Last Wednesday night, I stared at my payment sheet
Last Wednesday night, I sat in my Hangzhou rental, staring at the payment sheet: one client had delayed the second installment by 18 days, and a subscription tool had just renewed. Outside, everyone was debating the AI market. In my head, one question remained: if payments came in even more slowly next month, could I keep running my business the same way?
I used to get carried away by the idea that “AI is hot,” and that missing the wave meant being late. I got this wrong too—I kept registering new tools and cutting demo videos without looking back to see who was actually willing to pay. Hype and cash flow live on two different spreadsheets.
What this market shift is really telling us
Reports say investors pulled out of AI trading after the election, sending the Nasdaq down 5%. For a small team like us, that does not mean “AI is useless.” It feels more like a reminder: valuations built on imagination cool before real money does.
I know an independent consultant, Lin Xiao, who organizes client records. She automates workflows for four small consulting firms in Chengdu. In November 2025, instead of renewing an expensive package, she changed her offer to “per delivery plus monthly maintenance.” She said clients bought easily during the boom, but once the hype faded, they cared more about saving two hours of rework. The tool did not change; the reason for charging did.
Build a “payment sheet” with what you can spend today
The cost to replicate this is essentially 0 yuan and takes about 30 minutes. The only barrier is being able to read customer payment records; no investing knowledge required. Open the accounting or project-management app I already use, create a sheet, and enter just four fields: client, due date, amount paid, and expected payment date. Then mark the fixed expenses.
If a client is more than seven days past the agreed date, I send a short message confirming the payment schedule before considering any discount. If software I know I will pay for next month accounts for over 10% of revenue, I first work out where it creates value in the client workflow, then decide whether to renew. There is no rush to try this, especially if deals are not yet consistent.
Match the next move to the current stage
If I am just starting: I would first find one real client willing to buy again, rather than chasing the news and changing direction. I would keep the payment sheet filled out for eight consecutive weeks before discussing automation.
If I have one or two clients: I would break the service into a small, acceptance-based result to shorten the payment cycle. I would give a late-paying client a clear option instead of sending repeated reminders.
If I am scaling: I would plan the budget around the money confirmed to arrive in the next 30 days, linking subscription counts to client counts. I would review whether each renewal is worth it at least 30 days in advance.
When the market cools, it becomes easier to see who is actually paying. Maybe we can start by filling in the first payment sheet together?