Field notes for a calmer, more capable one-person business.

Practical Solo Ops
Menu

Open testing method

Automation Platform Comparison Test Method

A repeatable way to compare setup, permissions, maintenance, failure recovery, portability, and exit work—before publishing a recommendation.

Current testing status

Method published. Product results not yet tested.

Practical Solo Ops has not yet run this protocol inside Zapier, Make, or n8n accounts. No platform has a score, ranking, or recommendation. Every included identity, address, request, and outcome is synthetic.

One bounded workflow, five cases

Editorial test design. Each platform receives the same synthetic client-intake workflow: validate an incoming record, require consent, stop duplicates by request ID, write one accepted record, and issue one notification. Five cases cover a valid request, missing consent, a duplicate, a temporary destination failure, and malformed input.

The harness specifies observable outcomes rather than platform-specific modules. A tester may use native steps, but must record every connection, permission, paid-plan dependency, custom-code step, and deviation.

Record evidence before scoring

  1. Record the date, region, plan or edition, visible version, and workflow version.
  2. Capture setup time and every permission or connection request without recording secrets.
  3. Run the five cases in the published order and record observed writes, notifications, status, and run ID.
  4. For the temporary failure, document visibility, recovery options, retry scope, and whether recovery duplicates prior work.
  5. Export the workflow, inspect it for identifiers or secrets, import or restore it in an isolated destination, and record gaps.
  6. Map every deletion surface: workflow, connections or credentials, webhooks, stored data, history, and account data.

Score only observed behavior

Do not convert missing access into a zero. Mark it Not tested or Plan-limited. Narrative evidence comes first; any later score must trace to a run log, screenshot, export, or dated official limitation. Setup speed cannot erase unsafe permissions, duplicate recovery, or an unusable exit.

What official documentation establishes—and does not

Sourced facts. Zapier documents account-data export and deletion, replay behavior, and workflow JSON import/export for Team and Enterprise accounts. Make documents JSON scenario blueprints and separate deletion surfaces for scenarios, connections, webhooks, keys, devices, data stores, and the profile. n8n documents JSON workflow export/import, execution-data retention controls for self-hosting, and workflow permissions.

Untested boundary. Documentation proves that controls are described, not that they are easy, complete, reliable, or available in the plan a solo professional would choose. Those conclusions remain blocked until the owner creates or authorizes product accounts and the same test pack is run. No credentials should ever be sent in chat or stored in the evidence files.

Download the test pack

The JSON files define inputs and expected outcomes. The CSV files record one run at a time and the final evidence summary. Blank fields are intentional; they must not be filled from marketing claims.

Reference verification. The downloadable reference result proves that the five fixtures and expected-outcome rules are internally reproducible. It is a local model result, not a Zapier, Make, n8n, or other product test.

Official sources reviewed August 25, 2026

Future commercial boundary. If affiliate relationships are later added, the relationship will be disclosed clearly near the recommendation and link. Compensation will not substitute for testing or determine the conclusion.