Skip to content
Aakash GuptaAakash Gupta

How I Built This: 100M+ AI Startup

If you’ve ever thought you’re a great PM and should build something of your own with AI, today’s episode is for you. Matt, who built a $1B+ AI product company LogRocket, shares the full roadmap of his journey on: - How to build an $100M AI company - How to grow it from 0 to 1,000 customers - Then, how to hyper-scale it We're discussing: Introduction - 00:00 Step 1: Build Projects - 01:23 Step 2: Develop Skills - 04:31 Step 3: Find Opportunities - 09:08 Ad: Miro - 10:09 Product Growth Framework - 11:38 Step 4: Build MVP - 12:51 Step 5: Launch Strategy - 13:51 Step 6: Raise Funding - 16:40 Step 7: Growth Channels - 21:15 Step 8: Build Teams - 23:44 Step 9: Hire Executives - 26:21 Ad: Linear - 30:31 PM Workflow Solution - 31:29 Step 10: Funding Growth - 31:30 Step 11: Second Product - 35:29 Step 12: Expand Marketing - 39:28 Step 13: Product Portfolio - 41:53 Step 14: Develop Partnerships - 45:49 Step 15: Global Domination - 48:07 Founder's Lowest Point - 48:52 Starting Today's Advice - 50:12 Closing - 53:06 💼 Check out our sponsors: Miro: The innovation workspace: your team’s canvas - http://miro.pxf.io/PO4WZX Linear: Plan and build products like the best - https://linear.app/partners/aakash 👀 Where to find Matt LinkedIn: https://www.linkedin.com/in/matthew-arbesfeld-04b5429b/ LogRocket: https://logrocket.com/ 👨‍💻 Where to find Aakash: Twitter: https://www.twitter.com/aakashg0 LinkedIn: https://www.linkedin.com/in/aagupta/ Instagram: https://www.instagram.com/aakashg0/ 🔑 Key Takeaways 1. Start by building lots of projects. Build, build and build, just like Matt did since elementary school. His mobile game hit millions of users while an app he built for introverts failed completely. These hands-on experiences teach you more than any classroom ever could. 2. Develop specialized skills. Find your superpower skill combination. It was Matt's coding plus design expertise that created the perfect foundation for LogRocket. So what unique skills can you combine to solve problems others can't? 3. Deeply understand a problem domain. His time at Meteor showed him front-end development challenges firsthand. Put yourself in environments where you'll experience problems worth solving. Your insider knowledge will help you identify gaps that others miss completely. 4. Build an MVP that tackles a specific challenge. His team created user session replay technology by working nights after their day jobs. So, start with something small but valuable that demonstrates your core insight and solves a specific pain point. 5. Launch strategically with momentum. You can orchestrate your launch like a Hacker News campaign with friends ready to upvote. A coordinated push can generate thousands of day-one signups when executed with precision and timing. 6. Establish sustainable growth channels. LogRocket's technical blog became their acquisition engine with 200 high-quality posts monthly. Find one channel that works and then double down on it hard before diversifying your marketing efforts. 7. Be selective about fundraising. Matt successfully grew to $5K monthly revenue before raising $500K from Matrix Partners for his startup. In competitive markets, funding helps you outpace rivals, but remember each market requires a different capital approach to win. 8. Focus intensely on recruiting. Make hiring your obsession once you have product-market fit. Great early hires attract more great talent naturally, creating a self-reinforcing cycle of exceptional team building. 9. Choose executives that match your culture. Cultural alignment trumps raw skill when hiring executives. Look for leaders who enhance your existing culture rather than those who want to demolish and rebuild with their own imported approach. 10. Expand your product portfolio deliberately. Building second and third products is crucial for continued growth. LogRocket expanded from session replay to AI-powered issue detection, creating a broader suite while serving the same developer and product manager customers. #AI #startup 🧠 About Product Growth: The world's largest podcast focused solely on product + growth, with over 167K listeners. Hosted by Aakash Gupta, who spent 16 years in PM, rising to VP of product, this 2x/ week show covers product and growth topics in depth. 🔔 Subscribe and like the video to support our content! And turn on the bell for notifications.

Matt ArbesfeldguestAakash Guptahost
May 11, 202553mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

15-step blueprint for building LogRocket into a $100M+ startup

  1. The episode lays out a 15-step path from early project-building and skill development to identifying a real problem space and launching an MVP with strong initial distribution.
  2. LogRocket’s early traction came from a Hacker News launch and then a durable SEO/content engine built for front-end developers, emphasizing quality plus high-volume publishing.
  3. Arbesfeld details when and why to raise venture funding—especially in competitive markets—and how capital enables faster hiring, marketing, and staying ahead of well-funded competitors.
  4. Scaling required deliberate recruiting, culture building, and careful executive hiring focused on values fit, plus a preference for developing leaders internally over “fire fast” churn.
  5. Later growth depended on expanding beyond a single product into adjacent offerings (including AI that reviews sessions for issues), diversifying go-to-market channels as search behavior shifts to ChatGPT, and building integrations/partnerships as a moat.

IDEAS WORTH REMEMBERING

5 ideas

Starting a company is usually the compounding result of many prior projects.

Arbesfeld argues founders rarely succeed from a cold start; repeated side projects (including failures like “Cluck”) build the creator muscle and pattern recognition needed for a real company.

Build skills by shipping something you personally care about.

For beginners in coding/design, he recommends choosing a meaningful project and pushing through the painful first iteration; learning accelerates when the outcome matters to you or your friends.

A strong problem choice often comes from living inside a domain’s pain.

LogRocket emerged from time spent in the front-end framework ecosystem (Meteor/React era), where he noticed a gap: lots of backend monitoring existed, but little visibility into real user experience on the frontend.

Distribution can create the spark, but product quality determines survival.

The Hacker News launch worked partly due to coordinated initial momentum (friends upvoting), but he stresses that only a genuinely novel/useful product sustains organic sharing and real signups beyond the first push.

Early revenue traction makes fundraising and recruiting materially easier.

They raised after reaching meaningful MRR (on the order of ~$5k/month), which made the story credible to investors and allowed founder salaries that reduced personal survival stress while staying lean.

WORDS WORTH SAVING

5 quotes

I've never seen someone start a company if they haven't spent a lot of their life just working on projects on their own and trying stuff.

Matt Arbesfeld

It's hard to have 100% batting average.

Matt Arbesfeld

There's just a constant stress of starting a company, 'cause you're just bombarded with problems every single day.

Matt Arbesfeld

When you have something that's working in a go-to-market channel, like you're kind of in a golden era, right? Like, if you think of the, the Roman civili- like, when you're in a golden era, that's when you get all your gains. And when you identify that golden era, you have to just go all in.

Matt Arbesfeld

The hardest part of any A- AI product is the accuracy and quality.

Matt Arbesfeld

Early project reps and skill-building (coding + design)Problem selection via deep domain exposureMVP creation and Hacker News launch mechanicsFundraising triggers: traction, competitive category, venture scaleSEO/content engine strategy for developer audienceTeam scaling, location strategy (Boston vs Valley), exec hiringSecond/third products, AI accuracy challenges, partnerships/integrations

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome