n8n Insights: measuring the performance and time saved by your workflows

How much time do your n8n workflows actually save your team? How many executions failed last week, and on which workflows? Until recently, answering these questions meant hand-rolling your own queries against the execution history. In recent versions, n8n ships Insights, a native dashboard that aggregates these numbers with no external setup. This guide covers what it measures, how to configure it, and where its limits sit depending on your plan.

What n8n Insights actually is

Insights gives instance owners and admins a numeric view of how workflows behave over time. The feature comes in two complementary forms:

• A summary banner, shown above the workflow list, displaying key metrics for the last 7 days (total executions, failures, estimated time saved).
• A detailed dashboard, with per-workflow charts and historical comparisons, useful for spotting a gradual degradation or a spike in failures on a specific workflow.

Unlike technical instance monitoring, which answers "is my server still standing?", Insights answers a business-facing question: "which workflows run the most, which ones fail, and how much human time does this pipeline actually save?"

Banner vs. dashboard: what depends on your plan

The two tiers of Insights aren't available the same way everywhere:

Element	Access
Summary banner (last 7 days)	Every plan, including the free self-hosted Community Edition
Detailed dashboard with history	Paid plans only (Pro and Business on n8n Cloud, Enterprise license self-hosted)
30-day retention	Business plan
365-day retention	Enterprise plan

In practice, a free self-hosted instance keeps a useful day-to-day snapshot (the banner), but loses history as soon as you want to compare one month to another or justify an ongoing investment in automation. If your usage has moved past the test stage and you need to present numbers to leadership, see our n8n Cloud pricing comparison to see where this tier sits relative to other paid features (separate environments, Git-based workflow versioning).

Configuring time saved per workflow

The most useful metric for justifying automation to leadership isn't the raw execution count, but the estimated time saved. n8n offers two calculation methods, configurable per workflow in its settings:

• Fixed time: you set a number of minutes saved on every execution of the workflow, regardless of the path taken. Simple to configure, well suited to workflows with a uniform flow.
• Dynamic time by execution path: you attach a different estimate depending on the branch actually reached (a specific node, an output tag). More accurate for a workflow whose branches carry very different human cost — for example an email-sorting workflow where only the "urgent, AI-drafted reply" branch replaces real writing work, unlike the "newsletter, auto-archive" branch.

n8n then multiplies that estimate by the number of production executions over the chosen period, and displays the total in hours or person-days on the banner and dashboard.

Example: quantifying the Inbox AI Pack

On an email-sorting workflow similar to the Inbox AI Pack (€79), a realistic approach is to set a dynamic time: 3 minutes saved for an email classified and auto-archived, 8 minutes for an email where the AI also drafted a reply that's approved in a few seconds. Across 150 emails processed per day, Insights then computes a directly readable weekly total — a far more compelling number internally than a bare "the workflow ran 750 times this week."

What Insights doesn't replace

Two mix-ups come up often:

Insights is not a technical monitoring tool. It says nothing about CPU load, available memory, or queue state on an instance running in queue mode. For that layer, the Prometheus /metrics endpoint (enabled via N8N_METRICS=true) remains the right tool, as covered in our self-hosted n8n monitoring guide.

Insights doesn't measure AI call costs. A workflow can show an excellent time-saved figure while burning a disproportionate API budget on expensive models. To cover that dimension, pair it with dedicated tracking like the one described in our guide to tracking AI call costs — the two dashboards together give a real cost/benefit view per workflow.

A useful caveat: time saved is an estimate, not a measurement

The number Insights displays depends entirely on the estimate entered up front — and human estimates of time saved carry a documented bias. A landmark study in decision psychology, Buehler, Griffin and Ross (Exploring the Planning Fallacy: Why People Underestimate Their Task Completion Times, Journal of Personality and Social Psychology, 1994 — see on Google Scholar), shows that people systematically underestimate how long tasks actually take, even when they have comparable past experience to draw on. Applied to Insights, that suggests a simple rule: err on the side of caution when entering time-saved figures rather than optimism, and periodically cross-check the estimate against a real stopwatch run on a sample of equivalent manual tasks.

More broadly, the difficulty of measuring the real payoff of a digital tool isn't unique to n8n: economist Erik Brynjolfsson, in his foundational article The Productivity Paradox of Information Technology (Communications of the ACM, 1993 — see on Google Scholar), documented the persistent gap between companies' technology investment and measured productivity gains at the aggregate level. A dashboard like Insights doesn't magically solve that problem, but it brings the measurement down to the individual workflow, where it stays actionable — unlike a macroeconomic statistic.

Going further

Insights fills a real gap between "the workflow runs" and "the workflow pays off": the banner is enough for informal day-to-day tracking, and the paid dashboard becomes worthwhile once you need to justify an automation budget over several months. For workflows designed from the start with clear measurement points — volumes processed, human escalations, logged AI decisions — the FlowKit packs (Inbox AI Pack at €79, Assistant RAG Pack at €119, Compliance & Audit Pack at €149) provide an already-structured base: just plug Insights in and let the numbers speak.