Your €12,000 retainer starts on the 1st.
On working day 15, the client finally sends the approved brief and grants access to the ad account. You now have five days to produce the work you planned across the month.
You can rush it, move it into month two or deliver less. Each option costs margin or trust.
The damage started before anyone missed a delivery date. It started when the agency planned work against a contract date while the client inputs had no owners, no deadlines and no agreed consequence for arriving late.
What is client onboarding for an agency?
Agency client onboarding runs from contract signature to delivery readiness. At that point, the team can produce the first agreed outcome without waiting on a critical client input.
It has five jobs:
- Transfer what was sold from the commercial team to the people delivering it.
- Define the first observable client outcome.
- Collect and verify the inputs required to produce that outcome.
- Agree who decides, who approves and how delays or changes will be handled.
- Give both sides one current view of the plan, owners and next actions.
“Kickoff held” and “form submitted” tell you that an activity happened. They do not tell you whether the team can start. At the other extreme, waiting for every asset needed across a year-long retainer creates an intake project nobody will finish. Clear the inputs on the critical path to the first deliverable.
The first approved deliverable is a separate milestone. It is the best practical test of the onboarding, but combining the two hides the source of delay. If readiness took four days and delivery took four weeks, you have a delivery problem. If readiness took three weeks, you have an intake or dependency problem.
Where does month-one margin go?
The retainer example leaves three choices:
| Response | Immediate effect | Likely consequence |
|---|---|---|
| Deliver less | Protect the team’s workload | The client sees a weak first month |
| Compress the work | Protect the promised date | Overtime or disruption to other accounts erodes margin |
| Move the deliverable | Protect quality and workload | Month two begins with month-one work still open |
The financial effect depends on the contract and whether you can move the team to other paid work. Measure the hours you could not reassign, the rework caused by incomplete inputs, and any credit or unpaid recovery work you gave the client.
A useful estimate is:
Cost of delay = unused reserved capacity + rework + unpaid recovery work or credits
Pull up your last ten new accounts. If you cannot find the date each one became ready for delivery, you cannot tell how much month-one capacity onboarding consumed.
What should happen from signature to first delivery?
Work backwards from the first client outcome and clear the dependencies on that path first.
| Stage | Agency action | Client action | Exit condition |
|---|---|---|---|
| Within one business day of signature | Complete the sales-to-delivery handoff and create the shared plan | Confirm the day-to-day coordinator | Delivery owner accepts the account and open assumptions are visible |
| Before kickoff | Pre-fill the plan, send the short critical-input request and test any access already granted | Invite the necessary approvers and start the critical inputs | Kickoff can be used for decisions rather than information gathering |
| Kickoff | Confirm outcome, scope, owners, approval path, cadence and delay rules | Correct assumptions and accept their actions | Every critical dependency has one owner and date |
| Days 2 to 5 | Verify inputs as they arrive and follow up on blockers | Complete or escalate assigned inputs | The first workstream passes its readiness gate |
| Remainder of month one | Produce, review and refine the first deliverable | Give consolidated feedback and approval | The first client outcome is accepted or a specific decision is recorded |
The timing above suits a straightforward retainer. A website migration or consulting project may need more time. Keep the sequence and set dates that match the work.
The sales-to-onboarding handoff guide covers the commercial context that should reach delivery before the client is asked to repeat it.
Ask only for what blocks the first deliverable
Ask first for inputs whose absence would stop the first deliverable or force you to rebuild it. Request the rest later.
Use three tests for each requested item:
- What work does this unblock? If nobody can name the work, remove it from the first request.
- Who can provide it? Assign a person, not “the client” or a department.
- How will we know it works? A permission is not complete until someone has tested it. A brief is not complete until the required decision-maker has approved it.
The critical inputs differ by agency type:
| Agency type | Likely critical inputs | Inputs that can often wait |
|---|---|---|
| Performance marketing | Campaign objective, budget authority, ad-platform access, analytics access, tracking status | Full asset library, long-term campaign calendar, optional channel history |
| Brand or creative | Approved brief, decision criteria, brand constraints, approver, source files needed for the first concept | Complete archive, future campaign variants, secondary templates |
| Web or development | Agreed first release, repository or CMS access, hosting constraints, technical owner, acceptance criteria | Backlog items outside the first release, non-critical content migration |
| Consulting | Decision to be supported, sponsor, initial dataset, interview access, review group | Broader research requests and stakeholders outside the first workstream |
| Recruitment | Role scorecard, compensation range, hiring owner, interview stages, approval to approach candidates | Employer-brand assets not required for the initial search |
Universal intake forms get bloated because every team adds its questions to the same document. Build one short base template, then add a module for each service. Show the client the smallest request that lets the team begin, and hold the rest back until the work reaches it. The client onboarding questionnaire guide sets out the three waves and what belongs in each.
Make each request hard to misunderstand:
| Input | Why you need it | Owner | Due | Complete when |
|---|---|---|---|---|
| Analytics access | Validate the conversion baseline | Priya Shah | 18 Aug | The analyst can open the property and export the required report |
Put these requests, the current plan, files and decisions in one client-facing place. Your production tasks can stay in the agency’s internal project tool. The client should not have to reconstruct the process from an email thread, a form, a Drive folder and a project board they cannot open.
Use automatic reminders for routine late tasks. A useful reminder names the missing item, its owner, the original date and the milestone it affects. Call the client when the blocker needs a decision, a security exception or a change in priority.
Settle five decisions at kickoff
Use the kickoff to close decisions and assign work. Send information beforehand so you do not spend the call reading the intake form aloud.
Agree these five things before the call ends:
- The first outcome. Name the first thing the client can review or use. “Start the account” is vague. “Approve the first creative direction” or “release the first landing page” can be checked. Write down what the client will use to accept it.
- The dependency list. Give every critical input one owner, due date and acceptance test. If three people could provide it, one person still owns getting it done.
- The approval path. Ask: “When we send the first deliverable, who must review it before you can approve it, and how much time does that review normally need?” Record the final approver, required reviewers and review window. Ask one client owner to consolidate comments.
- The working rules. Agree the meeting cadence, response expectations, primary channel and where final decisions will be recorded. State which requests count as urgent.
- The escalation and change rules. Name one senior contact on each side who can unblock a decision. Agree what happens when a client input is late and what triggers a change request.
The customer onboarding kickoff guide provides a complete agenda and the questions behind each decision.
Decide whether the work is ready to start
Use a readiness gate for each workstream:
- Green: the required input is present, approved where necessary and tested. Start.
- Amber: an input is missing, but the work can proceed without material rework. Start the independent portion and record the assumption.
- Red: a missing or unapproved input could change the direction or invalidate the work. Do not start that workstream.
An audit, competitor research, project setup or a technical environment check may be safe while another input is late. Producing the final concept from an unapproved brief is usually red. Record every amber assumption so the client can correct it before it hardens into work.
How should an agency handle client-caused delays?
Put the delay policy in the statement of work and restate it at kickoff. It should answer three questions:
- Does the affected delivery date move one-for-one with a late client dependency?
- How long will the agency hold reserved capacity before rescheduling the work?
- When does re-planning, rush work or remobilisation require a change request or additional fee?
Follow the same sequence each time:
- Before the due date, send the automatic reminder with the dependency and consequence.
- On the due date, confirm whether the owner needs help or a different route to completion.
- When the date is missed, update the affected milestone rather than leaving an impossible date in the plan.
- If the item remains blocked, involve the agreed escalation contacts and present options.
A useful message is direct and neutral:
Analytics admin access was due today and still blocks baseline validation. If access is available by Thursday, the first recommendation remains due on 27 August. After Thursday, we can either move the review by the same number of working days or keep the date by reducing the first review to channels with verified data. Please confirm which option you prefer.
This turns a hidden delay into a decision while both options are still workable. For projects that have gone quiet rather than slipped, use the guide to restarting a stalled onboarding.
Stop scope creep at its source
Separate incomplete inputs from changed inputs and new requests. They require different responses.
| Situation | Example | Response |
|---|---|---|
| Incomplete agreed input | The brief omits the target region | Return it to the owner with the missing decision specified |
| Changed input | The approved audience changes after concepts are built | Re-estimate the affected work and timeline |
| New request | A second channel is added during kickoff | Put it through the agreed change process before adding it to the plan |
| Agency misunderstanding | Delivery interpreted an ambiguous promise incorrectly | Correct it without presenting the correction as client scope creep |
State revision limits and exclusions aloud at kickoff. Attach acceptance criteria to the first deliverable and keep decisions in writing. Scope often expands through ambiguous approval and forgotten decisions long before a client formally asks to change the contract.
When does onboarding software become worth it?
Use the lightest system that keeps the process visible to both sides.
| Situation | Sensible setup |
|---|---|
| One respondent, a short one-time intake and low onboarding volume | Form plus a simple shared checklist |
| Detailed information or document collection is almost the entire process | Specialist intake and document-collection tool |
| Several stakeholders, a shared plan, recurring approvals and concurrent onboardings | Client onboarding workspace plus the agency’s internal delivery tools |
| Resource allocation, time, billing and margin must be managed together | Professional-services automation platform |
For the third case, Valuecase is the strongest default in our agency onboarding software comparison. It combines a branded, login-free client Space with tasks, forms, files, content, reminders and visibility across active clients, while the agency can keep its internal production system. If onboarding is only one short intake, that is more software than the process needs.
Coordination complexity matters more than headcount. A six-person agency onboarding five clients at once may need a shared workspace more than a 60-person agency with one standardised intake each month.
Measure where the time went
Track readiness and delivery separately. Four metrics are enough to diagnose most month-one problems:
| Metric | Calculation | What it diagnoses |
|---|---|---|
| Signature-to-ready time | Delivery-ready date minus signature date | Handoff, intake and client-dependency delay |
| Ready-to-first-delivery time | First delivery date minus delivery-ready date | Agency production speed |
| First-pass acceptance rate | First deliverables accepted without material rework divided by first deliverables submitted | Brief and approval quality |
| Onboarding effort variance | Actual onboarding hours minus planned onboarding hours | Margin leakage from coordination and rework |
Segment the results by service line. Comparing a website build with a paid-search retainer produces a number that explains neither. Within each service line, review the slowest three onboardings every month and look for the dependency they share.
The onboarding metrics guide covers the wider measurement model.
Start with one service line. Record the delivery-ready date for its last ten clients, inspect the slowest three, and fix the dependency that delayed all of them.