Automation Audit: Find the Zaps and Scenarios Nobody Owns

List every Zap and Make scenario in one spreadsheet. For each, record who owns it, what triggers it, which apps it touches and whose login each connection uses, when it last ran successfully, how often it errors and what it costs in tasks or credits. Then give every automation a named owner, fix it, or switch it off.

For a business with 20 to 40 automations, allow two to four hours. The result is an inventory, a decision for each automation, and a few habits that stop orphaned automations piling up again.

How automations end up with nobody in charge

Almost every small business that has used Zapier or Make for more than two years has a few automations nobody fully understands. The pattern is predictable:

• The builder left. An office manager or a keen freelancer set things up, then moved on. The automations kept running because nobody touched them.
• Connections use personal logins. The automation reads a shared inbox or writes to your booking system, but it signs in as a named person's account. When that account is closed or its password changes, the connection expires and the automation stops.
• Errors go to the wrong inbox. Zapier sends error notifications to the account email by default. If the owner has gone, the warnings land in a mailbox nobody reads.
• Automations switch themselves off. Both platforms deactivate an automation that keeps failing. From the outside, nothing looks wrong; the reminders just stop.
• Nobody knows what it's for. An automation called "New Zap (3)" might be vital or pointless, so nobody dares switch it off.

The inventory: one row per automation

Start a spreadsheet with these columns. You'll fill most of them from the platform; the last few need a conversation with whoever uses the output.

Filled in, two rows from an illustrative six-person events-catering company look like this (shown sideways, one automation per column):

Neither automation was broken, and both were at risk. The deposit Zap would have stopped the day the coordinator's mailbox was deleted, and the enquiry scenario would have stopped when the freelancer cancelled his card, with allergy details passing through an account the business didn't control in the meantime.

Where to find the facts in Zapier

• Zap history. The Zap runs view can be filtered by owner, status, app, folder and date. Filter to errored runs first; that's your repair list. Zapier's help centre says it only guarantees around 60 days of run data, so export anything you want to keep.
• Task usage. A separate task usage tab shows billable tasks per workflow over a date range. Copy the monthly figure into your sheet. Remember that only successful action steps count as tasks; triggers and filters don't.
• App connections. The connections page can be filtered by status (active or expired) and by owner, and it shows who else has access to each connection. Every expired connection is a Zap that has stopped or soon will.
• Change owner. From the menu next to a Zap, choose "Change owner" and pick another user in your account. On Team and Enterprise accounts, removing a member lets you transfer all their Zaps and app connections to someone else; the Zaps keep their on or off status.
• The webhook catch. Zaps that start with a Catch Hook trigger receive data at a unique web address containing the owner's ID. When ownership changes, that address changes too, and you must paste the new one into whatever app sends the data. The tutorial on what a webhook is explains why.
• Error notifications. Choose how often errors are emailed, from "Immediately" through "Hourly summary" to "Never". On Team and Enterprise accounts, owners and super admins can set up custom notifications for any Zap in the account, which is how you make sure a current person hears about failures.

Where to find the facts in Make

Make is organised differently, and that changes where the orphan problem hides.

• Scenarios belong to teams, not people. Connections, scenarios, webhooks and data stores always sit in a team. That's safer than personal ownership, but it moves the risk into the connections: a team connection still signs in as whoever authorised it.
• Check team roles. Roles include Team Admin, Team Member, Team Operator and Team Monitoring. Leavers should have their role set to none, which removes them from the team.
• Incomplete executions. If a scenario has "Store incomplete executions" switched on, failed runs pile up in its incomplete executions tab. A long list there means a scenario that's broken but still looks active.
• Errors before deactivation. This scenario setting decides how many errors in a row Make tolerates before it switches the scenario off. Note which of yours are already off and why.
• Email preferences. Error and deactivation emails are set per person, per organisation, in each user's profile. Make sure at least one current person receives them for every team.
• Credits. Make bills in credits, from about $9 a month for the entry paid plan. Note each scenario's usage so you can see which ones cost the most.

The audit checklist

Once the sheet is filled, work down this list. Each item says what to check and how to confirm it.

Ownership

• Every automation has a business owner who still works for you. Confirm by asking them to explain what it does in one sentence. For a Zap nobody can explain, an AI assistant can help you read it. Copy the step list (app, event and the key fields of each step, with no customer data from the run history) and ask: "In one sentence, what does this automation do, who receives its output, and what would stop working if it were switched off?" An illustrative answer for a florist's mystery "New Zap (3)": "When an order email arrives, it adds the order to the delivery sheet and sends the customer a confirmation." Close, but wrong in the part that matters: the final Gmail step's "To" field was the delivery driver, not the customer, so switching it off would have left the driver without a route list. Treat the AI's sentence as a draft and check every recipient field yourself.
• No connection runs on a leaver's login. Check the connection account column against your staff list.
• Shared logins used by automations are stored in a password manager, not in one person's head. The approach in sharing tool logins safely with a password manager works for automation accounts too.

Health

• The last successful run fits the schedule. A daily automation that last succeeded three weeks ago has stopped.
• Errors are rare. My working threshold is fewer than one failed run in fifty over 30 days; anything above that goes on the fix list. A Zap that ran 420 times with 14 errors is at one in thirty, so it qualifies. Open the errored runs before fixing anything: in a typical case like that, all 14 might say "row not found" and all fall on Tuesdays, the day someone re-sorts the sheet the Zap writes to. The fix is a lookup on a booking reference rather than a row number, not a reconnection.
• A current person receives error emails for every automation. Confirm by checking notification settings, not by assuming.
• Every automation that's switched off is off on purpose. If nobody knows why, find out before turning it back on, because it may have been failing.

Data

• Every app that receives customer data is one you'd approve today. Old automations often feed a spreadsheet or a free tool nobody would sign off now.
• AI steps are on business accounts. If a Zap or scenario sends text to an AI model, note what data goes in and whose account or API key pays for it.

Cost and continuity

• The three biggest task or credit users are justified. One chatty automation can push you into a higher tier; see when Zapier gets too expensive.
• Each automation has a description in its notes or description field saying what it does and who owns it.
• Every webhook address is recorded alongside the app it's pasted into.

Triage: adopt, fix, merge or retire

Worked example: a hearing-aid practice with 33 automations

Take an independent hearing-aid practice with two audiologists and four admin staff that finds 27 Zaps and 6 Make scenarios, most built by an office manager who left four months earlier. These numbers are illustrative. The audit takes one afternoon and finds:

• Nine automations connected through the former manager's own email account, three of them already expired.
• Four Zaps switched off by Zapier after repeated errors. One of them sent the two-week follow-up booking message to new hearing-aid wearers. It had been off for seven weeks, and nobody noticed until patients started calling to ask when their check was.
• Three separate automations sending review requests, so some patients got two.
• Six automations with no runs in 90 days.
• One scenario sending full appointment notes to an AI summarising step on an API key registered to the former manager.

The decisions: adopt 14, fix 5, merge the three review requests into one, retire 9, and stop the AI summarising scenario that day, pending a rebuild on a business account with notes stripped of anything clinical it doesn't need. Monthly task use falls by about 40%, which is worth checking against the tier prices on Zapier's pricing page before the next renewal. The follow-up message is the lesson: the most important automation in the account was the one that had failed silently. The tutorial on stopping automations breaking silently covers the alerts that would have caught it.

Switching off safely: the two-week off test

1. Turn the automation off. Don't delete it.
2. Write the date and the reason in your sheet, and tell the team which automations are off.
3. Wait two weeks, or one full cycle for anything monthly, such as an invoice run.
4. If nobody has noticed, delete it. Zapier keeps deleted items in its trash for 30 days. In Make, export the scenario first (three dots, then "Export blueprint") so you have a copy; blueprints keep modules and settings but not connections, which you'd have to recreate.
5. If someone does notice, you've just found its business owner.

Step 5 happens more often than you'd expect. Imagine a garden centre that switches off an unnamed Zap copying web-form entries into an old spreadsheet, apparently a leftover from a retired newsletter. On day nine the part-time bookkeeper asks why the monthly gift-card sales summary she downloads from that spreadsheet is empty. The Zap was doing a second job nobody had written down. It gets a proper name, a description and her as its business owner, and a shorter-lived test would have deleted it before anyone noticed.

Stopping the problem coming back

• Name automations properly: "Bookings: new online booking to confirmation SMS (owner: front desk lead)" instead of "New Zap (3)".
• Put automations on the leaver checklist. Before anyone leaves, transfer their Zaps and connections and reassign anything that signs in as them. The staff offboarding checklist for AI tools and shared accounts has the full list.
• Connect through business accounts where your apps allow it, with the login stored in a password manager, rather than through whoever happened to set it up.
• Review quarterly. Thirty minutes: sort the sheet by last successful run and errors, and check the connections page for anything expired.

Further reads

• Zapier vs Make vs n8n for AI Automation: Which Fits Your Business? — Worth a look if the audit shows you're on the wrong platform.
• How to Add AI Steps to Zapier: Classify, Summarise, and Draft — Add AI steps properly once the basics are owned.
• Who Should Own AI in a Small Business? Roles and Responsibilities — Decide who holds the automation owner role.
• How to Build Your First AI Automation in Make, Step by Step — Rebuild a retired automation cleanly in Make.
• Native Integrations vs Zapier: Which Should Connect Your Apps? — Some Zaps are better replaced by a built-in link.
• AI Governance for a Small Business: Who Decides, Approves, Checks — Where automation ownership fits in wider AI rules.
• AI vs Rule-Based Automation: Which Does Your Task Need? — A simple test for choosing rules, AI or both, a side-by-side comparison, the hybrid pattern that keeps AI small, and a dry cleaner's jobs sorted.
• How to Scale AI From One Workflow to the Whole Business — How to copy your first working AI workflow across the business: a readiness gate, a playbook template, adjacency rules and a pet shop's nine months.
• How to Choose and Support an AI Champion in a Small Team — A weighted scoring sheet for candidates, a one-page remit template, a monthly check-in agenda and a music school example with time costs.
• AI Incident Response Plan for Small Businesses (With Template) — A two-page AI incident plan for small teams: what counts as an incident, who leads, severity levels, first-hour actions and a template to copy.
• How to Audit Your AI Subscriptions and Cut Wasted Spend — Find every AI charge, check real usage in each tool's admin reports, cut overlaps, and time changes to seat-reduction rules, with an architect-practice example.
• AI Adoption Stages: Where Is Your Business Now, and What's Next? — Place your business on a five-stage AI adoption scale with a ten-question check, then see the one move that gets you to the next stage.
• Is AI Too Complicated for Non-Technical Business Owners? — Four levels of AI use, from typing into a chat box to custom builds, with honest learning times and five tests for when to bring in help.
• Common AI Mistakes Small Businesses Make and How to Avoid Them — Twelve everyday AI mistakes in small firms, grouped by data, output, money and ownership, each with how to spot it and fix it in under an hour.
• Automating a Broken Process: Why It Backfires and What to Fix — Why automating a broken process backfires, the five-pass fix to do first, and how to split the cleaned-up process between rules, AI and people.
• How to Stop Retyping Data Between Apps With AI Automation — Find every place your team copies data by hand, match each to the cheapest fix, and use AI extraction only where the source is an email, PDF or free text.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Zapier help centre articles on Zap history, task usage, app connections, changing Zap ownership, removing team members and error notifications; Make help centre pages on teams and roles, scenario settings, incomplete executions and scenario blueprints.