跪拜 Guibai
← Back to the summary

90 Days to Your First Paid Customer: A Solo Developer's Monetization Playbook

One Person Is an Entire Engineering Team: A 90-Day SOP for AI-Era Super Individuals to Make Money

Author: 怕浪猫

cover-03.jpg

Let me be clear up front: this article is not teaching you how to "make 100,000 RMB a month."

You're tired of that kind of headline too. I want to talk about a smaller, harder, and easier-to-verify goal: completing one real transaction within 90 days.

Notice three words: real, transaction, 90 days.

Not "built a cool project," not "gained this many followers," but someone is willing to pay for your deliverable. Even if the first order is only 999 RMB.

Because going from 0 to 1 and from 1 to 100 are two completely different capabilities. Most people get stuck at 0 to 1, not because their technical skills are insufficient, but because they never complete the "demand—delivery—payment" loop.

1. Why the Window Is Opening Now

In the past, the biggest barriers to individual monetization were not technical, but "trust" and "channels."

You had to first prove you can do it, then prove you are reliable, and finally find someone willing to pay. Three mountains, climbed alone, slowly.

Now all three mountains have shrunk a bit.

First, the demand side has been educated.

Lovable announced in September 2026 that its ARR reached $600 million, with a valuation of $13.3 billion. [1] The significance of this number is not how much money it makes, but that: a large number of people who "shouldn't be building software" are now paying for "making usable software."

Demand has been validated. What remains is supply. And in supply, the scarcest thing is not "people who know how to use tools," but "people who can deliver something to the point of being usable."

Second, companies are starting to buy "fragmented professional capabilities."

Upwork's 2026 report shows: demand for skills explicitly involving AI grew 109% year-over-year, AI integration grew 178%, AI chatbot development grew 71%; 77% of business leaders say AI increases demand for "specialized, fragmented talent" rather than traditional full-time positions; nearly half of leaders are willing to pay a premium for "creative, innovative" independent talent. [2]

Another report, the Future Workforce Index 2026, shows: skilled freelancers as a share of skilled knowledge workers rose from 28% in 2025 to 38% in 2026; 58% of full-time employees are considering freelancing, up from 36% the previous year. [3]

Third, domestic budgets are also shifting toward AI.

In the first half of 2026, new AI job postings grew 789% year-over-year, with about 1 in 4 new jobs being an AI job. [4]

Most of these jobs are at big tech companies and funded companies. But the spillover effect of budgets is real: big tech builds platforms, small companies build applications, individuals fill the gaps. Gaps are characterized by: small amounts, very specific needs, ignored by big tech, and beyond the capability of small companies.

This is the survival space for the super individual.

By the way, here's an overused but still effective example: independent developer Pieter Levels, maintaining a dozen products alone, with public income consistently at the level of millions of dollars, and a team size of 0. [5]

His model is hard to replicate, but it verifies one thing: in the software industry, the relationship between "headcount" and "output" is no longer linear.

2. First Define Your "First Bit of Money"

"Monetization" is too vague a word. Let's split it into five concrete paths.

Path Startup Cost Payment Cycle Scalability Who It Suits
Delivery contracts (projects/consulting) Low 2–6 weeks Low Has engineering experience, can communicate independently
Templates/plugins/scripts Medium 1–3 months High Familiar with a specific niche scenario
Subscription tools (micro-SaaS) Medium-high 3–6 months Highest Can maintain long-term, understands customer acquisition
Diagnostics/consulting Very low 1–7 days Medium Has industry experience, can articulate
Content → services Low 1–3 months Medium Can write, willing to keep producing

If this is your first attempt, my advice is: first use "delivery contracts" or "diagnostics" to complete the loop, then use "templates/subscriptions" to scale up.

The reason is simple: delivery contracts allow you to get in touch with real paying users fastest and get real feedback; productization requires you to have a deep enough understanding of demand, otherwise it is easy to build something that "only you think is useful."

Collect money first, then optimize. If you reverse the order, you may spend half a year still "polishing."

fig-06-roadmap.png

Figure 1: 90-day roadmap. The three phases cannot be skipped, especially "collecting money."

3. 0–30 Days: Choose a Demand That "Someone Is Already Paying For"

The first month, do only one thing: find a real, specific demand that someone is already spending money to solve.

Note, not "find a good idea," but "find an existing expenditure."

Three Signals

  1. Someone is currently toughing it out with Excel / manual work / outsourcing. For example, exporting reports weekly, manual copy-paste every time, hiring interns for repetitive sorting. Manual labor handling it means the demand is real, just lacking a good tool.
  2. There is a clear budget line item. Marketing fees, labor costs, software procurement, outsourcing services—which pocket the money comes from determines how much you can charge.
  3. The decision chain has no more than two people. One person raises the demand, one person signs off. The longer the decision chain, the slower the sale, and the less an individual can afford it.

How to Find It

Don't sit at home thinking. Use three ways:

Then feed the 10 notes to AI, have it cluster the "most frequent pain points" and the "scenarios most willing to pay."

One-Sentence Value Proposition

Write out three things clearly:

I help [who] take [what process] from [current state] to [result], for a fee of [amount].

For example:

I help small and medium foreign trade companies take "inquiry classification + quote generation" from 2 hours per day to 20 minutes, charged 12,000 RMB in the first year.

If you can't write it, the demand isn't clear enough yet.

Three Anti-Patterns

Chapter 63 of the Tao Te Ching says:

Difficult things in the world must be done from the easy; great things from the small.

The first order doesn't need to change the world, only to save one person two hours of overtime.

4. 31–60 Days: Build a "Minimal Loop That Can Collect Money"

In the second month, turn the demand into a deliverable thing.

The keyword is "can collect money," not "feature complete."

Four "Ones"

Anything beyond this gets cut.

Tech Stack: If It Can Be Manual, Never Automate

Scenario Recommended Solution Reason
Web application Next.js + Supabase + Vercel Frontend and backend integrated, fast deployment, free tier is enough for validation
Internal tools Feishu multi-dimensional table + AI fields/robots No frontend development needed, business people can modify it themselves
Data processing Python script + scheduled tasks Fastest to get running, no need to agonize over architecture
Conversational Ready-made bot platform + knowledge base Validate demand first, don't build your own inference pipeline

The only principle: before the first 10 users, any design "for scalability" is waste.

Paul Graham wrote a famous essay, Do Things That Don't Scale. [6] The core point: early startups often need to do "non-scalable" things—manually serving users, pulling people in one by one, personally handling orders. These things look inelegant but keep you closest to users.

Individual development is the same. The first version can be semi-automated, even "AI + manual" hybrid delivery. Users want results, not how advanced your architecture is.

Acceptance Criteria

By the end of the second month, three conditions must be met:

  1. 3 real users have completely run through the process (not a demo, real data);
  2. Data can persist (it's still there next time, not a one-off demo);
  3. Errors can be rolled back (at least one backup and one "undo" plan).

Xunzi's "Encouraging Learning" says:

Gentlemen are not born different; they are good at borrowing external things.

In the AI era, the cost of "borrowing external things" has dropped to historical lows. The real difference is not whether you can use tools, but whether you know what to borrow and what to use it for.

5. 61–90 Days: Collect the First Bit of Money

In the third month, do only one thing: close the deal.

Pricing Ladder

fig-07-pricing.png

Figure 2: To reach the same 10,000 RMB monthly revenue, the higher the average order value, the fewer deals needed. Estimates based on assumptions in the article, not promised returns.

I suggest designing three tiers from the start, not one price:

The purpose of three tiers is not to "let the customer choose the cheap one," but to give the middle tier a reference point. With only one tier, customers only haggle; with three tiers, customers first ask "which fits me better."

Anchoring the Quote

Don't quote based on "how long I worked," quote based on "what the other party loses if they don't do it."

The essence of quoting is helping the other party do the accounting, not proving how hard you worked.

Customer Acquisition: 3 Channels × 20 Touches

Don't expect one viral piece of content to bring customers. In the early stage, the most effective is "narrow and deep" outreach:

  1. Acquaintances/former colleagues/former clients: one-on-one explain what you're doing, ask for referrals. 20 touches usually yield 2–3 leads.
  2. Vertical communities: industry groups, regional groups, tool groups. Don't post ads; first answer 10 specific questions, then leave a "I've done this problem, DM me" line.
  3. Platforms/content: Upwork, Zhubajie, Juejin, Xiaohongshu, WeChat official accounts. Content doesn't need to be much; one "how I solved problem XX" is enough, the key is leaving a contact entry point.

Four Elements of a Contract

The most common loss for individual contractors isn't price, it's scope.

Doesn't need to be complicated, one page is enough. But it must exist.

Unit Economy Calculation

Average order value Deals needed for 10K monthly revenue Avg delivery hours per deal Monthly delivery hours Main bottleneck
99 RMB 101 deals 0.5 hour 50 hours Traffic and conversion
999 RMB 10 deals 8 hours 80 hours Trust and case studies
9999 RMB 1 deal 40 hours 40 hours Case studies and referrals

This table is not for you to pick "the most profitable tier," but for you to see: each tier has a different bottleneck.

The 99 RMB tier is about traffic, the 999 RMB tier is about trust, the 9999 RMB tier is about case studies. Whatever you have at the moment, start with that tier.

6. After 90 Days: Service-ize or Productize

After completing one loop, you face a fork.

Service-ization: continue taking orders, raise the average order value, go deeper into the industry, rely on referrals and word of mouth. Advantages: fast cash flow. Disadvantages: still tied to your time.

Productization: extract the repetitive delivery parts into templates, tools, subscriptions. Advantages: low marginal cost. Disadvantages: requires continuous customer acquisition and maintenance.

My suggestion: survive first with service, then scale with productization.

Service lets you understand demand; product lets you escape time. Reverse the order and the product will go astray; only doing service and you'll never grow.

Sun Tzu's "The Art of War" - Chapter on Situational Advantages says:

In all battles, engage with the orthodox, achieve victory through the unorthodox.

The orthodox is stable delivery, keeping promises; the unorthodox is differentiation, the point others can't replace.

For an individual, the "orthodox" is on-time delivery, no accidents, keeping your word; the "unorthodox" is your understanding of an industry, your unique workflow, your accumulated case studies. Without the orthodox, the unorthodox is speculation; without the unorthodox, the orthodox only leads to price wars.

Three Risk Warnings

  1. Don't quit your job immediately. Only consider it after side income covers basic living expenses for 3 consecutive months.
  2. Don't pursue three directions at once. Within 90 days, only one main line is allowed; everything else is noise.
  3. Don't neglect cash flow. Individuals don't have financing; prepayments and payment cycles matter more than profit.

7. 90-Day Acceptance Checklist

Pin this checklist where you can see it, and check it every 10 days.

Day 30

Day 60

Day 90

8. In Conclusion

In the Records of the Grand Historian · Biographies of the Money-Makers, there's another quote:

When it's dry, prepare boats; when it's flooded, prepare carts. That's the logic of things.

It feels counterintuitive, but it's actually planning ahead—by the time everyone needs it, you already have it in stock.

The opportunity of the AI era is not where "everyone is flooding in," but in the gaps where "demand has appeared, but supply hasn't caught up yet."

One person is an entire engineering team, not because you know everything, but because you can finally hand the repetitive parts to machines and keep the judgment and responsibility for yourself.

90 days don't need to change your fate.

You only need to complete one real transaction, and then you'll know whether this path can work.

References

  1. AI Expert Magazine, Lovable Hits $600M ARR as Vibe Coding Goes Enterprise (2026-09): https://www.aiexpertmagazine.com/lovable-600m-arr-vibe-coding-enterprise/
  2. Upwork, In-Demand Skills 2026 (2026-02): https://www.upwork.com/press/releases/upworks-in-demand-skills-2026-demand-for-top-ai-skills-more-than-doubles-as-ai-is-embedded-into-everyday-work
  3. Upwork, The Future Workforce Index 2026: https://www.upwork.com/research/research-future-workforce-index-2026
  4. 36Kr, 2026 Autumn Recruitment, Does It Require Fresh Graduates to Have an Average of "3 Years of Work Experience"?: https://36kr.com/p/4025027072577414
  5. Pieter Levels' public products and income: https://levels.io/
  6. Paul Graham, Do Things That Don't Scale: https://www.paulgraham.com/ds.html
Comments

Top 1 of 2 from juejin.cn, machine-translated. The original thread is authoritative.

掘一二三桶金

Impressive

怕浪猫

Send