The five ladders at a glance
| Ladder | 12-month realistic income | Capital required | Audience required | Hardest part |
|---|
| Paid newsletter (Substack) | $0 to $20K total · median ~$4K/yr platform-wide | $0 to $500 | Pre-existing or built in public | Compounding without an audience seed |
| Indie SaaS | $0 to $10K MRR if lucky · most plateau under $1K | $500 to $5K | Useful but not required if SEO works | Distribution, not code |
| Video creator (YT/TikTok) | Often $0 to $5K total in year one | $0 to $2K (gear) | Built on platform | Sustaining output for 12+ months at zero feedback |
| Independent consultant | $60K to $250K possible by month 12 | $0 | LinkedIn + reputation in a niche | Pipeline consistency, not delivery |
| Boutique AI studio (1-3 people) | $150K to $600K firm-level | $5K to $25K | Founder network | Hiring the second person well |
LadderPaid newsletter (Substack)
12-month realistic income$0 to $20K total · median ~$4K/yr platform-wide
Capital required$0 to $500
Audience requiredPre-existing or built in public
Hardest partCompounding without an audience seed
LadderIndie SaaS
12-month realistic income$0 to $10K MRR if lucky · most plateau under $1K
Capital required$500 to $5K
Audience requiredUseful but not required if SEO works
Hardest partDistribution, not code
LadderVideo creator (YT/TikTok)
12-month realistic incomeOften $0 to $5K total in year one
Capital required$0 to $2K (gear)
Audience requiredBuilt on platform
Hardest partSustaining output for 12+ months at zero feedback
LadderIndependent consultant
12-month realistic income$60K to $250K possible by month 12
Capital required$0
Audience requiredLinkedIn + reputation in a niche
Hardest partPipeline consistency, not delivery
LadderBoutique AI studio (1-3 people)
12-month realistic income$150K to $600K firm-level
Capital required$5K to $25K
Audience requiredFounder network
Hardest partHiring the second person well
Ladder 1 · Paid newsletter writer
Substack is the most documented version of this path. The platform reported roughly 100K publications earning money as of April 2026, growth from 50K a year earlier, and gross writer revenue around $450M in 2025. Median writer earnings sit near $4,000 per year. The top quartile clears $16,000 per year. Roughly 50 writers earn $1M+ per year on the platform — the visible ceiling that fuels most of the marketing.
The twelve-month version is realistic only if you start with something to say. Months one to three: publish once a week, build the free list to 500-1000 subscribers through cross-posting in Notes, guest posts on adjacent newsletters, and Twitter/Bluesky. Months four to six: turn on paid tier at $5-8/month with one paid post per month; expect 1-3% of free subs to convert. At 1000 free and 2% conversion, that is ~$120/month. Months seven to nine: launch a paid cohort or annual founding-member tier; this is where serious money first appears if it appears at all. Months ten to twelve: either you have crossed $1K MRR and have a credible path to $5K, or you are looking at why you didn't.
AI stack is light. A language model (Claude or GPT) for outline pressure-testing and fact-check passes. A research-aggregation tool (Perplexity, Exa, or similar) for source surfacing. A grammar/style pass (the built-in one is fine; specialist tools add little for a serious writer). Image generation for occasional headers — most successful Substacks use almost no AI-generated images and lean on a personal visual identity instead. The AI does not write the newsletter; it shortens the loop between idea and shipped draft.
Newsletter · 12-month timeline
Month 1-2
Find the angle and ship weekly
Pick one beat. Publish 8 posts before doing any optimization. The first 8 are calibration — voice, cadence, what readers respond to. Free tier only.
Month 3-4
Network the launch list
Cross-post in Notes. Guest on three adjacent newsletters. Aim for 500-1000 free subscribers. Still no paid tier.
Month 5-6
Turn on paid
Set $5-8/month or $50-80/year. Move 1-2 of your strongest posts behind the paywall. Expect 1-3% conversion of free to paid.
Month 7-9
Add a second product
Founding-member tier at 5x-10x base price, or a cohort/course. This is where annual revenue meaningfully diverges. Without a second product, growth caps at subscriber growth.
Month 10-12
Honest assessment
If you are not at $1K MRR, the issue is almost always topic-market fit or distribution, not writing quality. Decide: pivot, double down, or sunset.
Ladder 2 · Indie SaaS
The reference case here is Pieter Levels, who publishes Stripe MRR screenshots and runs a portfolio reportedly past $250K/month in 2025-2026, with Photo AI alone around $130-138K/month as of late 2025. That is the visible top of a long-tail distribution. Tyler Tringas built Storemapper to roughly $40K/month before selling to SureSwift Capital, and has written extensively about the actual mechanics at TylerTringas.com. The Indie Hackers community publishes enough MRR data to triangulate: a working rough split is that ~30% of started projects never cross $1K MRR, ~50% plateau between $1K and $10K MRR, ~15% reach $10K-$100K MRR, and ~5% exceed $100K MRR. Median time from launch to $10K MRR for the cohort that gets there is 12 to 36 months.
The twelve-month milestone path that the data supports: month one, pick a problem you have lived with for at least a year and have at least five other people you know also live with. Months two to three, ship an MVP. Levels-style single-page sites with Stripe Payment Links work and avoid most of the SaaS plumbing tax. Months four to six, find your first 10 paying customers manually — not through ads, through DMs, communities, and personal email. Months seven to nine, the inflection point is whether usage compounds. If retention is bad, no growth channel will save it. Months ten to twelve, you should know whether the unit economics work; if churn is over 8-10% monthly for a B2B tool, the business does not exist yet.
The single most under-stated lesson from both Levels and Tringas: distribution is the hard part, not building. Tringas has been explicit that his first version of Storemapper was written on a flight; the next several years were about finding customers, not features. AI changes the build cost but does not change the distribution cost.
Indie SaaS · the AI stack that matters
- Code assistant in the IDE (Cursor, Claude Code, or similar) — primary leverage; cuts implementation time roughly 30-50% on common CRUD work based on public reports, less on novel architecture
- Hosted database with no-ops posture (Supabase, Neon, PlanetScale) — keeps you out of database administration that you will lose hours to otherwise
- Payments — Stripe Payment Links for V1, full Stripe Checkout once you need subscriptions; build no billing logic yourself for at least the first 12 months
- Email — transactional (Resend, Postmark) and marketing (Loops, ConvertKit) on separate rails so you do not lose deliverability when one breaks
- Analytics — PostHog or Plausible; resist Mixpanel-tier complexity until you have 500+ users to justify it
- LLM API for in-product AI features — Anthropic, OpenAI, or Google; budget at minimum 10x what your napkin math says because users hit edge cases
Ladder 3 · Video creator (YouTube / TikTok)
This is the lane with the worst median outcome and the highest visible ceiling, which makes the survivorship-bias problem maximally severe. As of 2025-2026 best-effort, YouTube creator RPM (the actual paid-out rate per 1000 views after the platform's 45% cut) averages roughly $1-15 depending on niche. Finance and tech channels can reach $15-30 RPM; gaming and entertainment commonly sit at $1-5 RPM. YouTube Shorts RPM is roughly $0.01-0.06 — effectively zero for income purposes, though useful for top-of-funnel discovery.
Doing the math honestly: a channel at 100,000 monthly views in a $5 RPM niche earns roughly $500/month from ad revenue. To replace a $100K/year job from ads alone requires roughly 1-2 million monthly views in a strong niche, or 5-10x that in a weak one. This is why every credible solo video creator monetizes a second way — sponsorships, a product, a course, a community membership, an affiliate stack — within the first 12-18 months.
Milestone path: months one to three, ship 20-30 videos to learn what works for you. Most channels die here. Months four to six, you should see one or two videos materially outperform; analyze them ruthlessly and lean in. Months seven to nine, hit YouTube Partner Program eligibility (currently 1,000 subscribers and 4,000 watch-hours, check provider docs for current thresholds) and turn on monetization. Months ten to twelve, the inflection question is whether you have a second income stream beyond ads. Without one, you are an ad-rev creator competing on volume against full-time studios.
AI stack: a script and outline assistant (any frontier LLM), a thumbnail-iteration loop (image gen for variants, A/B test via TubeBuddy or vidIQ), and a transcription/clipping pipeline (Descript, Opus Clip) for short-form repurposing. Voice cloning and full AI video pipelines exist but tend to ceiling at low quality for human-led channels — they are a force multiplier, not a substitute for a human in front of the camera.
Ladder 4 · Independent consultant
The fastest path to replacing a tech salary in 2026, and the least romantic. Public rate guides put senior independent AI consultants in the US at roughly $150-500/hour, with specialized generative-AI work commanding 20-30% premiums. Mid-level operators commonly start at $150/hour, push to $300/hour within 12 months, and clear $500/hour after that if they have a defined niche and a stream of referrals. Day rates ($1,200-3,000) and project-based pricing tend to displace hourly billing as the consultant matures.
The math is more forgiving than any other ladder on this page. At $250/hour and 20 billable hours per week (which is realistic; the other 20 go to sales, ops, learning, and rest), gross is $250K/year. The hard part is not the math; it is keeping pipeline full so those 20 hours are actually billable every week. Most failed consulting practices fail not on rate or delivery but on the boom-bust cycle of selling only when you are not delivering.
Milestone path: months one to three, define the niche tightly enough that a stranger can repeat it back to you. 'AI implementation for mid-market e-commerce' is a niche; 'AI consulting' is not. Build a portfolio of two or three deeply-written case studies, even from past employment, with permission. Months four to six, two outreach motions in parallel: warm network (former colleagues, alumni, communities you are already in) and one consistent public-output motion (LinkedIn, podcast guesting, conference talks). Months seven to nine, you should be able to raise rates 20-50% as your first batch of work generates referrals. Months ten to twelve, the question is whether you can productize a piece of delivery to escape pure hours-for-dollars.
Ladder 5 · Boutique AI studio (1-3 people)
The 'studio' path is consultant plus partner plus, at most, one or two specialists. Firm revenue ranges from $150K to roughly $600K per year for a tight three-person studio doing AI implementation, with founder take-home varying widely depending on how much of revenue gets reinvested in pipeline and tooling. The 'agency' label is technically applicable but carries cultural baggage; 'studio' captures the smaller, higher-craft, lower-headcount posture better.
The key strategic difference from solo consulting is partner economics. Two people with complementary skills (e.g., one ML-leaning, one product-leaning) can sell larger engagements that no single consultant could deliver alone, and can keep delivery moving while one of them is in sales mode. The classic failure pattern is hiring the second person too fast — before you have repeatable enough delivery to onboard them productively — or hiring them as a contractor when the work needs an owner.
Milestone path: months one to three, the first $50K of revenue should come in before you even talk to a second person. The studio thesis only works if the solo version already works. Months four to six, identify the partner with complementary skills, ideally someone you have already collaborated with. Months seven to nine, formalize the partnership with a written operating agreement; equity split, decision authority, exit clauses. Months ten to twelve, the studio should be repeatedly winning engagements that neither founder could win alone. If you cannot point to one of those by month twelve, the studio is functionally two consultants sharing a brand.
AI stack scales with the work. Codebase-wide assistants (Cursor, Claude Code, Cognition's Devin where appropriate) for engineering work. Evaluation harnesses (Braintrust, Langfuse, or roll-your-own) for any client work involving LLMs — the moment you ship an LLM-backed product without an eval loop, you are flying blind. A shared knowledge base (Notion, Linear, or similar) is non-negotiable as soon as you are two people.
Failure modes by ladder
| Ladder | Failure 1 | Failure 2 | Failure 3 |
|---|
| Newsletter | Quitting at month 4-6 right before compounding | Optimizing for free subs at the expense of conversion | Topic too broad to attract a paying tribe |
| Indie SaaS | Building 6+ months before talking to users | Solving an interesting problem nobody is paying to solve | Founder burnout before distribution compounds |
| Video | Output collapse around month 4-6 at zero apparent traction | Niche too crowded to win on signal alone | Ad-rev-only monetization with no product or sponsorship layer |
| Consultant | Pipeline boom-bust cycle (sell, deliver, idle, panic, repeat) | Refusing to raise rates after first 3-5 clients | Scope creep eating margin on fixed-price projects |
| Boutique studio | Hiring the second person before repeatable delivery exists | Partner mismatch on values or work pace, no operating agreement | Saying yes to engagements outside the niche, eroding referrals |
LadderNewsletter
Failure 1Quitting at month 4-6 right before compounding
Failure 2Optimizing for free subs at the expense of conversion
Failure 3Topic too broad to attract a paying tribe
LadderIndie SaaS
Failure 1Building 6+ months before talking to users
Failure 2Solving an interesting problem nobody is paying to solve
Failure 3Founder burnout before distribution compounds
LadderVideo
Failure 1Output collapse around month 4-6 at zero apparent traction
Failure 2Niche too crowded to win on signal alone
Failure 3Ad-rev-only monetization with no product or sponsorship layer
LadderConsultant
Failure 1Pipeline boom-bust cycle (sell, deliver, idle, panic, repeat)
Failure 2Refusing to raise rates after first 3-5 clients
Failure 3Scope creep eating margin on fixed-price projects
LadderBoutique studio
Failure 1Hiring the second person before repeatable delivery exists
Failure 2Partner mismatch on values or work pace, no operating agreement
Failure 3Saying yes to engagements outside the niche, eroding referrals
Honest about luck and survivorship bias
Almost every public success story you can name on these ladders has at least one of: a prior audience, prior savings runway, a topic of years-long expertise, a co-founder relationship that pre-existed the venture, or a launch moment that benefitted from platform momentum (early Substack, early YouTube algorithm preference for a niche, early ChatGPT-API wrappers). None of this invalidates the path. It means the timelines you see in public are not the median timelines, and most public retellings of those journeys have a clean narrative imposed on top of months or years of stochastic noise. Plan against medians, give yourself enough runway to survive the noise, and assume that the version of you who succeeds is going to be the one who can still ship a useful thing in month nine after eight months of zero feedback. The median Substack writer earns roughly $4,000/year. The median indie SaaS does not cross $1K MRR. The median YouTube channel does not hit Partner Program eligibility. These are not reasons not to start. They are reasons to start with eyes open and to define your own success criteria against the distribution, not the tail.
The support network nobody mentions
A peer group of 5-10
Single highest-leverage social asset
Not a Discord with thousands; a small group at roughly your stage where you can be specific about numbers and stuck-points. MicroConf, Indie Hackers in-person meetups, and small Twitter/Bluesky circles all work. The thing that matters is honesty about numbers.
One mentor 1-2 stages ahead
Compresses calendar months into weeks
Someone who hit the milestone you are trying to hit in the last 12-36 months. Far more useful than someone who hit it 10 years ago in a different macro environment. Pay them if needed; cheaper than the mistakes you avoid.
A non-business human anchor
Not optional, just under-discussed
Partner, family member, close friend, or therapist who is not invested in the outcome and will tell you when you are burning out. Solo operators chronically under-index on this until something breaks.
A financial floor
The single highest predictor of survival to month 18
Six to twelve months of runway in a separate account, untouched. Operators who run thin under-perform structurally because they take bad deals to make rent. Either accumulate runway first or keep a part-time anchor.
One vertical community per ladder
Discovery channel that compounds over years
Substack writers have the SubstackWriters communities and platform-internal threads. Indie SaaS has Indie Hackers and MicroConf. Consultants have niche Slack groups. Creators have format-specific communities. Pick one and contribute.
A weekly review ritual
Operating system, not productivity hack
Friday afternoon, 60 minutes, alone. What shipped, what stalled, what is the one bet for next week. Solo operators drift without this; the calendar becomes whatever email demanded most. Build it before you need it.
If you take only one thing from this page
Pick one ladder for the next 12 months and run it without portfolio-switching. The failure mode that ends more independent-operator runs than any other is jumping between ladders every 90 days because the current one isn't working yet. None of them work in 90 days. Pick the one that matches your actual leverage — audience for newsletter and video, technical depth for SaaS, domain expertise for consulting, partner relationships for studio — and give it until month 12 before re-evaluating. If you genuinely cannot stand the work for nine more months, that is a separate signal and worth respecting. But aesthetic dissatisfaction with the slowness of compounding is not a signal; it is the cost of admission.
