Question 1

What is n8n and when would you choose it?

Accepted Answer

Give a one-sentence definition, then explain your selection criterion. "n8n is an automation platform that lets you build a process as a visual workflow of nodes. I reach for it in two cases: first, when a process wires several systems together and writing a bespoke integration for each is expensive; second, when the logic is more than 'if this then that' — conditions, loops, data reshaping — so a tool like Zapier runs out of room." Then show the boundary: "I would not pick n8n for a high-throughput real-time service or a product with complex business logic — that needs a proper codebase."

Question 2

How do you assess whether a process is worth automating?

Accepted Answer

Walk through four criteria and tie each to an example. Frequency — something done several times a day pays back far better than something monthly Time — minutes per run × frequency = annual saving; compare that against the build cost Stability — automate when the steps are the same every time; if they change, stabilise the process first Cost of a mistake — if it is high, add an approval step or do not automate Strong close: "The biggest win is usually not the time saved but the latency removed — replying to a lead in 2 minutes instead of 4 hours moves revenue directly."

Question 3

If you introduced n8n to a team, which process would you automate first and why?

Accepted Answer

This tests strategy, not technique. A first workflow should be small, visible and low-risk. "I never pick the biggest problem first. My criteria: (1) the outcome must be reversible — a notification, not a payment; (2) the team should see the result daily so the value proves itself; (3) no more than one or two integrations. My typical pick is the daily report: pull the data each morning, write it to a sheet, post it to a channel. If it breaks, nothing is damaged, but everyone sees it every morning. Once that lands, you have the credibility to move on to more critical processes."

Question 4

How do you measure whether an automation succeeded?

Accepted Answer

Being able to talk about measurement is the middle/senior line. Show three categories: Process metrics — run time, latency (time to respond to a lead), number of remaining manual steps Reliability metrics — ratio of successful to failed executions, how often the error workflow fires Business metric — the outcome expressed in money or time "In practice I measure the baseline before switching the workflow on — average response time, say. Otherwise you cannot answer 'how much better is it?' afterwards. On the n8n side I watch the failure rate in the execution history; a rising failure rate is the first signal that something changed in an integration."

Area	Real automation example	What it saves
Sales / CRM	Lead from a website form → a card is created in the CRM → a Slack ping to the sales rep	Lead response time drops from hours to seconds
Finance / reporting	Every morning at 8, transactions are pulled from a bank API → written to Google Sheets → a summary email to management	The 30-40 minutes a day spent assembling the report by hand
Support / helpdesk	New ticket → AI assigns a category and priority → routed to the right team	Triage effort and mis-routed tickets
HR	New hire created in the HR system → Google account, Slack invite and equipment task are opened automatically	The forgotten steps in an onboarding checklist
Marketing	An RSS or social feed is watched → AI writes a summary → sent to Telegram for approval	The time spent monitoring content
IT / DevOps	A service health check every 5 minutes → on failure, a call/message to the on-call plus a Jira ticket	Time to detect an incident (MTTD)
Document processing	A PDF invoice arrives by email → text is extracted → AI fills the fields → written into the accounting system	Manual data entry and typos

What automation is and what you can automate

📚 Sources and documentation