PRODUCT TEST / N8N / 17 NOV 2026
BEST FOR HANDS-ON BUILDERSAI can build the flow; evidence still needs scrutiny.
The Assistant built and revised a search, fetch and summary workflow. An earlier Canva run produced a report that warned that only two of six pages were readable. A subsequent revised run was started but its final result was not verified.
A judgment of the observed journey, not a measured product-wide rating.
THE SCORE, EXPLAINED
3/5 reflects powerful AI-assisted construction and an explicit evidence warning, with meaningful source and output friction. It is a scoped editorial judgment, not a reliability or product-wide rating.
Scale: 1 = severe friction in the observed task; 3 = useful but meaningful setup or verification work; 5 = clear, smooth task experience with a verifiable outcome. We did not measure total completion time or test equal end-to-end scope.
THE JOB
Ask the product's AI to build a workflow that accepts a product name, finds public user reviews, summarizes the main pros and cons and produces a report with source links. We used Canva as the example. The task is a test of n8n, not a review of Canva.
WHAT WORKED / WHERE IT WAS HARD
What worked
- The Assistant generated and revised a multi-step workflow from the requested task.
- The earlier report explicitly warned when evidence was limited, instead of presenting inaccessible pages as readable.
- Nodes and output can be inspected, helping a hands-on builder understand what happened.
Friction
- Several review sites returned bot checks or no readable content. The earlier report relied on one review aggregation page and one editorial review.
- A subsequent initial attempt returned a truncated report and included irrelevant material as user review evidence; the Assistant attempted a fix. The final retest result was not captured.
- Trial assistant credits decreased substantially during a broader interval of conversations, modifications and tests; the exact cost of one run cannot be inferred.
THE TEST JOURNEY
- Requested a natural-language workflow to search user reviews, summarize pros and cons and produce a sourced report.
- The assistant generated a search, page-fetch, summary and report workflow.
- An earlier live run returned a limited-evidence report; it used two accessible sources from six selected pages.
- Requested separation of first-person comments and editorial sources. The first revised run had problems; a later rerun started but its final outcome was not verified.

THE BIGGEST IMPROVEMENT
Help builders distinguish accessible review evidence from search results and unrelated content.
Recommendation: Surface accessible-source counts and source type in the report, show the content passed to the model, and ensure output truncation is visible.
Tradeoff: Some third-party sites restrict automated access. A warning is valuable, but it cannot replace missing source material.
PRICE, CREDITS & TIME
n8n Cloud’s free trial lists 800 Assistant credits plus 100 AI credits. An earlier balance of 544/800 and later balance of 11/800 show a decrease of 533 over an interval containing multiple activities; this is not the cost of one run. Gateway was observed at $2.00 and later $1.91; attribution was not reconciled. Total time was not reliably measured.
n8n pricing and trial allowance ↗ · Plan information checked 17 November 2026; terms can change.
WHAT THIS TEST DOESN'T PROVE
One controller-led trial workflow with limited accessible sources. The final subsequent retest, source-by-source accuracy, complete timing and repeatability were not verified. Not a three-persona test.
Evidence-led product testing · This page describes observed actions and limitations. No fictitious persona quotations or unmeasured timings are presented as test results.