Hi all, founder here (disclosure: I built this, it has a free tier, and the node is MIT licensed).
What it is, in one line: a node you drop at the end of a workflow that records the execution in LumaTrack, a ledger that prices each run against a manual baseline you define (minutes a human would have spent, at a loaded rate) and turns the result into an auditable dollar-value report you can hand to a boss or a client.
The honest positioning first, because you already have Insights. n8n’s native Insights is good, and if what you want is “how much time are my workflows saving on this instance,” use it, it is built in and I will not pretend otherwise. What Insights gives you is minutes, per workflow, on one instance (plus the Time Saved node for dynamic per-path values). What it does not do, and what this node is for:
- Dollars/Pounds/Euros, not just minutes. A baseline of minutes x a loaded rate, with the productivity figure discounted by a visible conservatism factor (x0.50 by default) so the output is a number a skeptical finance person would have computed anyway. Minutes are an ops metric; budgets are argued in currency.
- Across every tool, not just n8n. The same ledger takes runs from Ansible, PowerShell, cron, CI, AI agents, anything that can send one HTTP call. If n8n is 60% of your automation estate, an n8n-only number undercounts you by the other 40%.
- Per client. If you run n8n for clients (MSPs, agencies), each client gets an isolated org, a white-label report the product mints as an “Executive Summary (QBR)”, and a monthly value-pack email for their stakeholders (each recipient can opt out). Insights has no per-client rollup.
- Client-facing and frozen. Reports are shareable by link, every figure drills down to the runs and the assumptions behind it, and closed months freeze so the number you showed in Q1 is the same in Q3. Failures are recorded and priced as costs, not skipped.
What the node does mechanically: one run event per execution. Fields: automation slug, status (success/failure), duration, optional metadata. External ID defaults to the n8n execution id, so retries are idempotent automatically. It honors Continue On Fail, so a LumaTrack outage never breaks your workflow. Wire the error path too: a LumaTrack node with Status = Failure in your error workflow is half the point, because failed executions cost money and save nothing, and a report that admits that is the one people believe.
Setup: Add a LumaTrack node to your n8n workflow, create a free LumaTrack workspace (5 automations, 25,000 run events/month, no card), mint an API key, add the credential. First dollar number in about 15 minutes. Docs: LumaTrack Developer Docs (the n8n page has the field-by-field reference).
Honest limits, so nobody is surprised: the run counts and outcomes are measured (they come from executions, auditable back to the event). The baseline is your assumption, and LumaTrack labels it as an assumption, grades its evidence, and discounts it rather than pretending to have measured it. If nobody in your org has to justify automation spend to anyone, you do not need this; Insights alone is plenty. This is for the moment someone asks “what is all this n8n stuff actually worth,” and you want an answer with the work shown.
Free tier is genuinely usable (it is the whole single-team story, not a trial). Feedback on the node, the fields, and what is missing is exactly what I am here for.