Automation consultant metrics: what to put in front of clients at renewal.

Most consultant dashboards measure the wrong things.

Walk into a quarterly review with a dashboard showing every workflow, every execution, every node — and you've handed your client a spreadsheet. They don't want a spreadsheet. They want one sentence: this is working, and it's worth what you're paying me.

The metrics that show up in most automation reports — total runs, nodes executed, MB processed — are activity metrics. They measure how busy the system is, not whether it's producing value. Activity metrics scale with traffic; nobody's ever renewed a retainer because workflow runs went up.

Renewals get won by outcome metrics: did things stay up, did they break in ways that mattered, and how much human time got saved. Five concrete ones make the case.

1. Uptime per critical workflow

Not n8n's instance uptime. Per-workflow uptime, scoped to the workflows that touch revenue or compliance — billing syncs, lead routers, inventory updaters, anything where downtime has a dollar value.

Express it as a percentage over the renewal window (last 30 / 90 days). Anything below 99% on a revenue-critical workflow is a conversation. Anything above 99.5% is the headline number you lead with. Show three workflows, not thirty.

2. Mean time to detection (MTTD)

When something breaks, how long is it broken before anyone — you or the client — knows about it? Pre-monitoring, this is days. With basic alerts in place, it's minutes. The delta is dramatic and easy to communicate.

MTTD wins renewals because it answers the unspoken client question: "what if you weren't paying attention?" A consultant who can point at "average 6 minutes from failure to alert" has answered it. A consultant who finds out a workflow broke when the client emails them has not.

The MTTD-vs-MTTR pairing

Mean time to recovery (MTTR) is the natural pair. If detection is fast but recovery is slow, the conversation is "we need a runbook." If detection is slow, the conversation is "we need better monitoring." Show both and the client gets a clear read on where to invest.

3. Time saved per month

The hardest metric to estimate, the most valuable one to report. Take every workflow currently in production, estimate how long the task wouldtake a human if it weren't automated, and multiply by run count.

Two rules: (a) use conservative estimates — 30 seconds, not 5 minutes, per row processed — because you want this number to survive scrutiny. (b)Don't restate the same hours every month. Show the trend: last quarter we saved 120 hours, this quarter 145 because we shipped two new workflows. Growth in time saved is what justifies a retainer increase.

4. Error rate, weighted by workflow importance

Raw error rate across the whole instance is noise. A debug workflow that fails 30% of the time mixed in with a billing workflow that fails 0.1% averages out to something meaningless.

Weight errors by the workflow's business impact tag (use 1 for low, 3 for medium, 10 for high — pick a scheme and stick with it). Now the number means something: errors trending up on tagged-as-high workflows is a problem; errors trending up on debug workflows is maintenance noise.

Recovery-adjusted error rate

Even better: subtract errors that auto-recovered on retry from the total. A workflow that fails once, retries, and succeeds isn't a failure from the client's perspective. That math turns a scary-looking 4% error rate into a calm 0.6% effective failure rate.

5. Automations shipped per quarter

Net new automations is the metric that prevents retainers from getting downsized. If the count is flat, you're in maintenance-only mode and the client will eventually wonder why they're paying you. If it's growing, you're delivering ongoing value.

Pair it with one line of context: "Shipped 4 new workflows this quarter — payroll reconciliation, refund handler, two reporting digests — net 31 hours/month saved." That one sentence is the retainer renewal.

What to leave out

Three metrics that sound impressive but don't belong in client reports:

• Total executions. Activity, not outcome. Trends with traffic, not value.
• Total nodes / total credentials / total integrations. These describe the size of the system, not its performance. Clients don't care.
• Data processed in MB.Unless you're billing on volume, it's a vanity metric.

How to surface them without writing reports by hand

The reason most consultants don't report these numbers is the same reason they don't track them: they'd have to query n8n's database, calculate the deltas, format the output, and rebuild it every month. That's an afternoon of work per client per renewal.

The fix is to derivethese metrics from the workflow inventory + execution history automatically. Workflow inventory gives you the "which workflows are critical" tag. Execution history gives you uptime, error rate, and run frequency. Alert dispatch logs give you MTTD. Time saved is the one estimate you have to write down once per workflow — after that, multiplication handles the rest.

WorkflowRadaris the tool we built to do exactly this. It reads your client's n8n instance, builds the inventory, scores per-workflow health, tracks alert dispatch, and generates the four-page PDF every quarter — branded with your logo, not ours. You hand it to the client; they renew.

The deeper move

You're not in the n8n maintenance business. You're in the "client doesn't have to think about this" business. The metrics above let you prove that — concretely, in writing, every quarter — so the renewal conversation never becomes a renegotiation.