real-time advertising campaigns, or even Netflix starting to ship AI-generated content in its shows. The TL;DR here is that real-time, personalized media is becoming less and less experimental, and will only become more core to how massive brands sell, engage, and build. fal is the infrastructure layer powering all that work, with inference solutions for AI-powered video, image, and audio applications used by customers like Adobe, Canva, Perplexity & Shopify. It’s a big boost with Meritech, Salesforce Ventures & Shopify Ventures coming aboard for the $125M Series C. As the technology continues to race forward, fal’s well-prepared for how fast the space is moving. 🔜 More support, more go-to-market, and a larger GPU fleet. Congrats to the whole team on all the hard work that went into this milestone, and all the ones that are still to come.

After almost 3 wonderful years @Dropbox, excited to announce my new role as Founder in Residence @FirstRound! (1/2) medium.com/@firstround/welco…

Launching Dropbox Professional! w/ Smart Sync and a new feature called Dropbox Showcase. Let us know what you think blogs.dropbox.com/dropbox/20…

Dropbox Design just DOMINATING the Dribbble popular page right now. dribbble.com/shots

It’s such a treat to be on @lennysan’s podcast today. We dig into Product-Market Fit Method, @firstround’s new framework and program to help B2B founders increase their odds of finding product-market fit. We cover a lot of ground, including: 🪜 The 4-levels of our PMF framework (nascent, developing, strong, extreme) 🧭 The goals and challenges at each stage 📖 Stories from companies at each level 🚩 Flags that you’re stuck 🛠️ Tools to get unstuck (persona, problem, promise, and product) It’s also a full circle moment — we’ve collaborated on First Round Review articles back in the early days of Lenny's newsletter (first hundred subscribers), and he’s an alum of our Angel Track program, so it was cool to experience the other side now as a guest on his show. One of the most fun podcast interviews I’ve done, Lenny is an incredibly warm host who asks great questions — the whole experience was 💯 Give it a listen and let me know what you think!

Super excited to kick off PMF Method this weekend with a group of exceptional, accomplished and kind founders 🎉 The reason @firstround decided to run this program is pretty simple — we strongly believe that Product Market Fit is the single most important thing for founders to think about in the first couple years of their company journey. Other than choosing your cofounder and building the early team, it’s really the only thing that matters in these early days of building a company. And after spending time as both an investor and a founder, I’ve realized there’s a huge problem with how we talk about PMF. We often treat it as a binary concept (you either have it or you don’t), when the reality is that it’s a progression. We’ve worked with hundreds of founders over the years to see if PMF is something that can be reverse engineered, and came up with our own framework and method for increasing the odds you’ll find it.  Super excited to work with this amazing group of founders! You can read all about Session 1 here: pmf.firstround.com/levels. More to come!

Seeing lots of design-focused founders who would have done a consumer startup 5 years ago now focused on SaaS. Future of work is ☀️

After @tobi’s well-known AI memo, what really changed @Shopify? @firstround just launched Applied Intelligence, a new publication about how companies are actually using AI in production. Our first story features none other than Shopify. This is the tactical sequel to the memo, where @fnthawar shares how Shopify created a cultural infrastructure to adopt and scale AI, along with their most impactful workflows. Read the first essay below, and stay tuned for deep dives from Carta, Atlassian, Linktree + more.

AI legal startups are a thing in 2024. But as @ironclad_inc’s @jboehmig puts it, “nobody was trying to buy an AI legal assistant back in 2015.” Ironclad has one of the most interesting and underexplored stories out there IMO. As we were developing @firstround's PMF Method, we learned so much from their journey — super grateful Jason took the time to share his insights for other builders. 🙏 Here were a few of my takeaways: 🔬 Zoom in to find focus “It was actually fairly easy to sign up early customers — what was difficult was finding a product that could address that market. That took us several years of iteration. We had to try to figure out what pieces we could peel off into a repeatable, discrete software product. We quickly realized the really interesting part of the problem was in repeatable business transactions — sales agreements, employment agreements, NDAs licensing agreements, partnership agreements. A lot of our competitors tried to do everything that corporate legal teams do. But we were only doing the contract part.” 📣 Expand an existing category (with an existing buying cycle) But the initial AI legal assistant positioning wasn’t resonating. I’ve talked to 100s of founders about PMF and the story of how Ironclad got unstuck is one of the wildest ones I’ve heard. “100% of the time I had to explain what an AI legal assistant was. We had a hello@ironclad.ai email on our site. One day I got a one-liner message that said, ‘Hello, are you a CLM?’ I was so close to hitting archive, but it was from someone at a publicly traded company. But what was a CLM? Turns out it’s a Contract Lifecycle Management platform that helps enterprise companies create and manage their legal contracts. By that definition, we were. So of course I wrote back, ‘Yes, we are definitely a CLM, we would love to come demo our CLM for you.’ But while we were really great at creating contracts with our AI legal assistant product, we hadn't put a lot of thought into how you deal with contracts afterwards, with a feature called a repository. And so we had set up this demo with the legal team from this publicly traded company, and I turned to my co-founder Cai, and said, ‘By the way, we have 3 hours to build a repository.’ We took the train from SF to San Jose and he built the first version of a repository, which we demoed live at the end of the train ride. This customer was in a CLM evaluation cycle that had 12 other solutions in it, but they loved the demo. So we went and actually built the full product, and we won. And after that, of course, we changed our messaging. We got serious about building CLM functionality and that's our flagship product to this day. There were lots of people out there trying to buy a CLM so we just got to participate in a lot of buying cycles, but with the AI legal assistant buying cycle, we had to create every one of those.” 👥 Artificially constrain the buyer and build community early “One of the things that we did which was really helpful in hindsight was we artificially constrained the buyer we were going after. Once we decided to make the shift to enterprise, instead of trying to address the whole US market or the whole global market, we decided we only cared about being the number one CLM in SoMa. We got a list of every company that could use the CLM in SoMa and got intros to them — it just provided a ton of focus for us. It's how we also stumbled into doing community. We would host these community dinners and if you were a general counsel of a company based in SoMa, you probably knew other people that were coming to them. We just started to get this buzz of ‘Are you going to the Ironclad thing tonight?’ There's a ton of value if you can discover a part of the organization that no one cares about, and connect that part of the organization to a larger business problem.” 📚 Founder-led sales is learnable I was impressed to learn that Jason still sends cold outreach himself to this day. But founder-led sales didn’t come naturally. “A misconception I had about early-stage startups was that the cartoon character salesperson who's slapping everyone on the back and is a total extrovert is the best salesperson. And it's actually the person who's almost like an engineer in their mindset — super methodical, sends great follow ups, could be very shy. It's a very learnable skill.”

B2B sales is hitting a ceiling with the “more is more” approach (more data, more emails, more calls). But the best sellers have always known that quality beats quantity. I’ve worked with the @actively_ai team super closely since leading their seed in early 2022, and I can tell you firsthand that @mgarimella and @agupta_108 are intensely focused on substance and real results, not volume and hype. (For example, they never even announced their seed. 😆) Unlike the many AI sales assistants or copilots that layer basic AI or signals on top of a single rep’s knowledge and deliver shallow personalization and relevance, Actively trains a custom reasoning model for each company. That means replicating the strategic depth of your best seller, not running a “spray and pray” play with an (at best mid) AI SDR. This results in AI that doesn’t just automate or assist, but reasons to make the best possible decisions. It picks up on the nuances of your business and synthesizes millions of data points about every prospect — from your custom Salesforce fields to minute 27 of a demo call where someone mentioned renewal dates — to determine who to target, when to reach out, and why. What's been particularly impressive is how they've built for and alongside elite revenue teams. They've worked hand-in-hand with organizations like Ramp, Ironclad, Navan, Dialpad, and Verkada, not just collecting logos, but creating a product that revenue teams truly need to drive hundreds of millions in added revenue. Congrats to the team on their Series A milestone today! It’s well-deserved, and there’s much more to come. More details below.

Back in 2008, @amasad wasn't trying to launch a startup. He was a college student in Jordan with a wild idea: What if you could code collaboratively in a browser? "I was always attracted to the entrepreneurship track, but I've always ultimately been guided by my worldview when it came to creating technology," Amjad told me. So he built @Replit — the first collaborative coding environment in a browser. Then he abandoned it to pursue a career as a software engineer, working at Codecademy and Facebook. But here's what's wild: Even though Amjad wasn’t actively working on it anymore, Replit took on a life of its own. It kept growing, mainly among student developers, racking up so many early product-market fit signals that soon, it became impossible for Amjad to ignore. Years later, Amjad circled back to his college project. But turning it into a company wasn't easy: • YC rejected him 3 times before finally accepting Replit • He struggled to raise initial funding • VCs kept saying Replit wasn't a sustainable business But Amjad's conviction never wavered. He made bold, unconventional bets like: • Supporting multiple programming languages (like Python) instead of betting on just JavaScript (this was back in 2017, as everyone else was doubling down on JS) • Pivoting from browser-based to cloud-based execution, even though browser had been his initial breakthrough • Investing heavily in AI before it was mainstream These decisions paid off big time. Replit's growth has been explosive: 2018: 1M users 2021: 6M users 2023: 20M users 2024: 30M users Now, Replit has raised over $200M and just launched Replit Agent to help users create and deploy full applications from a few sentences. No code required. Check out my full interview with Amjad in the latest installment of our Paths to Product-Market Fit series for more of the details on how he navigated each of these moves, linked in the comments.

Last day in the @twoffice. Gonna miss this building and all the people in it. Also, obligatory laptop+badge photo.

I find there are two types of people. A) Respond critically in areas they know Respond skeptically in areas they don’t know B) Respond patiently in areas they know Respond curiously in areas they don’t know I prefer B.

Craziest beta I have ever opted in to!! 😱😨

Pretty wild seeing @gorkem & @burkaygur on stage today at @fal's first Generative Media Conference. From trying to convince people this category would be a thing in 2023, to so many builders and creatives showing up just 2 years later. Amazing to see how far they've come.

A lot of founders ask me: “How narrow should I go to find product-market fit?” Product success can happen both ways — narrow and wide. As @amasad from @replit shared on our In Depth podcast, they made a key bet to support multiple languages, not just JavaScript, which he called an “agonizing decision” at the time. Episode here 👇 review.firstround.com/podcas…

We recently published the 10th article in our Paths to PMF series! I've learned so much already, and it's reaffirmed how incredibly challenging (but rewarding) the journey to finding product-market fit is. Here are my top 10 takeaways from my conversations with founders:

Just left a meeting with Dylan and Jake*. A total waste of time!! I asked what is going to happen this morning, is Dylan going to BRUSH HIS TEETH? Dylan said, NO. I said bye-bye, nothing else works!! SAD! (*my two young children)

When they see a consistent pain point often enough, they start by *manually* doing the work for those people. Through this, they learn the hard parts of doing the task and figure out whether it’s possible to “productize” the work.

Very cool product for WFH -- Remotion. remotion.com/ Just launched today on Product Hunt. Alex(@embirico) was a superstar PM at Dropbox. Excited to see what people think! producthunt.com/posts/remoti…

To celebrate @LaminiAI’s $25M Series A news today, I revisited @realSharonZhou and @GregoryDiamos’ original seed pitch deck from August 2022. Here’s a few observations on what I think they got right that might be helpful to other founders starting out on a similar journey: 📚 Expertise is great — being a learning machine is better: What stood out perhaps most from their original pitch was the team slide. While many founders these days are newcomers to generative AI and LLMs, that can’t be said for this team. They are quite literally world-class experts in AI and high-performance machine learning systems Sharon earned her Ph.D. in generative AI @Stanford under ML pioneer @AndrewYNg. She’s also taught AI in @DeepLearningAI short courses and to 250k students on @Coursera. Greg conducted his PhD research and spent his entire career working at top ML companies (Nvidia, Baidu, Landing) on optimizing the training of LLMs. He has 14,000 citations, including AI scaling laws and Tensor Cores. Everything about the team slide gets that message across — while also giving you a sense of their personalities and what's important to them. But credentials aren’t everything. What was perhaps more impressive was that they both referenced as steep slope, coachable “learn it alls.” After working with them closely these past 2 years, I’m a strong +1 here and would give the same reference to anyone who asked. ⏱️ Set an aggressive timeline — and do what you said you’d do: When you’re a founder pitching investors to raise your seed round, you’re painting a picture of possibility. Like every startup, many parts of Lamini’s initial idea changed, but looking back now, I’m really struck by the original timeline slide they shared — for one, it was bang on for the May 2024 Series A raise timing. 😆 "Hire a lean team of 5 engineers" and "Close 5th design partner" actually came true in roughly those time buckets as well. They really nailed their original plan, even if many of the details shifted, which in my experience as an investor is pretty rare — and indicative of how ambitious and focused this team is. 💰 Don’t shy away from the commercial side — even if you’re very technical and it’s super early: “Talked to 100 customers” and "Close 5th design partner" stood out to me from the timeline in the original pitch as well. Very few technical founders are willing to do the more commercially-minded work of conducting 100 customer interviews and managing design partners so early on in the company building journey. It was clear that Sharon and Greg prioritized this from the jump, and still do today as Lamini expands its customer base and partnerships. Congrats to the entire team on all the hard work that made all this possible — super excited for what the next few years have in store. 🦙 Link in the next post to read more about their progress.

Excited to announce Dropbox's integration into iOS 11! blogs.dropbox.com/dropbox/20…

Exciting day for @crossmint! Proud to be supporters of @rodrifernandezt and @alfongj for the last 3 years. They were quick to recognize that blockchain adoption required eliminating complexity, and they’re also early to making things accessible for an entirely new class of users: AI agents. As AI will increasingly start to manage our transactions (booking travel, buying groceries etc.), these agents will need new systems that don’t require them to circumvent captchas, two-factor prompts and manual approvals designed for human users. Their goal is to help move every business (and AI agent!) onchain, with an all-in-one platform that supports any use case. Instead of cobbling together 5-10 different vendors and spending months on implementation, companies can use Crossmint's suite of low-code APIs to handle everything from wallets and payments to tokenization. They’re already powering many major Fortune 500s building on crypto rails as well as the largest web3 startups behind the scenes, with a 1,100% increase in subscription revenue over the past year, and 40K+ companies and developers using their platform across 40+ blockchains. Can’t wait for what these next few years will bring!

Feeling privileged to have worked @twitter – thanks @kevinweil, @aroetter, @utkarsh and everyone else who makes this place so special.

If you’re an early-stage founder, finding out what your customers *really* want so you can build the right product is critical. I’m excited to share some of @FirstRound’s PMF Method program with @GaganBiyani on Friday! Gagan has founded 3 companies including Udemy, Sprig, and Maven. We’re hosting this live conversation for product leaders & founders on how to reach PMF. You’ll learn: - How to identify your customers' real problems - How to test demand by selling before building - When to write code vs. build a no-code MVP RSVP to join on June 14th (it’s free): bit.ly/3VBAK69

Learn more about PMF Method here: pmf.firstround.com Apply here by 5/7: pmf.firstround.com/apply FAQs answered here: pmf.firstround.com/#faq And read our framework in full in this essay here: pmf.firstround.com/levels

A SaaS founder recently shared a great framework for identifying painkillers (vs. vitamins): 1. Does the customer have a problem? 2. Are they aware of it? 3. Are they actively looking for a solution? 4. Have they cobbled together a solution? 5. Would they pay to fix the problem?

"The type ahead or chat interaction paradigm of AI maxes out at double digit productivity gains because [humans] are inherently rate limited by how fast they can interact with the system...the AI agent model, where you can run many agents in the background in parallel, actually can deliver multiples in productivity gains." Agents are the future. Finding ways for humans to interact with them (pull, push, ambient, autonomous) is one of the most interesting UX questions in tech right now. @mgarimella and @agupta_108 write really thoughtfully about this 👇

Some tips on finding good PMs :-) firstround.com/review/find-v…

One reason (of many) I love working at @Dropbox -- I've never been at a company where people take the time to write hand-written cards to each other so often. It's such a special, warm company.

When our 3 yr old sees a Q-tip he calls it a “covid test” 😂

If it seems possible, then they build the product and iterate with those early customers until the product is able to do 80-90% of the task for them. Then they launch the product, and no surprise, lots of people want it!

Christina (@christinacaci) is an awesome founder. She uncovered the brilliant insight that became Vanta by interviewing folks about security and listening for their real needs. She had never built a security company before and now runs one of the best. forbes.com/sites/phoebeliu/2…

1/ How do you evaluate products as a seed investor? (part 3) 🧵 We see a lot of pre-product pitches @firstround. Here's how we evaluate when all we have are: 🖌️ Figma mockups 💻 Slides 🧦 An adjacent proof of concept ('Uber for X') 🔮 A ‘dream with me' scenario

If you’re curious about angel investing, consider applying to Angel Track! It’s the program that helped me discover that I wanted to switch from building to investing ☺️

What does good retention look like? lennyrachitsky.com/p/what-is… By @lennysan and @onecaseman — great resource for founders, investors, and employees.

I'll be on Lenny's Podcast this Thursday talking more about Product-Market Fit Method — tune in!🎙️

Today's the last day we're accepting applications for @firstround's PMF Method! (Get yours in here: pmf.firstround.com/apply) We've been blown away by the response, and are capping things off with a brand-new resource for B2B builders on The Review today:  review.firstround.com/20-les…

Part of what makes pre-product-market fit investing rewarding is when you back technical founders with deep expertise, watch them adapt as markets evolve, and see it (eventually) start to compound into something extraordinary. That’s the story of @fal in a nutshell. $1M to $40M ARR in just one year is off-the-charts-impressive. This week’s @FortuneMagazine writeup on their $49M Series B is cool to see. From the outside, it looks like a story of a startup rocketing up through every level of product-market fit in a single year. But what I’m most proud of is how the team navigated the 2022-2024 period in nascent PMF, the heads-down work of figuring out the right idea, laying the technical foundation, and evolving the product. When they started back in 2021 (and when @firstround invested in their seed round in 2022), @burkaygur and @gorkem were focused on building a fast serverless Python environment. Their infrastructure expertise from Coinbase and Amazon coupled with an obsession with performance turned out to be perfectly timed as the generative AI media wave hit. In 2023 they recognized that generative image, video, and audio technologies would evolve into a distinct market segment, separate from LLM services. Last year, they focused on images and audio. This year, video is the next wave. Those bets paid off. They’ve grown to 1M+ developers, and now serve 50+ enterprise customers as well, generating billions of assets (image, video, audio) monthly and handling over 100 million inference requests daily with 99.99% uptime. Bigger scale, same obsession with speed, making things happen, and building ahead of the curve. Proud to have had the chance to work with this team from the very beginning of their story.

Evaluating products as a seed investor (part 4) 🧵 In my @firstround Angel Track session on product, the last thing we covered was evaluating business models — these are often tough to predict accurately at seed, and we try to avoid false precision. Here's what we do instead.

#WorldCup war room ramping up!

1/ I love building + investing in “power tools” — complex under the hood but simple and powerful to the user: Dev tools Data tools Design tools Productivity tools Etc. I like that founders are rethinking all of these with AI. These are some ideas I’m searching for lately:

It’s a great day for @rupa_health as they announce their Series A with @BessemerVP! @TaraViswanathan is a truly special founder who recognized the huge potential in root cause medicine early on and built a platform beloved by doctors and patients. bit.ly/3swNbS9

So proud of all the teams at Dropbox who worked towards today's launch! blogs.dropbox.com/dropbox/20…

Opening up Fall 2025 applications today for @firstround’s PMF Method! If you’re exploring early B2B ideas, it’s the single best way to de-risk the jump. It’s completely free (we take 0% of your company) and only 4 days. And you’d be in good company. Previous PMF Method founders are a seriously impressive bunch. They’ve collectively raised $295M+ (and most of the companies are <18 months old). Almost half are building their second company (many previously started unicorns). The rest are CTOs and senior leaders from companies like Stripe, Plaid, Rippling, Anthropic, Neuralink, Retool & Vanta. Alongside builders like these, you’ll learn what it took legendary companies months (or years) to figure out on their own: - How to run customer discovery that actually validates willingness to pay - Whether your market can support a venture-scale business - How to land your first enterprise contracts - When it’s time to pivot Here are a few key dates and details: - The Fall 2025 PMF Method cohort runs 10/5 - 10/8. - Application deadline is 11:59pm PDT 8/15. - For early B2B SaaS founders, whether you’ve been working on an early idea for the past year, or you’re still at your day job, a few months away from leaving to start your company. - You don’t have to be based in SF, just need to be able to get to Sonoma. - If you’re in, you’ll know by 9/12 at the latest. The app only takes 10 minutes, so don’t overthink it, just apply. (If previous cohorts are any indication, we’ll likely see thousands of founders apply for ~30 spots.) More below 👇

Super excited to launch Twitter Highlights for Android!

New @Dropbox iOS app with 5 new features launching today (including signing PDFs with your finger...so handy!) blogs.dropbox.com/dropbox/20…

My wife just told me I am much better at doing complicated things than simple things. Is this a compliment?

There are many ways companies approach product definition, but we tend to see these 3 paths over and over: 🎨 Creating a new product category 🛠️ Reinventing an existing product category 🔐 Unlocking new (or latent) behavior

I’m so excited to welcome @cjc to @firstround!! Between her deep work company building at Stripe and Notion to her extensive experience angel investing, I know she’ll be an incredible asset to our team and founders. 🙌

Open application for our next partner @firstround — please share! We strongly believe that our next partner can and should be from outside our current network.

What are the traits of a great Chief Product Officer? From what I’ve seen, it’s someone who: ✅ Thinks holistically across product and GTM ✅ Acts as the cross-functional anchor (across execs and execs+1) ✅ Attracts great talent and aligns it with company strategy What else?

East Bay Engineering friends! I'm super excited about Snackpass, a food pickup app that's insanely popular on college campuses. They're backed by YC & First Round. I'm proud to be an advisor. Snackpass is hiring great engineers in Berkeley. Apply here! jobs.lever.co/snackpass/79cd…

We just opened up applications for the Spring 2025 cohort of @firstround’s PMF Method, and if our previous cohorts are any indicator, the caliber of builders in this group will be off the charts. Here’s a small subset of our alumni: - @snirkodesh, former head of eng at Retool and 2nd-time founder exited to Lyft - @realdealpatil and @utsav_sha (both ex-Vanta) working on a stealth security startup - @mboufford (former Greenhouse CTO) & @intrater (former design at Coinbase, Okta) building Otti - @liuxcathy & @roopakv building Caro, headcount mgmt for scaling teams Whether you're pre-product or pre-idea, whether you’re working full-time or planning to leave your job soon(ish), don’t overthink it — just apply. We’re hosting the retreat in Sonoma from March 30 - April 2. Applications close at 11:59pm PDT on 2/7.

Felt super lucky to be part of @firstround’s Angel Track — an amazing group of budding angel investors learning from some of the best. Still keep in touch with so many folks from my cohort. Building deep, lasting community is First Round’s greatest strength as a firm.

I love @reductoai. Such a powerful and well-built product. Enterprises are lining up because it just works. @aditabrm & @raunakdoesdev are the 🐐s! And props to my partner @lizwessel for her tireless support of the company!

“S/he was a superstar at our company, and a total pain in the ass to manage” ➡️ great thing to hear when doing reference calls on a founder

And that’s a wrap on @firstround's PMF Method Summer '24! An incredible retreat to summarize the last 12 weeks, where we shared learnings, answered burning questions, and gave each other feedback 🎉 I'm so inspired by this group of 20 founders, and can’t wait to see what they do next. We had such overwhelming demand for this program, so stay tuned for more PMF Method news coming shortly….

Excited for all the new @Dropbox productivity tools launching today -- congrats to the teams who built them! techcrunch.com/2016/06/22/dr…

Proud of the Dropbox Design team (and their Eng/PM partners) for launching the redesigned dropbox.com today!

When text-to-image models first burst on the scene, it was pretty mind blowing. But as you started using them more, it could also be kinda slow and tedious to get things just the way you wanted. If you’ve had an ultra fast image generation experience recently, chances are it was powered by @FAL. What stood out most to me about @burkaygur and @gorkem when @firstround first invested in their seed back in 2022 was that they are just performance engineering nerds at their core (previously the first ML hire @coinbase and an Amazon infra vet). They obsess over coding optimizations at every level in the stack. (Our original investment memo had phrases like “warning: this is technical,” and “they went viral on data Twitter,” aka my favorite kind of founders)... With their early work to build a fast serverless Python environment, they were well-positioned when LLMs burst on the scene. They’ve always been technically excellent and obsessed with performance — and you can see it reflected in the product. When they first showed me a demo of their hyper-optimized image gen a year ago in our office, my mind was blown — not just by how they turned me into George Clooney, but also by how fast the performance was. They were doing stable diffusion inference continually at 30 frames per second, so fast it basically became a video! (Of course, they’ve somehow made it even faster now.)

...the only hopeful things right now are the images of police officers marching with protesters in New Jersey, Santa Cruz, and Michigan, and black protesters protecting stores from white looters. Everything else is tragic and chaos.

Founder-led sales with Emery Rosansky (PMF Method Session #6 @firstround). Emery is the GTM 🐐!

1/ Continuing with Part 2 of evaluating products as a seed investor. 🧵 It's really important to adapt your approach depending on the stage of the product when you see it. As seed investors at @firstround, the pitches we see generally fall into one of 6 buckets...

🔍 Here's a question I often ask myself when meeting a new company: If I squint, can I see this product eventually creating a category, reinventing a category, or unlocking new or latent behavior?

Excited to share the latest from our Paths to PMF series! I loved chatting with @retool's @dvdhsu. As a true first principles thinker, he shares a ton of thoughtful (and often unconventional) company-building advice for early-stage founders. review.firstround.com/retool…

Wrapping up the first weekend of PMF Method and feeling so excited for these founder journeys! In Session 2 we talked about “dollar-driven discovery” — how to validate that customers will truly need and pay for your product. This means before jumping on customer calls to validate your startup idea, you need a hypothesis. Who’s your buyer? What critical problem will it solve for them? Most founders stay at the surface level, using one liners like “we’re the first [X] to enable [Y]” during customer calls. At @firstround we developed a framework called the 4 P’s to help founders dig deeper: persona, problem, promise and product. Taking time to validate your 4 P’s with customers is how you’ll stress test whether you’re building something that customers truly need. In Session 3 we covered markets, which is a critical (but often underdeveloped) choice in company building. We don’t look at markets using high level TAM slides or McKinsey-ish top down numbers. Instead, we define a market as a collection of buyers who have a specific problem that your product solves, who buy software in a similar fashion, and who expect the product to provide value in a similar way. Given our definition, most generational businesses end up serving many adjacent markets over time. Some characteristics of a great market, starting with an initial white-hot center: •​​ Enough buyers with an urgent problem to drive significant initial revenue, but ideally not enough that it’s saturated or obviously huge (and thus hyper-competitive) •​​ Buyers are unhappy with the status quo…or the problem is very new •​​ The category is entering a period of high velocity software adoption •​​ Buyers have high (or at least medium) sensitivity to product quality •​​ Budget already exists or will be easy to obtain •​​ There’s willingness to pay a price point that can support the right GTM modality We loved digging into the first 6 months of company building with this group over the past few days, covering "How do I tell if this path is going to lead to PMF?" "How do I know if I'm targeting the right buyer?" and "Is this actually a good market to be building in?" Excited to continue the sessions with the founders over the summer! More to come on positioning, customer discovery feedback and founder-led sales.

I knew these founder stories would be interesting but had no idea how many incredible insights and tales of grit and determination we’d uncover over a couple months of conversations. I hope you enjoy reading about their journeys and learn the same valuable lessons I did!

Engineering friends! Vanta, an awesome early-stage security company (YC '18) is hiring. Great team of ex-Dropboxers w/ early product-market fit, growing quickly. vanta.com/jobs - securing the internet, increasing trust in software companies, keeping consumer data safe

tfw you’re listening to a podcast and hear @rahulvohra from @Superhuman quoting your wife 😂 👋@hiiamArielle (via @getshuffleapp)

So much of this lately. Am I the only one experiencing this?

Sometimes getting the right positioning for your startup is more challenging than getting to the product itself. “Often it’s not that you don’t know who the customer is, it’s that you’re not picking — you haven’t committed to one hypothesis over the other. That’s what creates the confusion about what the product is, what the feature set looks like, and what the language that you use in the product even is.” @clayrunhq's @kareemamin is one of the most thoughtful founders I know on this topic.  Before Clay was the GTM power tool that it is today, it was struggling to choose between broadly building a horizontal product and narrowly focusing on outbound sales. “We had a bunch of initial customers, but they were all doing very wildly different things,” says Kareem. “One company actually sent me their code base because they wanted to send all of their data to NetSuite, but they didn’t have time to do it, so they asked us to reverse engineer it. It was in the same general sphere of doing some data transformation in a no-code way, but it had nothing to do with sales or recruiting. We just had this running in the background for a long time.  They were paying us. But suddenly our servers would go down and we were like, ‘What's going on?’ And it was just this one company sending a bunch of data to us that needed to get transformed and sent to NetSuite. We had a customer, but it was the wrong customer.” In the latest installment of our Paths to PMF series on The Review, I got to sit down with Kareem to unpack the seven-year “overnight success” story of Clay, drilling deep into the debate between building a horizontal vs. vertical product, and how to find focus as a founder. (Article below)

great idea you will love this job and find it easy to navigate