Originally published at PickGearLab — practical AI tutorials for writers, freelancers, and small business owners.
By Shahid Saleem
|
Founder & Editor, PickGearLab
|
5 min read
Client project management has a specific failure mode: things fall through the gap between conversations. A deliverable gets approved on a call, someone forgets to update the project board, no one sends the status email, and three days later the client is chasing you wondering what’s happening.
I spent two years solving this with manual discipline — religiously updating Trello after every call, drafting and sending status emails, writing end-of-week summaries. It worked, but it was friction-heavy and the first thing to slip when things got busy.
The stack I use now handles most of that automatically: Trello for project tracking, Zapier for automation triggers, and Claude for drafting the communications. Here’s how it’s wired together.
How the Three Tools Divide the Work
Trello is the source of truth. Every client project lives on a Trello board with cards representing deliverables. Cards move through lists: Backlog → In Progress → Review → Done. Everything that matters about a project — status, notes, attachments, due dates — lives in Trello.
Zapier watches Trello for events and triggers actions when things change. When a card moves to a new list, Zapier can send a notification, create a task, or fire a webhook. It connects Trello’s state changes to the rest of the workflow without any manual intervention.
Claude handles the communication drafting. Rather than sending automated robotic status emails, I use Claude to turn structured project data into natural-sounding client updates. The output goes into a draft — I review and send — not directly to the client.
Step 1: Set Up the Trello Board Structure
Every client gets a board with these standard lists:
- Backlog — agreed deliverables not yet started
- In Progress — active work this week
- Awaiting Client — delivered, waiting for feedback or approval
- Review — client feedback received, in revision
- Done — completed and signed off
- Blocked — stalled, needs action from client or third party
Each card has a due date, a description of the deliverable, and a checklist of sub-tasks. This structure is what Zapier reads to generate status data.
Step 2: Set Up the Zapier Automation
The core automation: when a card moves to “Awaiting Client” in Trello, Zapier fires a trigger. The trigger sends the card’s details — title, description, due date, any checklist notes — to a Google Sheet row (or a Notion database if you prefer). This creates a log of every deliverable submitted for client review.
A second automation runs on a weekly schedule (Friday at 4pm): Zapier pulls all cards that moved during the week — what was completed, what is awaiting feedback, what is in progress — and compiles a structured summary. This summary gets sent to a specific email address or Slack channel where I review it before sending anything to the client.
Setting up these two Zaps takes about 30 minutes. The Trello-to-Google-Sheets Zap is straightforward — Zapier has a template for it. The weekly summary Zap requires a search step to filter cards by date modified, which takes a few more steps but is well-documented in Zapier’s interface.
Get the next tutorial in your inbox
One AI tutorial or comparison per week. No filler, no listicles.
Step 3: Use Claude to Draft the Client Update
Every Friday I copy the weekly summary from the Google Sheet — the structured list of what moved, what’s waiting, what’s coming next — and paste it into Claude with this prompt:
“Write a brief client project update email based on this week’s activity. Tone: professional but direct, not overly formal. Structure: one paragraph on what was completed this week, one paragraph on what is awaiting their feedback and what I need from them, one paragraph on what is planned for next week. Keep it under 200 words. Don’t start with ‘I hope this email finds you well.'”
Claude produces a draft in about 15 seconds. I read it, adjust anything that doesn’t sound like me or that needs client-specific context, and send. The whole process — reviewing the summary, drafting, editing, sending — takes five to eight minutes.
The quality difference over a manually written update is mostly in consistency. When I write these manually after a long week, they’re often thin or slightly off in tone. Claude’s drafts are consistently well-structured, even when I’m tired.
What Gets Automated vs What Stays Manual
Automated: tracking card movements, compiling the weekly data, generating a structured summary log.
Manual (intentionally): reviewing the draft, editing for context, sending. Client communication is not something I want fully automated. The five-minute review exists because sometimes things happened during the week that the Trello board doesn’t capture — a call where expectations shifted, a scope change discussed informally, a client concern that needs addressing. The automated draft is a starting point, not a finished output.
The Result in Practice
Before this setup: status updates happened when I remembered them, which wasn’t always weekly. Clients sometimes chased me. The mental overhead of “I should send an update” was constant low-level friction.
After: weekly updates go out every Friday without fail. Clients have stopped chasing for status because they trust the cadence. I’ve had two clients specifically mention the weekly update as something that made working with me feel more organised than other freelancers they’d used.
The time cost is five to eight minutes per client per week. The relationship cost of not doing it is higher than that.
About the author
Shahid Saleem writes PickGearLab — a practical blog about AI tools, tutorials, and automation workflows for people who want real results, not another listicle. Certified in Microsoft AZ-900, CompTIA Security+, and AWS AI Practitioner, with 10+ years in enterprise IT.
→ Connect on LinkedIn · More about Shahid · Latest posts
Top comments (0)