Did AI kill ShipFast?

Why are people asking whether AI killed ShipFast now?

ShipFast has fallen far enough that “Did AI kill ShipFast?” is now a legitimate business question, not just a provocative headline. Marc Lou’s Next.js boilerplate once generated more than $100,000 in a single month and dominated his business. Today, Stripe-verified TrustMRR data puts ShipFast at roughly $2,800 in trailing-30-day revenue, while Marc’s newer products generate multiples of that amount.

The timing makes the story tempting. ShipFast sells prebuilt authentication, payments, databases, email, SEO and other repetitive infrastructure so founders can launch without writing it all themselves. AI products such as Lovable, Replit, Bolt and Cursor increasingly promise exactly the same outcome through natural-language instructions.

But three different effects are tangled together. ShipFast experienced a major decline before the modern vibe-coding boom. It is a one-time-purchase product whose customers can reuse it forever, so some decay was structurally likely. And Marc has increasingly promoted and developed other products instead.

So the useful question is narrower: after separating those effects, did AI materially accelerate ShipFast’s decline and destroy enough of its original value proposition to change what the business is?

Is ShipFast actually making much less money today?

Yes. ShipFast revenue has collapsed from its peak, and the magnitude is too large to dismiss as normal month-to-month volatility. Marc Lou reported $123,400 from ShipFast in May 2024. By October 2024 it was $46,400, by June 2025 $19,100, and by the beginning of 2026 it was down to $17,200. Current Stripe-connected TrustMRR data shows roughly $2,800 over the last 30 days.

From $123,400 to roughly $2,800 is a 97.7% decline. Even using the much less flattering-to-the-thesis December 2025 baseline of $20,300, current sales are down about 86%.

The decline has not been perfectly linear. ShipFast fell to $11,700 in August 2025, then recovered to $16,800 in October, $21,100 in November and $20,300 in December. That rebound matters: it shows the product was still capable of generating meaningful sales well into the AI-builder era. But the subsequent fall into the low thousands is now much harder to describe as simple normalization.

Period	ShipFast revenue	Change from May 2024 peak
May 2024	$123.4K	—
October 2024	$46.4K	-62%
January 2025	$36.2K	-71%
June 2025	$19.1K	-85%
August 2025	$11.7K	-91%
December 2025	$20.3K	-84%
January 2026	$17.2K	-86%
Current trailing 30 days	~$2.8K	-98%

Is ShipFast dead as a business?

No. ShipFast is dramatically smaller, but calling it dead would ignore a substantial installed base and more than $1.26 million of verified lifetime revenue. TrustMRR currently records roughly 8,348 ShipFast customers, while the ShipFast website says about 8,400 makers have used the product.

The business still sells a $199 Starter version, a $249 All-in version and a ShipFast-plus-CodeFast bundle. It still has a functioning product, testimonials, documentation, a community and lifetime updates. Product Hunt continues to list 52 reviews with a 5.0 average rating.

A product doing a few thousand dollars in monthly one-time sales with almost no marginal distribution cost is not economically dead. What has disappeared is something different: ShipFast is no longer the breakout engine that defined Marc Lou’s business.

Did ShipFast start declining before the vibe-coding boom?

Yes, and this is the strongest argument against the simple claim that AI killed ShipFast. Most of ShipFast’s first collapse had already happened before Lovable, Bolt and Replit Agent became meaningful substitutes.

ShipFast generated $123,400 in May 2024, then $80,200 in June, $61,700 in July, $60,200 in August, $55,300 in September and $46,400 in October. That is already a 62% fall from the May peak.

Replit Agent was introduced in September 2024. Bolt.new launched on October 3, 2024. Lovable launched its product in November 2024. The tools most closely associated with turning a prompt directly into a deployable application therefore arrived after much of ShipFast’s initial decline had occurred.

There were already AI coding assistants before then, of course. Cursor, GitHub Copilot and ChatGPT could write substantial amounts of code. But the specific consumer proposition that now threatens ShipFast, “tell us what app you want and we handle the application around it,” was not yet operating at anything close to its current scale.

AI cannot reasonably receive most of the blame for the first leg down.

Was ShipFast’s 2024 peak itself sustainable?

Probably not. ShipFast’s $100K-plus months look more like an extraordinary distribution peak than a stable baseline that was later destroyed.

Marc described how outside exposure pushed ShipFast from roughly $50,000 a month toward $85,000 and then around $135,000 during its breakout period. His May 2024 income report shows ShipFast producing $123,400, roughly 94% of all the revenue he reported that month.

The concentration is telling. ShipFast was experiencing an unusually powerful moment when a viral founder story, a rapidly growing X audience, Product Hunt credibility and a very simple “save days of boring setup” message reinforced one another.

The first decline from $123,000 toward $50,000 therefore did not require a technological disruption. Even an unchanged market could have produced a large fall once the viral launch cycle cooled.

The better AI test begins later: ShipFast stabilized around the $15,000 to $25,000 range for much of 2025, then deteriorated toward the low thousands as AI builders became much more capable of shipping production software.

Which AI products actually threaten ShipFast?

End-to-end AI app builders threaten ShipFast much more than AI code editors do. Treating “AI” as one competitor hides an important difference.

Cursor, Claude Code and similar tools primarily make developers faster inside an existing codebase. ShipFast can actually benefit from them because its prebuilt structure gives the model context. Marc himself has argued that combining Cursor with ShipFast lets him turn ideas into businesses within hours, and the current ShipFast site explicitly markets the codebase as context for Cursor, Claude, Copilot, OpenAI, Windsurf and other models.

Lovable, Bolt and Replit attack a different layer. Their ambition is to remove the need to choose, configure and connect much of the codebase in the first place. Replit Agent can create and deploy an application from natural-language instructions. Lovable describes itself as an AI software engineer that lets users build web apps without technical knowledge. Bolt combines prompting, package installation, backends and deployment inside the browser.

The real competitive line has ShipFast plus AI coding assistants on one side, and AI platforms trying to absorb the entire startup setup layer on the other.

Have AI builders actually replaced the setup work ShipFast sells?

Increasingly, yes. AI builders now automate a striking percentage of the exact 22-plus hours of repetitive work ShipFast originally quantified as its value proposition.

ShipFast highlights payments, authentication, databases, emails, SEO, protected routes and landing-page setup. Its homepage estimates that those jobs collectively create more than 22 hours of work before founders even build their differentiated product.

Replit has progressively absorbed these jobs into Agent. It introduced one-prompt authentication in May 2025, one-click deployment, built-in databases, custom-domain purchasing and then integrated Stripe payments that synchronize subscription status with the database. Its 2025 product recap explicitly describes the platform becoming “Agent-first.”

Lovable now lets a founder describe a Stripe checkout or subscription in chat and automatically creates the payment logic, database structures and UI. Its Supabase integration handles authentication, storage and backend secrets, while its current positioning says it handles infrastructure from hosting and authentication through payments and integrations.

That is genuine substitution, not marketing similarity.

Job ShipFast originally removes	ShipFast	AI builders now
Authentication	Prebuilt Google OAuth / magic links	Replit Auth from a prompt; Lovable + Supabase auth
Database	MongoDB / Supabase setup	Replit built-in DB; Lovable configures Supabase
Payments	Stripe / Lemon Squeezy code	Replit integrated Stripe; Lovable chat-driven Stripe
Deployment	Production-ready Next.js base	Replit and Lovable deploy directly
Domains	Documentation / normal hosting stack	Built-in domain workflows
SEO	SEO tags and blog structure	Replit now includes SEO tooling and generated metadata

How big did the AI-builder alternative become?

Enormous. AI app building moved from a theoretical substitute to a mass-market software category within roughly a year.

Bolt went from zero to $4 million in ARR within four weeks of its October 2024 launch and to around $40 million ARR within five months. StackBlitz said Bolt had reached millions of users, while Anthropic’s Bolt case study explicitly describes the product as automating boilerplate and taking users from idea to production deployment.

Lovable moved even faster at larger scale. It passed $100 million ARR within eight months, doubled to $200 million later that year, crossed $400 million in annualized revenue in early 2026 and reported more than $500 million in annualized revenue by June. Lovable also said more than 50 million projects had been created and that users were generating about one million new projects every week.

Replit provides another independent confirmation. The company said annualized revenue went from $2.8 million to $150 million in less than a year, and its CEO later said the company was tracking toward a billion-dollar annual run rate.

These are not niche ShipFast alternatives. They represent hundreds of millions of dollars being spent on a new way of starting software projects.

Why didn’t ShipFast immediately collapse when Bolt and Lovable appeared?

Because the first generation of vibe-coding tools was much better at creating prototypes than replacing a trusted production stack. ShipFast still had a meaningful practical advantage during that transition.

Even as AI builders exploded in popularity, investors and users repeatedly raised the production gap. TechCrunch reported in 2025 that leading vibe-coding tools were excellent at prototypes but still struggled with the infrastructure required to launch complete applications. That gap encouraged newer competitors to emphasize databases, storage, authentication and payments rather than code generation alone.

ShipFast revenue reflects that transition. It generated $22,300 in April 2025, $19,300 in May, $19,100 in June and $18,200 in July. After falling to $11,700 in August, it recovered and finished the year at $16,800 in October, $21,100 in November and $20,300 in December.

That is not the revenue curve of something instantly rendered useless.

The more convincing disruption came as the AI products moved beyond generating interfaces and began integrating the unglamorous production jobs ShipFast had monetized: auth, payments, databases, deployment, domains, storage and increasingly security.

Can Cursor and Claude actually make ShipFast more valuable?

Yes, for experienced developers ShipFast and AI can be complements rather than substitutes. A stable, opinionated codebase can give an AI coding agent much better context than an empty directory.

ShipFast now explicitly makes this argument. Its homepage says the codebase gives AI editors the naming conventions, file structure and existing patterns they need to build additional features consistently. Marc has separately described using Cursor on top of ShipFast rather than replacing ShipFast with Cursor.

There is sound logic behind that position. AI can generate authentication code, but an experienced founder may still prefer a stack whose payment flow, protected routes, email configuration and database conventions have already been used repeatedly. Instead of asking an agent to make every architectural decision from scratch, the founder asks it to extend a known system.

So boilerplates are not technically pointless.

What changed is what customers are paying for. Before AI, much of ShipFast’s value came from saving the labor of writing boilerplate. With AI, the stronger remaining value is reducing uncertainty: a known structure, tested integrations, conventions and reusable context.

That is a narrower market.

Did ShipFast lose distribution, or did buyers stop wanting the same product?

The evidence points much more toward weaker product demand than weaker founder distribution. Marc Lou’s audience expanded massively while ShipFast revenue moved in the opposite direction.

Marc says his X account crossed 100,000 followers in July 2024 and 200,000 in August 2025. TrustMRR currently shows about 374,000 followers.

ShipFast generated $61,700 in July 2024, when Marc had roughly 100,000 followers. Around the time he passed 200,000 followers, ShipFast was making $11,700. Today, with close to 374,000 followers, ShipFast is doing roughly $2,800 over 30 days.

From the 100,000-follower milestone to now, the audience increased by roughly 274%. ShipFast sales fell by about 95%.

Audience size is obviously not the same as qualified buyer traffic, and Marc now distributes attention across many products. But this divergence is still powerful. ShipFast did not shrink because Marc disappeared from the internet. The founder became dramatically more visible while the boilerplate became dramatically less productive per unit of visibility.

Something changed in what that audience wanted to buy.

Did Marc Lou cannibalize ShipFast himself?

Yes, probably to a meaningful degree. ShipFast now competes not only with AI builders but with Marc Lou’s own products for the same entrepreneur audience.

CodeFast is the clearest example. It teaches beginners how to build online businesses, explicitly encourages students to “use AI to code for you,” and walks them through authentication, databases, subscription payments and deployment. Those are adjacent to the jobs ShipFast solves.

The pricing is even more revealing. TrustMRR currently lists CodeFast at $299 and the CodeFast-plus-ShipFast bundle at the same $299 price. ShipFast itself sells for $199 or $249 depending on the edition.

In other words, ShipFast can now appear as an effectively free addition to a CodeFast purchase.

That does not prove that every bundle sale would otherwise have been a standalone ShipFast sale, and public revenue reporting does not let us cleanly allocate bundle economics. But strategically, the positioning has changed. ShipFast used to be the transformation. Now it can be one component inside a broader “learn to build with AI and launch a business” product.

That weakens the case for attributing the full revenue decline to external AI competition.

Is ShipFast’s one-time pricing model also working against it?

Yes. ShipFast’s one-time lifetime model naturally produces revenue decay unless the flow of new buyers keeps expanding.

TrustMRR reports no recurring ShipFast subscriptions. The product is sold once, includes lifetime updates and can be used to build unlimited projects. ShipFast therefore does not monetize an existing customer again when that person launches a second, fifth or twentieth startup.

That was an advantage when ShipFast was spreading rapidly. A $199 or $249 lifetime purchase is easy to understand, and Marc has argued that recurring subscriptions introduce unnecessary purchase friction for this type of product.

But the model also means the installed base is progressively exhausted. More than 8,000 indie makers already own ShipFast. Every satisfied long-term user is someone who can keep benefiting without generating another dollar of ShipFast revenue.

Lovable, Bolt and Replit have almost the opposite economics. Their usage-based or recurring plans capture value whenever users return to build more software.

So a falling ShipFast sales curve does not map perfectly onto falling ShipFast usage. The business can become financially smaller even while thousands of people continue launching products from copies they bought years ago.

Has ShipFast become an add-on rather than Marc Lou’s flagship?

Yes. Marc Lou’s own portfolio has effectively demoted ShipFast from the center of the business to a supporting product.

In May 2024, ShipFast contributed roughly 94% of the revenue Marc reported for the month. By November 2025, Marc himself described his business as becoming balanced across CodeFast, ShipFast, DataFast and TrustMRR instead of depending almost entirely on his original two products.

The transition has since gone much further. Current TrustMRR figures show TrustMRR at roughly $40,000 over 30 days, DataFast around $27,000, Ship or Die around $6,800, CodeFast around $4,600 and ShipFast around $2,800. Among those five largest displayed startup revenue sources, ShipFast represents only about 3.4% of the combined total.

The portfolio math is brutal: whatever word we use for ShipFast itself, the ShipFast era of Marc Lou’s business is over.

Product	Current revenue, roughly 30 days
TrustMRR	$40.0K
DataFast	$27.0K
Ship or Die	$6.8K
CodeFast	$4.6K
ShipFast	$2.8K

Is ShipFast being maintained aggressively enough to fight back?

Not really. ShipFast remains maintained, but it does not currently look like a product in an aggressive feature race against AI builders.

The current ShipFast site says the last update was seven months ago. Marc explains that he continues using ShipFast himself and updates it when his own projects require fixes or improvements. That is a perfectly rational maintenance strategy for a mature boilerplate, but it is very different from the development cadence of the platforms attacking the category.

Replit says it shipped hundreds of features during 2025 alone, including several Agent generations, authentication, one-click deployment, Stripe, domain purchasing, databases, security tooling and mobile-app support. Lovable has similarly expanded from a generator into cloud hosting, payments, security, integrations, domains and production infrastructure.

Feature velocity alone does not make a product better. A small, stable boilerplate may intentionally avoid constant changes.

But the strategic consequence is obvious. ShipFast is defending a fixed and well-understood slice of the development workflow while heavily funded AI platforms keep expanding the amount of that workflow they absorb.

Is the SaaS boilerplate market itself dead?

No, but the generic boilerplate has been heavily commoditized. There is still a reason to buy a high-quality starter kit, especially for developers who want explicit architecture, predictable code ownership and less agent-generated uncertainty.

ShipFast still has thousands of customers. MakerKit continues to maintain and market a substantial Next.js SaaS boilerplate. Current comparison guides still contain ShipFast, MakerKit, Supastarter and several smaller competitors, so the category has not disappeared.

But the competitive floor has collapsed. A new boilerplate no longer competes only against other boilerplates. It competes with free open-source starters, AI IDEs, prompt-to-app products and AI agents that can generate custom setup code on demand.

TrustMRR illustrates how difficult the lower end has become. Fast SaaS, a newer FastAPI boilerplate, shows only $3,657 of verified lifetime revenue. NextBoiler, which even includes an AI page generator, shows virtually no revenue. Another recent Next.js ShipFast-style product reports less than $1,000 lifetime revenue. Those individual examples do not establish a complete market average, but they show that simply packaging auth, payments and database setup is no longer an automatic indie-hacker opportunity.

The category still exists, but its scarcity value is largely gone.

What part of ShipFast did AI actually kill?

AI hit the value of “we already wrote the boring code for you” hardest. A trusted starting architecture still has value.

That distinction explains nearly all the conflicting evidence.

ShipFast launched when saving 20 hours of integration work could feel transformative. Today a founder can tell Replit to add authentication, ask Lovable for three subscription tiers tied to Stripe, or have an AI coding agent generate and debug a large amount of Next.js infrastructure. The marginal cost of producing ordinary application code has fallen enormously.

What AI has not eliminated is the cost of bad decisions, broken integrations, security mistakes or inconsistent architecture. That leaves ShipFast a useful product for a narrower buyer: someone technical enough to care how the codebase is organized, but impatient enough not to rebuild the foundation.

The original market was “developers who want to avoid writing boilerplate.”

The surviving market increasingly looks like “developers who want a known boilerplate as context for their AI.”

That second proposition can still make money. It is simply much smaller and harder to differentiate.

How much of ShipFast’s decline can we reasonably blame on AI?

We would blame AI for a large part of the later decline, but not for the full 98% fall from the peak.

The 2024 evidence prevents that conclusion. ShipFast had already fallen more than 60% before Bolt, Lovable and Replit Agent had meaningful market presence. The peak itself was boosted by unusual founder distribution and publicity, so comparing every later month against $123,400 exaggerates the amount of destruction requiring an external explanation.

The 2025 and 2026 evidence points more strongly toward AI. ShipFast managed to hover around $15,000 to $20,000 for much of 2025 while vibe-coding products were still filling production gaps. Those same competitors subsequently added precisely the infrastructure ShipFast sells, while Lovable reached hundreds of millions in annualized revenue and Replit moved toward a vastly larger business. ShipFast then fell from roughly $20,300 at the end of 2025 to only a few thousand dollars today.

We cannot turn that correlation into a clean causal percentage. Marc redirected his attention, launched competing products, bundles ShipFast with CodeFast and operates a lifetime-purchase model that naturally decays.

But the direction is difficult to dispute. AI arrived after the decline began, then made a recovery increasingly unlikely by attacking the exact source of economic scarcity on which the product had been built.

So, did AI kill ShipFast?

Mostly yes, but not in the simplistic sense. AI did not cause ShipFast’s original collapse, and ShipFast is not literally dead. What AI appears to have killed is ShipFast’s position as a uniquely valuable shortcut for launching software.

The business had already normalized sharply from its viral 2024 peak before modern vibe coding arrived. That first 60%-plus decline is better explained by an extraordinary launch cycle, one-time-product economics and the natural maturation of a breakout indie product.

After that, however, the evidence changes. Lovable, Replit and Bolt turned authentication, databases, payments, deployment and other former “boilerplate work” into increasingly automated platform features. Their combined adoption became enormous. Marc’s own audience continued expanding, yet ShipFast revenue kept shrinking. By now the product contributes only a few percent of the revenue generated by Marc’s leading products, and he effectively gives ShipFast away inside a $299 CodeFast bundle that costs the same as CodeFast alone.

The cleanest judgment: AI did not kill ShipFast as a product. It helped kill ShipFast as a breakout business.

ShipFast can still be useful, still sell copies and still generate high-margin income for years. Its strongest remaining use case may even involve AI: a proven codebase that gives Cursor or Claude a reliable starting point.

But the original promise, “pay us because integrating all this yourself takes days,” has been structurally weakened. Increasingly, founders can ask an AI agent to do the work instead.

That is not a temporary competitive problem. It is a change in what boilerplate code is worth.