Building a 4-App Micro-SaaS Portfolio in One Year as a Solo Vibe Coder

Building a 4-App Micro-SaaS Portfolio in One Year as a Solo Vibe Coder

Most solo builders treat their first app like it has to be the one. They spend ten months polishing every screen, chasing every feature request from three beta users, and hoping it turns into a life-changing exit instead of the first piece of a real micro-SaaS portfolio. That is the slow way to build wealth as a vibe coder, and it is also the riskiest.

The operators actually compounding their income run a portfolio instead. Four apps a year, not one. Two of them flop by week six. One breaks even. One quietly turns into $8,000 a month. The math works because you are not betting a year on a single outcome. You are running four small, time-boxed experiments with an actual decision process behind each one.

This is that process: a four-quarter build calendar, a scorecard for deciding what to kill and what to nurture, and the real numbers from solo operators who already run this system at a scale most weekend builders would not believe.

Why a Micro-SaaS Portfolio Beats One Big Bet

Marc Lou runs 15 to 16 products at once. Not sequentially, concurrently, with most of the revenue coming from two or three of them at any given time. In June 2026 his portfolio pulled in $83,701 across the month: TrustMRR did about $30,000, DataFast around $21,000, Ship or Die roughly $15,000, CodeFast about $9,000, ShipFast around $3,000, and a long tail of smaller apps split the rest. He publishes these numbers publicly every month, and the pattern never changes: a couple of winners carry the group.

Tony Dinh shipped about ten products and kept four as long-term businesses. TypingMind, a ChatGPT wrapper he built in a single day, crossed $1 million in lifetime revenue within 20 months and now runs $130,000 to $160,000 a month. DevUtils peaked around $20,000 a month and now ticks along at roughly $8,000 on autopilot. The other six products he tried got dropped, not because they failed dramatically, but because they never earned another quarter of his attention.

Pieter Levels has shipped more than 40 products since his original "12 startups in 12 months" challenge. His active lineup, Nomad List, RemoteOK, Photo AI, and Interior AI, brings in roughly $250,000 a month combined, with Photo AI alone carrying close to 70% of that. He does not run a formal kill process. He ships fast, watches which idea gets organic pull, and reinvests his time there.

None of these builders spent a year on one idea before finding out if it worked. They shipped in weeks, watched the data, and only committed real time to the one that pulled its own weight. A portfolio strategy only pays off if you actually run the cycle four times instead of getting attached to app number one.

The Four-Quarter Build Calendar

Treat the year as four 12-week sprints, one app per quarter, with a fixed rhythm inside each:

• ·Weeks 1-2: Build. MVP only. If it is not usable by day 14, the scope is too big.
• ·Weeks 3-4: Launch and polish. Ship to a small list, fix the obvious rough edges, get the first 10 real users.
• ·Weeks 5-10: Traction sprint. Push distribution, not features. Post in the communities where your buyer already hangs out, collect feedback, watch for organic pickup you did not force.
• ·Weeks 11-12: Decision week. Run the app through the scorecard below and commit to kill, nurture, or scale before starting the next quarter.

The tooling in 2026 makes this cadence realistic in a way it was not two years ago. Cursor's Composer 2 update reads up to 15 files before making an edit, which cuts down the back-and-forth when you are working across four separate codebases in the same year. Replit's Agent 4, released in March 2026, runs parallel tasks across isolated micro-VMs and resolves roughly 90% of merge conflicts on its own, useful if you are running two builds side by side during a transition week. Lovable's pricing covers multiple projects under one account rather than charging per seat the way Bolt does, which matters once you are juggling four live codebases instead of one.

Supabase moved to organization-based billing in July 2026, so your usage quota applies across every project in the org instead of resetting per app. The free tier still covers two projects, Pro runs $25 a month plus usage, and database branching costs $0.01344 per branch-hour if you want an isolated preview environment for each app during the build sprint. For a four-app portfolio, that one change means you are not paying four separate platform bills just to keep everything running.

If you have not settled on a stack yet, our comparison of Cursor, Bolt, and Lovable breaks down which one exports the cleanest code for a buyer's due diligence, which matters more when you are planning to sell three of the four apps you build.

The 30/60/90 Kill-Nurture-Scale Scorecard

Vague gut feel is how builders end up carrying five dead apps into year two. Score each app at day 30, day 60, and day 90 of its traction sprint against fixed thresholds:

This is a tighter version of a rule that circulates in indie hacker circles: sunset anything that has not reached meaningful recurring revenue within six months. Running it on a 90-day clock instead of six months fits a four-quarter calendar and forces the decision before you have sunk another sprint into an app that already told you the answer.

"Kill" does not mean delete the repo. It means stop building new features and either list it for a quick sale or leave it running untouched. "Nurture" means one more quarter of light iteration before the next checkpoint. "Scale" means this app becomes your quarter 3 and quarter 4 focus, and the remaining build slots in the calendar shrink to maintenance only.

Picking Niches That Won't Waste a Quarter

The niche you pick determines whether the scorecard even gets a fair shot. A few categories are pulling real demand in 2026: vertical and compliance-adjacent tools for healthcare, legal, or finance operators, payment recovery and dunning tools that run at 70% to 90% margins, AI support-agent add-ons for small SaaS teams, and unglamorous workflow tools like wedding vendor coordination, Amazon seller inventory tracking, or podcast show-note generation. None of these will get you a conference talk. All of them have a buyer who is already annoyed enough to pay.

Where most vibe coders get this wrong: they build a bare ChatGPT wrapper because it is the fastest thing to ship, then wonder why it never gets past 10 signups. Industry estimates for 2026 put wrapper-app failure near 90%, with 60% to 70% of them generating zero revenue. A thin prompt with a login screen is not a product a buyer will pay 2-3x SDE for later. Pick the boring, specific problem instead. It is slower to explain at a party and faster to sell on Acquire.

What Happens When One App Wins

Somewhere around app two or three, one of them will clear the scale threshold. This is the moment the whole system pays for itself, and it is also where builders make their next real decision: hold the winner and sell the rest, or flip the winner too while it is at its peak multiple.

Tony Dinh sold BlackMagic.so, a Twitter growth tool, for $128,000 after it reached $14,000 MRR, then kept building TypingMind into a seven-figure product. Danny Postma sold Headlime for $1,000,000 after it hit $20,000 MRR and kept HeadshotPro, which was doing eight-figure annualized revenue by the end of 2023. Both made the same call: sell the app that has plateaued, keep the one still climbing.

The pricing data backs up that timing. Micro-SaaS businesses under $1 million ARR sold for an average of 2.85x SDE in 2026, with the top quartile of deals reaching 6.13x, based on a dataset covering more than 500 transactions. Apps sitting at $1,000-$5,000 MRR with stable retention land in the 2-3x SDE range. Push past $10,000 MRR with low churn and buyers pay 4-6x. That gap is the entire argument for the scorecard: an app you nurture into $10K MRR before listing it is worth roughly double what the same app sells for at $3K MRR, for maybe two more quarters of work.

Once an app clears "scale" and you decide to sell rather than hold, the transition from builder to seller has its own checklist. Our 30-day pre-listing checklist covers the documentation, metrics snapshot, and support-log cleanup a buyer expects to see before they wire funds.

Running the System Twice

The real payoff is not the first four-app year. It is the second one. By year two you already know which niches you can build fast in, which stack keeps your bills low across a portfolio, and roughly where the kill line sits for your own risk tolerance. That is the same compounding effect behind the serial vibe coder mindset: each cycle makes the next one cheaper to run and easier to price.

Four apps in a year is not about being four times as busy. It is about running the same disciplined sprint four times and letting the scorecard, not your attachment to the idea, decide what earns your next quarter. Pick the first niche, block out 12 weeks, and start the clock.