The founder's time budget: how to ship a SaaS in 90 days with Avery

Most solo SaaS founders never ship. Not because the product is bad. Not because they lack skills. Because they don't have a real time budget and so they drift.

A real time budget makes the difference. It defines what gets done each week, surfaces when you're behind, and forces the cuts that turn a sprawling idea into a shippable product. Without it, you have nine months of "almost done" and an unlaunched product.

This post is the 90-day plan that worked for me. The constraints, the weekly milestones, the failure modes I hit, and the honest economics of a solo founder shipping a real SaaS with Avery doing the build.

The frame

90 days. 3 focused hours per day. Real SaaS shipped at the end, with paying customers.

3 hours sounds short. It's actually about what most founders can sustain when they have a day job, family obligations, or both. The constraint is the point. If you can't ship a SaaS in 3 hours a day for 90 days, you either need to hire help, cut scope dramatically, or accept that this isn't the right product.

The 3 hours have to be focused. Not 8 hours of distracted work. Not 3 hours of meetings, email, and Twitter. 3 hours of CR writing, PR review, customer calls, and copywriting. Real work, no context switching.

If you can't carve out 3 contiguous focused hours, you're not running a founder schedule, you're running a hobby schedule. Hobby schedules don't ship SaaS.

Week 1: Scope and CR breakdown

Day 1 to 5. The goal: write the elevator pitch in 3 sentences and break the product into 10 to 15 CRs.

The elevator pitch test: if you can't describe what the product does in 3 sentences, you don't know what you're building. Examples:

Bad: "An AI-powered productivity platform that helps teams collaborate."

Good: "A scheduler for podcast guest interviews. Hosts share a link, guests pick a time, calendar invites get sent automatically. Free for the first 5 bookings, $19/month after."

The CR breakdown: list every feature your MVP needs. For each, write a one-paragraph CR. If the list exceeds 15 CRs, cut. If a single CR is too big to ship in a day, split.

What stays in MVP: authentication, the core workflow (whatever your product does), payments via Stripe, transactional email, a landing page.

What gets cut for MVP: integrations beyond the bare minimum, admin dashboards, analytics, mobile apps, anything labeled "nice to have."

The discipline of week 1 is saying no. Most founders fail here because they can't.

Weeks 2 to 6: Build with Avery

Day 6 to 35. The goal: ship CRs at a rate of 2 to 3 per week.

Each day:

Pick one CR from the backlog (45 minutes including any thinking required).

Write the prompt to Avery, attach acceptance criteria (15 minutes).

Wait for Avery to PR the work (background time, you do other things).

Review the PR carefully. Run it locally. Test the happy path and one edge case (60 minutes).

Either merge or request changes with specific feedback (15 minutes).

Update your CR backlog with anything you discovered (15 minutes).

That's the 3-hour day in build mode. Roughly one CR shipped per day if you're focused, sometimes two if the CR is small.

By end of week 6, you should have shipped 10 to 15 CRs. The product is functional. It's not pretty, it's not feature-complete, it's not ready for a big launch. It's functional.

If you're behind at week 6, you scoped too big. Cut more.

Week 7: Customer interviews

Day 36 to 42. The goal: stop building. Talk to 10 to 15 potential users.

This is the most important week. It's also the one most founders skip because building feels productive and selling feels uncomfortable.

The interview structure:

Find people in your target market. LinkedIn, Twitter, Slack groups, Reddit, your existing network.

Reach out with a short message: "I'm building [thing]. Can I show it to you for 20 minutes and get your honest reaction?"

In the interview, demo for 5 minutes, then shut up. Ask: what's the most frustrating part of [the problem your product solves] for you right now? Do you currently use anything to handle this? What would have to be true for you to pay for this?

Take notes religiously. Don't argue. Don't sell. Listen.

Patterns will emerge. Maybe nobody cares about the feature you spent 3 weeks on. Maybe everyone asks for a feature you skipped. Maybe your pricing assumption is wrong by 5x.

The product you're building at week 2 is almost certainly wrong by week 7. The interviews tell you how wrong and where to course correct.

Weeks 8 to 10: Refine and launch prep

Day 43 to 63. The goal: apply interview feedback, write the landing page, set up payments, prepare to launch.

You'll have new CRs from the interviews. Maybe 5 to 8 of them. Ship the highest-leverage ones. Defer the rest.

Landing page is the underrated work of these three weeks. Most founders write landing pages in an afternoon and wonder why nobody signs up. A good landing page takes a full week if you're thoughtful about it.

The landing page job: explain in 30 seconds what the product does, who it's for, what the customer outcome is. Show one screenshot or demo video. Make the call to action obvious.

Common landing page failures: too much jargon, no clear outcome, no social proof (which you don't have yet because you haven't launched), a CTA that's "join the waitlist" when it should be "start your free trial."

Stripe setup: takes a day. Don't underestimate the time. Pricing tiers, products, webhooks, customer portal. All of it.

Transactional email: another half day. Welcome email, payment confirmation, password reset. Use Resend or similar. Don't try to write your own email server.

This is also the week most founders give up. The build is done; the launch feels intimidating. The work shifts from coding to marketing, and many engineers are uncomfortable with marketing.

The trick is to schedule the launch before you feel ready. If you wait until you feel ready, you'll never feel ready. Pick a date in week 11 and commit publicly.

Week 11: Launch on Product Hunt

Day 64 to 70. The goal: ship the launch and ride the spike.

Product Hunt launches on Tuesdays for best traffic. Show up at 12:01 AM Pacific time. Have your team and friends ready to upvote in the first two hours.

The pre-launch week:

Send the link to your network 24 hours before launch. Personal asks, not mass emails.

Coordinate with anyone who agreed to upvote or comment in the first two hours.

Write a great launch comment. Founder voice. Honest about what works and what doesn't.

Pre-write social posts. Twitter thread, LinkedIn post, email to your list (if you have one).

On launch day:

Start at 12:01 AM PT. The first 2 hours determine your ranking.

Reply to every comment within 15 minutes.

Email your full list at 8 AM PT.

Share progress on Twitter every 2 hours.

Cross-post to Indie Hackers, Reddit (carefully), and any communities you're part of.

You'll get more traffic in 24 hours than you got in the previous 10 weeks combined. Some of that traffic will convert to signups, some to paid users. Most will bounce. That's normal.

Week 12: Post-launch

Day 71 to 90. The goal: convert launch traffic, learn from real users, plan the next 90 days.

The post-launch week is where most founders crash. The launch was exciting. The post-launch is grinding.

Reply to every customer message. Bug reports, feature requests, complaints, praise. Treat every message as a signal.

Schedule customer calls with the first 10 paying users. Learn what they actually do with the product, what almost made them not sign up, what they wish was better.

Ship 1 or 2 CRs based on the early feedback. Don't ship everything. Pick the highest-leverage one or two.

Plan the next 90 days. What's the next milestone? More paying customers? A specific revenue target? A new market segment? Choose one.

Honest economics

The actual numbers from my 90-day build:

Avery subscription: about $200 in compute over 90 days (varied by CR complexity).

Vercel hosting: $20 per month.

Supabase Postgres: $25 per month.

Stripe fees: percentage of revenue, didn't matter much in first 90 days.

Domain: $12 for the year.

Resend for email: free tier covered the first 90 days.

Total infrastructure: about $400 over 90 days.

Revenue: 47 beta signups during launch week, of which 12 converted to paid at $19/month. Month 4 MRR: $228.

Not life-changing. But real.

The honest point: 90 days does not produce a profitable SaaS. It produces a real product with real users and a foundation to build from. The next 90 days is about growth, not building.

Common failure modes

Scope creep. Every week you'll discover something you wish was different. Don't change scope. Add to backlog. Ship MVP first.

Avoiding customer interviews. Building feels productive. Selling feels uncomfortable. The founders who don't interview never get product-market fit. The founders who interview ship something people want.

Perfectionism on the landing page. You'll spend three weeks tweaking the landing page if you let yourself. Set a hard deadline (end of week 10) and ship whatever's there.

Skipping the launch. "I'll launch when I have more features" is the death sentence. Launch with what you have. Ship features post-launch based on user feedback.

Burning out by week 7. 90 days at 3 hours per day is sustainable. Working weekends and evenings on top of a day job is not. If you're already burned out by week 4, you're working too hard. Pace yourself.

Going too cheap on infrastructure. Free tiers will hit limits during launch traffic. Pay for the next tier 24 hours before launch. The $50 is worth not crashing at hour 3.

What 90 days teaches you

The 90-day frame teaches you three things you can't learn any other way:

How long things actually take. Most founders dramatically underestimate. After 90 days you have calibrated estimates.

What your customers actually want. Pre-launch, you have hypotheses. Post-launch, you have data.

Whether you can sustain founder mode. Some people thrive in 3-hour daily focus. Some can't sustain it past week 4. Better to find out at week 4 than at year 2.

If the 90 days produced a product with paying customers, you have something to build on. If they didn't, you've learned what doesn't work for a fraction of the cost of a typical startup attempt.

The Avery angle

Avery's role in this is the CR-execution layer. You're the founder. Avery is the engineer who never sleeps.

The leverage: you focus on what only you can do (talk to customers, decide what to build, write the landing page, run the launch). Avery handles the part that previously required hiring an engineer or burning out doing it yourself.

3 hours of CR review and customer work beats 8 hours of writing code yourself. Both because you ship more, and because the customer work compounds (each interview makes the next CR more targeted).

The pattern: most solo founders fail because they can't do everything. The 90-day frame with Avery lets you do the things only you can do, while the AI does the rest.