Anton Osika, the CEO of Lovable, just shared something really interesting — their top 4 actual use cases for vibe coding in businesses right now.
And the data behind it is hard to ignore. Lovable is now the second-fastest growing software product among businesses on Ramp as of February 2026. The company has crossed $300M in ARR, is reportedly raising at an $8B+ valuation, and over 100,000 new projects are being built on the platform every single day.
Meanwhile Replit did $240M in revenue in 2025 and is raising at $9B and has been just as big of a vibe coding rocketship. This is not a niche trend anymore.
But what are all these people actually building? Who is threatened, who should panic, and who is being truly disrupted? Let’s dig in, because the answer is more nuanced — and more practical — than most of us assumed.
Use Case #1: Rapid Prototyping Without Waiting on Engineering
This is the killer app. The #1 use case. And it makes total sense.
Think about how software actually gets built in most organizations. A product manager has an idea. They write a PRD. It goes into the backlog. Engineering triages it against 47 other priorities. Six weeks later — maybe — someone picks it up and builds a rough v1.
Now? That same PM opens Lovable and has a working prototype in 20-60 minutes.
Replit CEO Amjad Masad recently made the same point, and added a critical nuance. He shared that a public company CEO told him AI coding has had negligible impact on his engineering teams — whatever time saved generating code is lost debugging, reverting bugs, and running security audits. But the real transformation? It’s been on their product and design teams using Replit. Non-technical teams gained what Masad calls “a fundamentally new super power of being able to make software.”
That’s the key insight here. Vibe coding isn’t making engineers 10x faster (yet). It’s giving product people, designers, and executives an entirely new capability they never had before.
Masad said CEOs are now showing up to meetings with working prototypes saying “Look what I built.” He called product managers “some of the best vibe coders” because they’re trained to break problems into clear steps and communicate precisely — exactly what AI needs to build effectively. Even Google CEO Sundar Pichai and Klarna CEO Sebastian Siemiatkowski have publicly talked about vibe coding their own prototypes. Siemiatkowski put it perfectly: “Rather than disturbing my poor engineers and product people with what is half good ideas and half bad ideas, now I test it myself.”
Uber is using Lovable to cut design concept testing from six weeks down to five days. Zendesk moved from idea to working prototype in three hours instead of six weeks. McKinsey engineers used Lovable to build in a few hours what they’d been waiting four to six months for their internal team to deliver. That’s not a marginal improvement. That’s a category shift in how product teams validate ideas.
Anton Osika’s phrase for this: “Demo, don’t memo.” Instead of writing a 12-page deck arguing for a feature, you just… build it. Show it. Let people click around. Get real feedback before a single engineer touches it.
This is the real unlock. Vibe coding isn’t replacing engineering. It’s replacing the bottleneck before engineering. It’s replacing all those ideas that died in a Google Doc because nobody had the bandwidth to build a prototype.
Use Case #1 Is Bigger Than You Think
Here’s what makes the rapid prototyping use case so interesting from a market perspective: it’s not just replacing one tool. It’s collapsing an entire stack.
Think about what a product team uses today to go from idea to validated prototype. You’ve got Figma for design (now public, trading as FIG on the NYSE, just crossed $1B in annual revenue run rate — but the stock has cratered 80% from its post-IPO highs, in large part due to AI disruption fears). Jira and the broader project management software market ($10B+ and growing at 15%+ CAGR). Miro or FigJam for whiteboarding. Notion or Confluence for documentation. The prototyping software market alone is estimated at $700M-$1B and growing at double digits. The broader “product development tools” stack across design, prototyping, project management, and collaboration easily exceeds $25B.
And what’s the actual workflow? A PM writes a spec in Notion. A designer mocks it up in Figma. That goes through review cycles in Jira. Engineering eventually picks it up, maybe 6-8 weeks later. The whole process of going from “I have an idea” to “someone can click on a working thing” takes weeks to months and costs tens of thousands of dollars across all the people involved.
Vibe coding tools compress that entire workflow into a single step. Describe what you want, get a working prototype. Not a static mockup — a functional app that people can actually use and give feedback on.
This is why Figma’s stock has gotten hammered. The company acknowledged the threat in its S-1 filing, mentioning “AI” over 200 times and warning that its products may not remain competitive. Figma rushed to launch “Figma Make” — an AI prototyping tool — and CEO Dylan Field called AI “the most important technology shift of our lifetimes.” But the market isn’t buying it yet. FIG opened at $85 on IPO day and briefly hit $143. It’s now trading under $30. Investors see the same thing Anton and Amjad see: the prototyping workflow is being fundamentally disrupted.
The legacy product development toolchain isn’t going away overnight. But the prototyping and validation phase — the most time-consuming and bottlenecked part — is being radically disintermediated. That’s a massive opportunity.
Use Case #2: Building Internal Tools That Actually Match Your Process
Every company has them — those internal tools that have been on the eng backlog for 18 months. The ones that would save Operations 10 hours a week but never quite make the priority cut because they’re not customer-facing.
Now non-technical teams are just building them themselves.
Masad shared a perfect example of this at Replit itself: an HR team member named Kelsey was looking to replace their org chart software. Every option on the market sucked in its own way. So she built her own in three days — with all the features she actually wanted. No engineering resources consumed. No vendor evaluation. No procurement cycle.
And this is the really profound shift Osika highlighted: the people who actually need the tools are now the ones building them. The sales ops person who knows exactly what the commission tracker should look like. The customer success manager who knows what the renewal dashboard needs. The nurse at a healthcare organization who suggested a patient journey visualization app — and then it actually got built and incorporated into every invoice.
As Masad put it: non-product teams like HR “can for the first time solve problems where vendors don’t have the exact solutions they’re looking for.” Verizon executives are reportedly ditching slide decks for Replit demos. A media company replaced costly software entirely, reallocating savings to AI builders.
We did this ourselves at SaaStr. We built an internal tool we call “10K” — essentially an AI VP of Marketing. We didn’t want to build it. We tried to buy an off-the-shelf AI marketing agent. But every solution out there just writes mediocre content, and we already have more content than we know what to do with. What we needed was orchestration: deep analysis of 4+ years of our data via Claude Opus, feeding into a Replit app that designs every campaign, every offer, every single marketing action for every day through mid-2026. It pulls real-time data from all our vendor APIs, predicts which sponsors to talk to and when, and auto-generates a daily workflow calendar. Nobody else has built this tool. Nobody will. It’s a niche of n=1. And it does more than any human on our team ever did in this role. That’s the ultimate internal tool use case.
Use Case #3: Turning Slide Decks into Real, Interactive Presentations
This one is more niche but surprisingly popular. Instead of building yet another static PowerPoint, people are vibe coding interactive demos and presentations.
It makes sense when you think about it. If you’re pitching an internal initiative, showing a working interactive prototype is 10x more persuasive than showing mockups on slide 14 of a deck nobody reads past slide 6. And if your sales team can show a custom-built interactive demo tailored to each prospect instead of a generic slide deck? That’s a real competitive advantage.
Use Case #4: Replacing Simple SaaS With Custom-Built Solutions
Here’s where it gets interesting for the SaaS world. And notice: this is #4, not #1.
People aren’t vibe coding their own Salesforce. They’re not rebuilding Workday or replacing Snowflake. They’re replacing the $49/month SaaS tool that does 80% of what they need and 40% of what they don’t.
Simple B2B apps. Simple ones.
The kind of tool where you’ve been paying $200/month for a team license to software that frustrates you every day because it almost does what you want but the workflow is just slightly wrong. Now you can describe exactly what you need and have it built in an afternoon.
This is meaningful for B2B founders to pay attention to. The “simple tool” end of the market is going to get compressed. If your product is essentially a CRUD app with some business logic on top, and you’re charging $30-50/seat/month, your customers now have an alternative that costs a few dollars in vibe-coding credits and does exactly what they want.
What We’ve Built at SaaStr (We’re Pushing the Envelope a Bit)
Full disclosure: we’re not just writing about this. We’re living it. And honestly? We’re pushing the envelope beyond what Anton’s top 4 use cases describe — at least for now.
Over the past year, we’ve vibe coded 12+ production apps on SaaStr.ai on Replit. Not prototypes. Not demos. Real software used at real scale:
- SaaStr AI VC Suite — Almost 1,000,000 uses. 3,000+ VC pitch decks reviewed with detailed AI grades and feedback. Nearly 1,000 VC intros facilitated. 600,000+ startup valuations run. This is our flagship, and it’s the real deal.
- AI Mentor — 100,000+ conversations with founders, drawing on 20 million+ words of SaaStr content to give real, tactical B2B SaaS advice.
- SaaStr.ai Website — A complete AI-powered rebuild of our main platform, serving thousands of daily visitors.
- SaaStr AI Annual & London Event Websites — Full conference sites with integrated AI speaker review that scores 4,000+ speaker applications automatically. That AI speaker review system alone saves us $180K+ a year — we used to pay an agency $200-300K to manually review applications. Now it’s instant.
- Founderscape — A full startup simulator game (47,000 lines of code!) where you go from founding to YC to Series A to IPO with real dilution math, real VCs, real cap tables.
Plus internal tools for social media tracking and various other utilities.
Now here’s the honest part. Each of Amelia (our Chief AI Officer) and I personally spend about 30-60 minutes every single day maintaining these apps. Every. Single. Day. It’s not just that stuff breaks — though it does. It’s that the features have to evolve rapidly. It’s a massive amount of mindshare. This is not “set it and forget it” software.
So are we squarely within Anton’s top 4 use cases? Mostly. The AI VC suite started as rapid prototyping (Use Case #1) and internal tools (#2). The event websites are glorified interactive presentations (#3). Some of the internal tools replaced simple SaaS we were paying for (#4).
But with the AI VC Matchmaking, the AI Mentor, and Founderscape? We’re pushing into territory that’s more complex than “simple B2B apps.” We’re building production software at scale — the kind of stuff that traditional wisdom says you need a real engineering team for.
Is that sustainable long-term without professional developers? I honestly don’t know yet. The daily maintenance burden is real. But the impact has been enormous, and I’d do it all again in a heartbeat.
The point is: the top 4 use cases are where vibe coding delivers clear, reliable value today. Pushing beyond that? You can do it. We’re proof. But go in with your eyes open. It’s not a weekend project — it’s a commitment.
What This Means for B2B
Let’s be honest about what this is and what it isn’t.
What it is: A massive expansion of who can build software and how fast ideas get validated. The TAM for “software creation” just got 100x bigger because 99% of people who had ideas couldn’t code before. SaaStr itself has pushed 10+ apps into production used 1m times. Versus zero in the rest of our history. Soon, that could be almost everyone.
What it isn’t: The end of mission-critical SaaS. Not yet, at least. Complex, mission-critical, deeply integrated software isn’t getting replaced by vibe-coded apps anytime soon. The code quality concerns are real. The maintenance question is real. The security issues are real — there was a recent incident where a vibe-coded app leaked 72,000 images including GPS data and user IDs.
But Masad isn’t pulling punches on where this is headed. He recently predicted there will be fewer public software companies in five years, and that “companies that embrace agents and become platforms for agents will survive and thrive.” He sees vertical CRMs for niche use cases being vibe-coded on Replit, challenging Salesforce in smaller segments. Not replacing Salesforce — but chipping away at the edges where one-size-fits-all SaaS doesn’t quite fit.
But the long tail of simple SaaS? The internal tools market? The prototyping phase of every product org? That’s all getting fundamentally rewired. Very, very quickly.
The Numbers Behind the Vibe Coding Explosion
Let’s just look at the scale of what’s happening here, because the numbers are staggering.
Lovable has crossed $300M in ARR and is reportedly raising at an $8B+ valuation — up from $6.6B just weeks ago when it closed a $330M Series B led by CapitalG and Menlo Ventures. This is a company that was at $1M ARR barely a year ago. $0 to $200M ARR in 12 months. Faster than Cursor (12 months to $100M), Wiz (18 months), and Deel (20 months). And it’s still accelerating. Osika was adding $8-15M in new ARR every month as of last fall, and targeting $1B ARR within 12 months. Over 100,000 projects are built on the platform every single day.
Replit is on a remarkably similar trajectory. It did $240M in revenue in 2025 — up from $2.8M just a year prior — and is reportedly in advanced talks to raise $400M at a $9B valuation, with Georgian leading. Replit’s CEO Amjad Masad is targeting $1B in revenue in 2026. They now have 150,000+ paying customers and have launched mobile apps that let you describe an app idea in plain English and publish it to the App Store.
Cursor (Anysphere) raised $2.3B at a $29.3B valuation. Vercel closed at $9.3B. The combined valuation of the top vibe coding startups has grown from roughly $7-8B in mid-2024 to over $36B.
The demand is not theoretical.
And It’s Not Just Web Apps. Mobile Is Exploding Too.
A16z just published data from Sensor Tower and Wells Fargo that tells a striking story: iOS apps released each month were essentially flat for three straight years — hovering around 0% year-over-year growth from 2022 through late 2024. Then agentic coding tools hit the market. Since then? New iOS app releases are up 60% year-over-year as of December 2025. On a trailing twelve-month basis, the acceleration is just as clear — from 5% YoY growth in early 2025 to 24% by December.
ReveneCat powers mobile subscriptions for 50% of mobile subscription apps and has seen a similar explosion:
That’s not a gradual trend. That’s a step function change that lines up almost perfectly with the release of agentic coding tools. After years of flat-to-declining new app creation, we’re seeing an explosion of new software being built — and not just web prototypes. Full mobile apps, published to the App Store.
Replit just launched the ability to create, publish, and monetize native iOS apps using nothing but natural language prompts. Masad described it as dissolving the last major friction point: the byzantine requirements of provisioning profiles, SDKs, and App Store submission that kept casual builders out of mobile entirely. Now that barrier is gone.
The implications are massive. More software is being created right now than at any point in history. Not by engineers — by everyone.
The fundamental lesson from these use cases is that vibe coding isn’t threatening the core of enterprise software. Not yet, and probably not for a while. What it’s doing is eliminating the gap between having an idea and seeing it work. And that gap — that was always the most expensive, most frustrating part of building software inside any organization.
The people who need the tools are now the ones building them. That’s the shift.
And if you’re a B2B founder selling a simple tool? You’d better make sure your product does a lot more than what someone can vibe code in 20 minutes. Because that bar at least is moving up fast.
