A new client signs on Friday and wants “everything managed” by Monday. The temptation is to push agents everywhere at once and sort out the mess later. Onboarding has a correct order — structure before agents, baseline before patching — and following it is the difference between a clean start and a month of apologies.
Onboarding in Endpoint Central MSP is a sequence, and each step makes the next one safe. Create the customer and remote offices first so every agent lands in the right container. Decide direct or distribution server per office before rollout. Push agents in waves, take a baseline before you change anything, and only then switch on patching, policies and reports. Here is the order, with the reason for each step.
Install the agent on a pilot group of three to five machines first — one ordinary desktop, one laptop, the accounts PC if they agree. Confirm they appear in the Scope of Management under the right office, report inventory correctly and answer a remote session. Only then roll out to the rest, office by office. A pilot wave catches the site-specific surprises — an aggressive antivirus, a proxy, a naming mess — while the blast radius is five machines and a phone call, not eighty machines and a crisis.
Run full inventory and patch scans and save the results as the client's baseline — day-zero truth. This document protects everyone: pre-existing problems (the already-failing disk, the two-year patch backlog) are recorded as inherited, not caused by you. Share the baseline summary with the client in the first week. It sets honest expectations and quietly demonstrates why they hired you — most clients have never seen their own network counted before.
Start patch management with the client's test group and Ring 1 rhythm rather than a heroic catch-up weekend — a two-year backlog installed in one night is how onboarding becomes an incident. Layer configurations next: security baseline, then convenience items. Bring the distribution server live before the first big software deployment if the site needs one. Each layer lands, soaks and is confirmed before the next — boring, and exactly what a client is paying for.
See Endpoint Central MSP running live — customers, offices and agents on one screen, step by step.
▶️ Watch Full VideoThe console work — customer, offices, agent push, first scans — fits in a day for a typical small office. The honest schedule after that is a week of settling: pilot wave, baseline review, first ring of patching. Promising “fully managed by tomorrow” skips the steps that keep it safe.
No. Record it in the baseline, then clear it through the normal ring rhythm — test group first, production after. A two-year backlog installed in one night mixes dozens of changes into one incident if anything breaks.
Three things: name a contact person who can approve windows and reboots, provide an always-on machine if a distribution server is planned, and agree the maintenance window. Everything else is your team's work — that is the point of hiring an MSP.