Client Asks for Progress, I'm Still Digging Through Chat History
One Tuesday afternoon a client asked, "When are the spreadsheets we discussed going to be ready?" I stared at the WeChat log for half a day and had lost track of the progress myself. I panicked in that moment too: it's not that I don't know how to do the work — it's that the information is scattered across chats, memos, and email. Only after going freelance did I realize what actually slows us down isn't output. It's "where exactly did I record that promise."
Let AI Turn Conversations Into One Trackable Work Line
Later I set up Feishu Bitable (多维表格) plus AI to organize client conversations: after each chat, I drop the relevant messages into a table, and AI extracts the client's requirements, owner, deadline, and next step, then generates a to-confirm checklist. It's not a chatbot and it doesn't require code — the point is turning scattered information into a fixed record. Xiaolin, who runs an independent store design shop in Shenzhen, feeds it her client-group feedback every Friday, turning it into "pages to revise + owner + due date," and sends it to the client for confirmation the same day. I got stuck too: the first time I just asked AI to summarize, and it dropped something the client said verbally — "let's look next week." So I changed it to keep "items pending confirmation" as its own column, and now mistakes are obvious at a glance.
Replicating It Today: One Hour, Less Than a Cup of Coffee
Cost: Feishu's base features are usually free; AI organizing may incur model call fees, billed by actual usage — my own test volume runs under 20 yuan per month. Time: about 30 minutes to set up a basic version, plus another 30 minutes to get comfortable with the fields. Technical barrier: no programming needed, but you have to be willing to standardize how you record clients. First step: open Feishu Bitable, create a client follow-up table, click "Fields" or "Add field," and start with five columns: client, original message, requirements, deadline, next step. Not everyone needs this tool — if you only take on one or two clients a day, a plain memo might be less hassle. There's no rush to try it either.
Match It to Your Stage, No Need to Do It All at Once
If I had just started out, I'd use one table with only client name and next step, so promises don't slip. With one or two clients, I'd add deadline, items pending confirmation, and owner — after each chat, let AI draft the summary, then I check it myself. If I'm scaling up, I'd add automatic reminders, templates, and a weekly digest so the workflow doesn't depend on one person's memory. My biggest takeaway wasn't "how smart AI is" — it's that humans have to state the rules clearly. I might test it first with one old conversation with names removed, rather than importing every client file right away.