If you are still copying intake forms into Trello, pasting approvals into Slack and checking Stripe to see who has paid, you are doing this by hand. This page names the three handoffs worth automating first, what Zapier and Make cost at studio volume, and which decisions should stay with a person.
Automate a handoff you can already explain. "When a brief is ready, create a queued request" has a clear trigger and result. "Keep everything in sync" leaves unanswered which system wins when records disagree.
Three handoffs worth automating
Each of these moves data in one direction, from one system that owns the fact to one that displays it. That is why they are safe to start with.
- Intake form to Trello card. Trigger: a client submits the request form (an Airtable form or Zapier Forms both work). Action: create a card in your Intake list with the brief fields filled in. Failures go to whoever owns the queue that week.
- Approved card to client message. Trigger: a card moves to Approved. Action: post a short note in that client's Slack channel saying the request is done and what comes next. Failures go to the designer on the card, who tells the client by hand.
- Paid Stripe invoice to a note on the client board. Trigger: Stripe marks an invoice paid. Action: add a comment to the client's board so the team knows the account is current. Failures go to whoever runs billing, usually you.
Small and one-way is how the best-known solo studio did it. In his 2021 Indie Hackers AMA, Brett Williams of DesignJoy described running 30+ clients on Trello, with IFTTT passing records from Trello into Airtable. One trigger, one destination, nothing flowing back. You do not need more than that on day one.
What Zapier and Make cost at your volume
Will a free plan cover you? For one simple handoff, maybe. For three steps per request, probably not on Zapier. The two vendors meter differently: Zapier counts tasks and Make counts credits. Read each vendor's own definition before you compare them, because a Make credit and a Zapier task are not the same unit.
| Plan | Price | Monthly allowance | Catch for a studio |
|---|---|---|---|
| Zapier Free | $0 | 100 tasks | Two-step Zaps only, so a trigger plus one action |
| Zapier Professional (entry) | $29.99 a month billed monthly, $19.99 a month billed annually | 750 tasks | Multi-step Zaps; the price climbs on a task slider |
| Make Free | $0 | 1,000 credits | No time limit; runs at most every 15 minutes |
| Make Core | $9 a month as listed | 10,000 credits | Make shows a "Save 15% or more" annual option without a firm figure |
| Make Pro | $16 a month as listed | 10,000 credits | Same credit allowance as Core, more features |
Here is an illustrative example. Your studio receives 25 requests a week. Each one runs a three-step automation: create the card, post to Slack, copy a row to a log. That is about 108 requests a month, or roughly 325 action steps.
- Zapier Free runs out in the first week at 100 tasks, and it cannot run a three-step Zap at all. You would be on Professional at $29.99 a month, with 750 tasks leaving room to double your volume.
- Make Free offers 1,000 credits. Whether 325 steps fit depends on how many credits your scenario actually uses per run. Build it, run ten test requests and read the credit count in Make's history before you trust the headroom.
The default we would pick: automate the form-to-card handoff alone, on whichever free plan you set up first. Add a failure alert that reaches a named person before you add a second automation. Pay only when the monthly count tells you to.
Test the failure cases before you rely on it
What happens when an automation fails silently? A client submits a brief, nothing appears in Trello, and you find out three days later when they ask where it is. Run these four tests with fictional records before a real client touches the form:
- Duplicate event. Submit the same form twice and check that you get one card and one Slack post, not two.
- Missing field. Leave out a required field and confirm the run stops and alerts a named person instead of creating an empty card.
- Disabled connection. Disconnect Trello, submit a request, and check that someone hears about the missed run the same day.
- Manual recovery. Reconnect, then recreate the missed card by hand without replaying the ones that already went through.
Record the results, the owner of each alert and who can revoke each connection in the software evaluation and migration worksheet. Give every integration only the access it needs: a Trello connection that can create cards on one board does not need admin rights to every workspace.
Scope, disputes and billing exceptions stay with a person
Do not buy Zapier or Make to compensate for nobody owning the request queue. Automation moves a request into the queue; it does not decide what the queue works on next, so you still need a person who owns it. The requests and delivery guide sets out the request states worth automating between.
Keep a short manual checklist instead while your rules are still changing week to week. Some decisions stay with a person every time. They are a request that goes beyond agreed scope, an approval the client disputes, and a billing exception such as a pause mid-request or a failed card. A Zap that posts "paid" is fine. A Zap that decides whether to keep working for a client who has not paid is not.
Before you wire anything, ask whether the handoff should exist. If your project management tool already holds the request, a second copy in Airtable may be the problem, not the fix. The agency software stack guide helps you decide which system owns each record, and removing a duplicate often removes the handoff with it.