Client Onboarding Checklist for Software Agencies (2026)

Quick Answer. A client onboarding checklist is the operational sequence an agency runs in the first ten business days after a contract is signed, before real delivery work begins. It moves a new client from signed paperwork to a working relationship: access and credentials collected, the kickoff run, scope confirmed against the statement of work, the communication cadence agreed, and the first milestone dated. Below is a five-phase checklist you can copy and paste, a worked example for a fixed-price AI app build, and the three onboarding failures that quietly cost agencies their margin. Updated July 2026.

The week after a client signs is the most underused week in most agencies. The deal is closed, the delivery team has not fully started, and the client is at peak trust and peak anxiety at the same time. That gap is where scope confusion, missing access, and unspoken expectations take root. None of them are visible yet. All of them surface later as rework, as a delayed launch, or as a renewal that never happens.

A good onboarding checklist does one thing: it converts a signature into a working system before the first line of delivery work is billed. It is the cheapest quality control an agency owns, because every hour spent here removes several hours of confusion downstream.

Why onboarding is a margin decision, not an admin task

Agencies tend to file onboarding under administration, somewhere between invoicing and password resets. That framing is why it gets skipped when the pipeline is busy. The reframing that matters: onboarding is the moment you either lock in the scope you sold or quietly inherit the scope the client imagined.

Three things are decided in the first ten days whether you manage them or not.

• Scope reality. The client signed a statement of work, but they remember the sales conversation. Onboarding is where you reconcile the two on the record, before either side has spent delivery hours on the wrong build.
• Access and dependency risk. Most delivery delays in the first month are not engineering problems. They are a missing API key, an unavailable stakeholder, or a repository nobody granted permission to. These are all knowable on day one.
• Communication expectations. How often you report, who signs off, and how change is requested are all cheaper to set as defaults now than to renegotiate mid-project when someone is already frustrated.

If you take one thing from the framing: onboarding is not the paperwork after the sale. It is the first delivery task, and it protects the margin on everything that follows.

The five-phase client onboarding checklist

The checklist below is organized into five phases that run across roughly the first two weeks. Phases one and two should be complete before the kickoff meeting. The kickoff itself is phase three. Phases four and five set the operating rhythm for the engagement.

Phase 1: Contract and administrative close

This phase confirms the deal is genuinely closed and billable before anyone schedules delivery work.

A common failure here is starting delivery on a signed contract with an unpaid deposit. The signature is intent. The cleared payment is commitment. Do not schedule engineers against intent.

Phase 2: Access, credentials, and dependencies

Nothing in this phase is glamorous, and every item in it is a potential two-week delay if it slips.

Collect these with a single request, not a trickle of emails. A short shared form or document with every required item listed lets the client hand off the whole list to their own team at once. Tools such as Notion are commonly used to host that intake document and track which items are outstanding.

Phase 3: The kickoff meeting

The kickoff is not a status meeting. It is the moment the whole delivery team and the client agree on what is being built and how they will work together. Keep it to sixty minutes with a fixed agenda.

• Restate the goal of the engagement in one sentence, and get the client to confirm it.
• Walk the scope from the statement of work out loud, item by item, so nothing lives only in the sales memory.
• Name what is explicitly out of scope. This single step prevents most later disputes.
• Agree the first milestone and its date.
• Confirm the communication cadence and the single point of contact on each side.

A well-run kickoff produces a one-page summary the same day. If you want a structured agenda to adapt, the public write-up from Asana on project kickoff meetings is a reasonable external reference for the meeting mechanics.

Phase 4: Communication and reporting cadence

Set the operating rhythm now, as a default, so it never has to be negotiated under pressure.

The change-request default matters most. When a client can request changes casually in a chat thread, scope drifts for free. Route change through a defined channel so each request is visible, estimated, and either absorbed or quoted.

Phase 5: Delivery setup and first milestone

The final phase turns the plan into a scheduled, resourced piece of work.

• Break the first milestone into tasks the delivery team can start on Monday.
• Confirm the milestone is staffed against real availability, not optimistic headcount. Your capacity planning template is the check here: do not commit a date the schedule cannot hold.
• Set the environment and tooling the client will use to review progress.
• Send the one-page onboarding summary: confirmed scope, out-of-scope list, cadence, first milestone date, contacts.

When phase five is complete, the client has moved from signed to started, and both sides are working from the same written picture.

Copy-paste onboarding checklist

Paste this into your project tool or client document and adapt the owners to your team.

CLIENT ONBOARDING CHECKLIST (first 10 business days)

Phase 1 - Contract and administrative close
[ ] Countersigned agreement filed
[ ] Deposit or first invoice paid and cleared
[ ] Statement of work version locked and confirmed in writing
[ ] Primary client contact confirmed (name, email, timezone)
[ ] Internal delivery owner assigned (one name)

Phase 2 - Access, credentials, dependencies
[ ] Repository and environment access granted and tested
[ ] Third-party API keys and secrets received and stored securely
[ ] Design assets and brand kit received (correct versions)
[ ] Analytics and hosting access verified
[ ] Decision-maker booked for kickoff

Phase 3 - Kickoff meeting (60 minutes)
[ ] Engagement goal restated and confirmed by client
[ ] Scope walked item by item from the SoW
[ ] Out-of-scope list stated explicitly
[ ] First milestone and date agreed
[ ] Communication cadence and single points of contact confirmed
[ ] One-page kickoff summary sent same day

Phase 4 - Communication and reporting cadence
[ ] Weekly written status day fixed
[ ] Live check-in scheduled
[ ] Change-request channel defined
[ ] Sign-off authority named per milestone
[ ] Escalation path shared

Phase 5 - Delivery setup and first milestone
[ ] First milestone broken into startable tasks
[ ] Milestone staffed against real capacity
[ ] Client review environment and tooling set up
[ ] One-page onboarding summary delivered

Worked example: onboarding a fixed-price AI app build

A four-person agency signs a fixed-price build for an internal operations tool with an AI assistant layer. Contract value is modest, timeline is eight weeks, and the client is a non-technical operations lead at a mid-sized company.

Phase 1 surfaces the first risk immediately. The agreement is countersigned but the deposit has not cleared, and the client contact is the operations lead, who cannot grant engineering access alone. The delivery owner holds the schedule until payment clears and a technical contact is named. Two days lost here save two weeks later.

Phase 2 is where the real dependency appears. The AI layer needs a model provider API key the client does not yet have, plus access to an internal data source behind the client's own approval process. Because this is caught on day two rather than in week three, the client starts their internal request early and the key arrives before it blocks anyone.

The kickoff reconciles a genuine gap: the sales conversation implied the assistant would handle three workflows, but the signed proposal scoped one, with the other two as a later phase. Stated out loud in the kickoff, this is a five-minute clarification. Discovered in week six, it would have been a margin-eating dispute.

By the end of phase five, the client has a one-page summary that says exactly what the eight weeks will deliver, what they will not, and when the first working milestone lands. The engagement starts on a shared picture rather than two hopeful ones.

Three onboarding failures that cost margin

• Starting delivery before access is complete. Engineers begin, hit a missing key or permission, and stall. The hours are burned but the meter of client patience keeps running. Complete phase two before scheduling delivery.
• Leaving scope in the sales memory. The most expensive disputes come from work the client was sure they bought and the agency was sure was a later phase. The kickoff walkthrough against the written scope is the cheapest insurance you will ever run.
• No default change channel. Without one, every casual request becomes unbilled scope. Set the channel during onboarding, before there is anything contentious to route through it.

None of these are exotic. All three are prevented by a checklist that someone actually runs.

If you take one thing from this: onboarding is the first delivery task, not the paperwork after the sale. Run it as a checklist and you protect the margin on everything that comes after it.