Monday morning, one client line made me hesitate: “Our budget is only 5,000 yuan.”
Last Monday at 10 a.m., I was revising a quote in a Hangzhou café. The client wanted a sales sheet with automated reminders, and the budget was 5,000 yuan. I replied “we can do it” based on instinct, but felt uneasy: Would late-night revisions count as overtime? Who would pay when the scope changed?
So I broke the project into smaller tasks, estimated each one, listed the deliverables, and asked AI to simulate three client budgets. Then I realized I had originally planned to charge less than I should have.
I wasn’t using a magic app—just asking AI to be a tough pricing assistant
The method is simple: write down the client’s requirements, project boundaries, and target total price. Ask AI to estimate the time under “optimistic,” “normal,” and “conservative” scenarios, then flag work I might overlook, including revisions, communication, and maintenance.
The point isn’t to have AI invent a number. It’s to force myself to turn vague promises into a checklist.
I know an indie developer, Lin Yue, who takes on small business-tool projects in Shanghai. She keeps a “project time-estimate sheet” in Feishu and asks AI to play a demanding client before every quote: keep pressing on delivery dates, revision rounds, and after-sales coverage.
She has made mistakes too. She once wrote “unlimited revisions” into the scope and lost three evenings. Now she limits the project to two revision rounds, 48 hours apart, and charges extra for anything beyond that. Her negotiations have gone much more smoothly.
My setup today: 0 yuan, 20 minutes, no coding—the first step is copying three prompts
The cost is 0 yuan. A basic chat version of AI is enough; if you already pay for a subscription, you can use that. Set aside 20 minutes. The technical barrier is knowing how to type and copy-paste.
First step: open the AI chat tool you use most, start a new conversation, and paste these three prompts:
“You are a freelance consultant. Break the following requirements into tasks and estimate the hours for each.”
“List the communication, revision, and after-sales costs most likely to be missing from a quote.”
“Simulate a client saying the budget is too low, and help me explain the value without inventing fake experience.”
Then paste in the client brief.
One caution: don’t paste the client’s name, the full contract, or confidential material. Replace details with something like “build a client management system.”
If AI sounds too polished, ask it to label its assumptions and answer: “What missing information could make this estimate inaccurate?” AI can support the estimate, but I still make the final call on price. Not everyone needs this method, and waiting is fine. A small project worth around 1,000 yuan is enough to practice.
How I’d use it at different stages
If I’m just starting: I’d create a “minimum acceptable price” sheet that records my time, revision rounds, and deliverables. That keeps me from undercharging for my first client.
If I have one or two clients: I’d ask AI to review my actual hours, compare the quoted amount with the time spent, and add a 10% buffer for communication and rework on the next project.
If I’m scaling: I’d standardize my task-breakdown template, revision rules, and acceptance checklist. AI could help new teammates practice estimating, while I keep a human review on every quote. I’d also enjoy comparing notes with other freelancers about the mistakes we’ve made and which way of scoping projects fits each industry best.