Field notes on what it takes to run more. Every claim is sourced; where we sell something, we say so.
Per-client WABAs, template approval queues, the 24-hour window: the operational setup agencies discover after the first integration.
Activepieces vs n8n for client work: verified pricing, isolation and white-label tiers side by side — and the $36,000/yr gap nobody mentions.
7 min read
The three real routes to white-label n8n — embed licence, client-owned instances, or a control plane — and what each actually costs.
7 min read
The ceiling isn't a client count, it's a time budget: 2-3 days to onboard, 1-2 hours a month each. The arithmetic, and what moves the number.
5 min read
Hosting runs €4-25/mo — it was never the expensive part. The real cost structure of running n8n for clients, and what to price into a retainer.
5 min read
A €4 Hetzner box runs n8n fine. What changes with isolated instances per client, where the ops burden shows up, and what breaks first.
6 min read
Why one n8n instance per client beats a shared instance — and the tracking, billing and OAuth overhead isolation forces you to solve.
5 min read
n8n's Sustainable Use Licence permits internal use but restricts hosting it for clients or white-labelling. What that means for an agency, in plain terms.
5 min read
Shared instance, projects, or instance-per-client? The four isolation patterns for running n8n across multiple clients, honestly compared.
6 min read
Loud errors wake you up. Quiet failures just stop — and keep not running while you invoice. Why silent workflow failures cost retainers.
5 min read
Sourced retainer ranges for automation work — what agencies actually charge per client per month, and the arithmetic behind the numbers.
5 min read
GoHighLevel is a CRM with automation attached; n8n is an automation engine with no CRM. Why many agencies end up running both, and what that hybrid stack costs.
5 min read
The research, the numbers, and what we're building. No digest, no roundup.