Most onboarding advice is a checklist: welcome email, contract, kickoff call, shared folder. All necessary, none of it the reason engagements go wrong. The relationships that sour in month three were usually damaged before anyone sent a welcome email.
Onboarding starts during the sale
This is the part checklists can't fix, and it's where most of the damage originates.
During a pitch, everyone is optimistic. Timelines get described at their best case. A result achieved for one client gets mentioned in a way that sounds like a forecast. Scope gets discussed in conversational terms — "we'd look at your email too" — that never make it into the contract but absolutely make it into the client's memory.
Then delivery starts, and the work meets the actual constraints. Nothing has gone wrong yet, but the client is already comparing what's happening against what they heard. Onboarding is where that gap becomes visible; it isn't where it was created.
The uncomfortable rule Every expectation set during the sale is a promise delivery has to keep. If the person selling isn't the person delivering, that gap is where your churn lives.
The practical fix is unglamorous: whoever runs delivery reads the proposal and the sales notes before kickoff, identifies anything that can't be honoured, and resets it explicitly — in writing, at kickoff, while there's still goodwill. That conversation is awkward for ten minutes and prevents an argument in month three that costs you the account.
The scope conversation nobody has
Contracts describe what's included. Almost none describe what isn't, and nearly every scope dispute lives in that silence.
The client isn't being unreasonable when they ask you to look at the landing page, or write the email that supports the campaign, or join the call with their developer. Each request is small and sensible. Cumulatively they're an unbilled second engagement, and the moment you decline one, you become the agency that says no — even though you'd been saying yes for months.
So state the exclusions plainly, at kickoff, without apology:
- What's outside the scope — named specifically, not implied by absence
- What's available at additional cost — which frames exclusions as options rather than refusals
- How additional requests get handled — the actual mechanism, so the first one isn't a negotiation
- What you need from them — approvals, content, access, response times. Their obligations are part of scope too
That last point is underused. Delays caused by slow client approvals get attributed to the agency by default unless the expectation was set beforehand. Put their side of the bargain in writing at the same time as yours.
Access is the real timeline, not your process
The single most predictable cause of a slow start, and the one most agencies fail to plan for.
The gap between "contract signed" and "actually able to do the work" is usually two to three weeks, and almost none of that is your process. It's waiting for access — from people who weren't in the sales conversation, don't share the urgency, and sometimes no longer work at the company.
| What you need | Who usually holds it | Difficulty |
|---|---|---|
| Analytics and search console | Marketing, or a previous agency | Easy — unless the previous agency owns it |
| Ad accounts and business manager | Marketing or finance | Moderate — billing ownership complicates it |
| CMS admin | Whoever built the site | Moderate — often an external developer |
| DNS / domain registrar | IT, or a founder's personal account | Hard — the classic multi-week blocker |
| Email platform | Marketing | Easy |
| CRM | Sales operations | Moderate — different team, different priorities |
| Brand assets and usage rights | Design, or a former agency | Moderate — rights often undocumented |
Two rules that save weeks. Send the entire list on day one, not as you need each item — sequential requests turn a two-week delay into a six-week one. And ask for a named owner beside each line, because "we'll sort it out" reliably means nobody has been made responsible.
One principle worth holding firm on: always request permissioned access to accounts the client owns rather than creating accounts you own. It's slower at the start and it protects both sides. An agency holding the client's ad account hostage is a relationship-ending event, and being the agency that visibly cannot do that is a trust asset.
GoHighLevel
All-in-one agency platform with CRM funnels email and white-label client management
Best for: All-in-One Agency Tool
The kickoff meeting most people waste
Kickoffs default to introductions and a walk through the plan. Both are already in the documents. Use the hour for the things that only exist in people's heads.
What's been tried before, and what happened. The most valuable and least-requested information in any engagement. Clients rarely volunteer failures, and you'll otherwise spend three months rediscovering that the thing you're proposing was tried in 2024 and didn't work — for reasons that still apply.
Who actually decides. Not who's on the call. Who signs off on creative, who controls budget, and who can quietly veto work after it's approved. That last person is real and is almost never mentioned.
What success looks like to them personally. The contract says pipeline or traffic. Your contact may need something narrower — a specific number for a board meeting, a channel their director keeps asking about, evidence that hiring you was a good call. Knowing that shapes what you show and when.
Constraints nobody has mentioned. Legal review cycles, brand rules, a competitor they can't be seen near, a founder with opinions about colour. These emerge eventually. Better in week one.
The measurement baseline. Agree what the numbers are today, in writing, before you change anything. Without it, month-six reporting becomes an argument about the starting point — and it's a harder argument than it used to be, given how much attribution has degraded. Agreeing what you'll both accept as evidence, at the start, is worth more than any dashboard.
The expectation curve
The most useful document you can produce in week one, and the rarest.
Marketing work has a delay between effort and result that clients consistently underestimate — particularly in SEO and content, where months of work precede visible movement. The client who's told "results take time" hears a hedge. The client shown a month-by-month curve understands a mechanism.
So write it out. Not promises — a sequence:
- Month 1 — access, audit, baseline, foundations. Almost nothing visible. Say so now.
- Month 2 — first work shipping. Leading indicators only.
- Month 3 — early signals. Enough to know whether the direction is right.
- Months 4–6 — compounding. Where the results you were hired for typically begin appearing.
Then name what to watch in each phase. Month-one impatience is almost always caused by a client with no idea what they should be looking at, so give them leading indicators to hold onto while the lagging ones develop. This pairs naturally with setting goals and KPIs properly — the curve is what makes those KPIs legible month to month.
A 2026 complication worth anticipating: clients increasingly arrive with expectations formed by AI tooling, having watched a model draft a campaign in seconds. The reasonable question underneath — "why does this take three months when AI can write it instantly?" — deserves a real answer about what the time is actually spent on. It's the same conversation running through how agencies are pricing AI-assisted work, and it now belongs in onboarding rather than in a defensive email six weeks later.
Set the communication rhythm before you need it
Most agency-client friction is not about the work. It is about not knowing what is happening, which is a different problem with a much cheaper fix.
Agree four things in week one and write them down. Which channel is for what — a standing call, a shared board, email for anything needing a record. How often you report, and in what depth. What your response times actually are, including who to contact when something is genuinely urgent. And who on your side the client speaks to by default.
The reason to fix this early is that unmanaged clients fill silence with worry, and a worried client generates more work than a demanding one. A predictable weekly note costs fifteen minutes and prevents the "just checking in" thread that eats an afternoon.
Reporting deserves its own decision. Agree at the start which numbers you will show, how often, and — importantly — what you will show in the months before the headline numbers move. A dashboard is not a report; someone still has to say what changed and what it means, and that interpretation is a large part of what a client is paying for. If your engagement is one where results compound slowly, this is doubly true, because compounding strategies look like nothing at all until they suddenly don't.
Onboard your own team too
Consistently skipped, and it's why clients repeat themselves.
Everyone touching the account needs the context the salesperson gathered: what the client actually sells, who to, what's been tried, who decides, what's off-limits, what success means here. Without it, the client explains their business three times to three people and quietly concludes you aren't organised.
A useful side effect: the same document is what makes your own team visible as specialists rather than anonymous resource, which is quietly how small agencies win referrals — the mechanism behind winning clients through your team's visibility.
One internal document per account, written during onboarding and kept current. It also means an account survives someone leaving — which matters more than most small agencies plan for.
Agree the ending at the beginning
The section nobody writes, and the easiest ten minutes in the whole process.
While everyone is optimistic, agree in writing: who owns each account and asset, what gets handed over if the engagement ends, in what format, and with how much notice on either side. Ownership of content and creative should be explicit rather than assumed.
This costs nothing now and saves weeks later. Clients also tend to read the willingness to raise it as confidence — an agency comfortable discussing the exit is signalling it doesn't rely on lock-in. Handled well, an ending produces referrals; handled badly, it produces a review.
When not to onboard
Some engagements should be declined at this stage, and the signals are visible if you look. Any one of these is survivable; two or more predicts trouble:
- No single decision-maker. Approval by committee with no tiebreaker means every deliverable takes three rounds.
- Refusing account ownership. If they won't give permissioned access to accounts they own, you'll be blamed for results you can't influence.
- Two agencies fired in eighteen months. Sometimes bad luck. Often a pattern, and you're next.
- Expectations you've already corrected twice. If the reset didn't hold during onboarding, it won't hold in month four.
- Urgency without cause. "We need results by next month" attached to a six-month discipline is a disagreement being deferred, not resolved.
Declining costs you revenue once. Accepting costs you revenue, team morale, and the case study you'd hoped for. This is also a pricing question — engagements that need heavy management should be priced for it, which is the substance of pricing agency services for profitable rates.
A workable sequence
- Day 0 (signature) — welcome note, full access list with named owners, kickoff booked, delivery lead reads the sales notes.
- Week 1 — kickoff covering history, decision-makers, constraints, baseline. Scope and exclusions confirmed in writing.
- Week 1–2 — access arriving, audit underway, internal account brief written.
- Week 2 — expectation curve delivered. Communication rhythm agreed. Ownership and exit terms documented.
- Week 3 — first work shipping, with a note on what to watch now versus later.
- Week 6 — a short check on the relationship itself, separate from a results review. Cheap, and it surfaces friction while it's still small.
That last one is the highest-return item on the list. Asking "how is this going for you?" before anyone is unhappy catches the small misalignments that otherwise compound quietly into a non-renewal.
If you're on the other side of this process — choosing an agency and wondering what good onboarding should feel like — the answers above double as an evaluation checklist. An agency that sends the full access list on day one and volunteers its exclusions is telling you something useful about how the next year will go, and it's reasonable to expect the same from any content and marketing partner you're considering.
The short version
Onboarding starts during the sale, so reset anything that can't be honoured before delivery meets it. State the exclusions as clearly as the inclusions. Send the whole access list on day one with named owners, and ask about DNS immediately. Use kickoff for what isn't written down — history, decision-makers, constraints, baseline. Give the client a month-by-month curve so patience has a shape. And agree the ending while everyone still likes each other.
Looking for a partner who onboards like this?
Clear scope, day-one access lists, and a month-by-month plan you can hold us to from the start.
Explore Working With Us →Frequently asked questions
How long should client onboarding take?
Two to three weeks from signature to productive work is realistic for most marketing engagements, and the limiting factor is almost never your process. It is access. Getting into ad accounts, analytics properties, the CMS, DNS records and third-party tools requires people on the client side who often are not in the room, sometimes no longer work there, and rarely treat it as urgent. Budget for that explicitly rather than promising a fast start and then explaining a delay you could have predicted.
What should be included in a client onboarding process?
Six things. A written scope that states what is not included as clearly as what is. A single named decision-maker with approval authority. A complete access request sent immediately after signature, with named owners for each item. A kickoff meeting focused on goals, constraints and history rather than introductions. An expectation curve showing what happens in which month, so nobody is waiting for results in week three that arrive in month five. And an agreed communication rhythm covering channel, frequency and response times.
Why do client relationships go wrong in the first three months?
Usually because of something that happened before onboarding started. Most early friction traces back to expectations set during the sales process that delivery cannot honour — a timeline that was optimistic, a result implied rather than promised, a scope that felt broader in conversation than it reads in the contract. Onboarding is where that gap becomes visible, but it is not where it was created. The most effective fix is to reset expectations explicitly at kickoff, in writing, even when that means contradicting something said during the pitch.
What access do you need from a new marketing client?
Typically analytics and search console properties, ad accounts across each platform in use, the business manager or equivalent, the CMS with appropriate permissions, DNS or domain registrar access, the email platform, any CRM the work touches, and brand assets with their usage rights. Request all of it in one list immediately after signature with a named owner beside each item, because the delays come from the ones nobody can locate. Always ask for permissioned access to client-owned accounts rather than creating new accounts you own.
Should you discuss offboarding during onboarding?
Yes, and it is easier at the start than at any point afterwards. Agree in writing who owns each account, asset and piece of content, what gets handed over if the engagement ends, in what format, and with how much notice. Doing this while everyone is optimistic takes ten minutes. Doing it during a difficult exit takes weeks and damages a relationship that might otherwise have produced a referral. Clients also tend to read the willingness to discuss it as confidence rather than pessimism.
