The client onboarding call: an agenda that holds.

The kickoff is the first thing you actually deliver.

Everything before it was a promise. This hour is the client’s first evidence of how you work, and they will read it that way whether or not you intend them to.

Which is why treating it as a formality is expensive. A client who leaves the first call unsure what happens next, who to contact, or what you need from them will spend the next fortnight generating the exact work you were hoping to avoid: chasing emails, a slipped first deliverable, and a quiet recalibration of how much they trust the engagement. None of that is a relationship problem. It is an agenda problem.

The fix is unglamorous. Run the same call every time, in the same order, with the same document in front of it and the same recap behind it. A system beats charisma here precisely because it works on the day you are tired, and because the second person on your team can run it without having watched you do it eleven times.

Three ways the hour goes wrong

All three feel fine while they are happening. That is what makes them worth naming — you cannot correct a failure you experienced as a good conversation.

Send the document that lets the hour start at minute zero.

Most first calls lose their first fifteen minutes to information that could have been read in four. The pre-call document buys those minutes back and does something more useful with them.

Send it two to three days ahead. The night before is too late for anyone with a real calendar, and a week ahead means it is forgotten by the time it matters. Keep it to a page or two: this is a briefing, not a deliverable, and a twelve-page onboarding pack signals process rather than clarity.

Mark the plan inside it as a draft, visibly and in words. A plan that arrives looking finished invites either silent disagreement or a defensive reading. A plan marked DRAFT invites the correction you actually want, and gives the client a reason to arrive with something to say.

Subject: Before Thursday — the agenda and two questions

[Name] — looking forward to Thursday at [Time] [Timezone].

The document below is the agenda for the hour, the plan in
outline, and the first thing we intend to deliver. It is a draft
on purpose. The call is where it becomes real.

[Link to the document]

Two questions before then, and short answers are genuinely fine:

1. In your words, what does a good outcome look like six months
   from now?
2. Who has final say on decisions, and who else should be in
   the loop?

If you get to the document, great. If you do not, the call still
works — I will walk through it.

[Your name]

Then assume they did not read it

Design the hour so it works either way. If they read it, you move faster and spend the saved minutes on their answers. If they did not, nothing breaks — you walk the same ground slightly slower. What you must not do is build the call on the assumption that they have, and then spend the first ten minutes visibly recovering from the fact that they have not.

The two questions at the end of that email do more work than the document. Answers give you their language before the call, which is what lets you open Part 02 with you said rather than tell me about. No answers is information too: a client who does not reply to two short questions before the kickoff is telling you something about the feedback windows you are about to agree.

Five parts, in an order that does not collapse.

Minutes are attached because an agenda without them is a list of topics, and a list of topics always spends its time on whichever item the room finds most comfortable.

Why this order

Their expectations come before yours because the person who speaks second gets to respond, and you want that advantage on the harder half. Expectations come before the plan because a plan presented into unstated assumptions gets nodded at rather than examined. The plan comes before the first deliverable because the deliverable only means something once its place is visible. And the norms come last, next to the recap, because that is where they get written down — the part of the hour people forget belongs beside the part that produces the record.

Ninety minutes, and thirty

Take ninety when several stakeholders need to speak or the engagement crosses teams; the extra time goes to Part 02 and to letting each stakeholder answer the outcome question separately, which is usually where the disagreements surface. If you only have thirty, cut the plan walkthrough and send it in writing. Do not cut the expectations exchange: it is the only part of the hour that does not work as a document.

Five minutes, and none of it about the weather.

The opening sets whether this is a meeting or a chat. Warmth here comes from being organised, not from small talk — the client relaxes when it becomes obvious that somebody has thought about the hour.

1. 01Names and roles, both sides. One line each, and specifically what each person is responsible for in the first month. Not job titles — responsibilities. “Priya writes the copy and will send you drafts” tells the client who to reply to.
2. 02What this hour is for. One sentence: to agree what good looks like, what happens first, and how we work together. Naming the destination stops the call drifting toward whatever comes up first.
3. 03The shape, out loud. The five parts and roughly how long each takes. This is what earns you the right to say “let us come back to that in the plan section” twenty minutes later.
4. 04The recording line. Ask here, before anything worth recording has been said. One sentence, asked rather than announced — the wording is in the recording section below.

Then stop. The most common mistake in the first five minutes is filling them, because silence early in a call feels like a failure. It is not. Part 02 needs the client talking, and a client who has just listened to nine minutes of you is harder to get talking than one who has listened to four.

Expectations, stated in both directions, out loud.

This is the part that gets skipped, and it is the part the rest of the engagement rests on. When an engagement sours in month three, it is usually against an expectation nobody said out loud in week one.

Ask them first. Their answer gives you language you will use for the whole engagement, and it puts you in the position of responding rather than presenting. Write the phrase down exactly as they say it — not your tidied version. “I want to stop being the bottleneck” is worth more in month four than “improve operational efficiency”, because you can hand the first one back to them and they will recognise it.

Before I tell you how we work, I want to hear it from you.
When this is finished and it has gone well, what is different
about [their business, their team, their week]?

[Stop talking. Write the phrase down in their words, not yours.]

Here is what you should be able to expect from us — and then
what we will need from you, because the second half is the part
that usually goes unsaid until it is a problem.

Then say your half. The order matters: what they can expect from you comes first, because it earns the right to ask for the second half. Said in that order it is a commitment followed by a request. Said the other way round it is a list of conditions.

Say the uncomfortable one anyway

The line most people swallow is the feedback window: if review takes longer than agreed, the date moves. Said in week one it is a reasonable operating fact that the client will usually accept without blinking. Said in week five, after a date has already slipped, it is an excuse — the same sentence doing completely different work depending on when it was first spoken.

Phases on the screen, and the first two dates fixed.

Thirteen minutes is not enough to explain a six-month plan, and it is more than enough to agree its shape. Aim for the second thing.

1. 01Phases, not tasks. Three to five named stages with a rough duration each. Task-level plans invite task-level negotiation, which is a conversation for the working sessions and not for the hour where you are establishing how the engagement runs.
2. 02What they will see, and when. For each phase, the thing that arrives in their inbox. Clients do not experience your phases; they experience the artefacts. Naming those is what makes a plan feel real rather than administrative.
3. 03What is explicitly not in scope. Two or three items, named out loud. This is the single highest-return sentence in the hour: the thing they assumed was included, said now, costs a short conversation. Said in month two, it costs the relationship a layer of trust.
4. 04The dependencies that live on their side. Access, approvals, data, the person whose sign-off gates phase two. Name them as items with owners and dates, not as a general request for cooperation.
5. 05Two dates, fixed before you move on. The first deliverable and the next call. Everything else can stay indicative. A plan where nothing is fixed is a document; a plan with two fixed dates is a commitment with a shape.

Do not improvise the plan on the call. If a question exposes something you have not thought through, say so and put it in the open-questions section of the recap with a date attached. A consultant who says “I do not want to answer that badly, so I will answer it Thursday” reads as careful. Inventing an answer under mild time pressure produces the commitments you spend the next month regretting.

One thing, dated inside the first week.

The gap between signing and seeing something is where new clients get quietly anxious. Closing it early is worth more than the deliverable itself usually is.

It does not have to be large. An audit summary, a draft structure, a first session booked and prepared, an access review, a single page of findings — anything that turns a payment into evidence that work has started. What it must be is dated, specific, and yours: a deliverable that depends on something the client has not sent yet is not a first deliverable, it is a first dependency.

Say what it is not, as well

“A structure for the programme, not the content itself.” “The findings, not the recommendations — those come in phase two.” Bounding the first deliverable out loud prevents the most common early disappointment, which is not a missed date but a delivered thing that was not the thing the client had pictured.

Then name what has to arrive from their side for the date to hold, and put a date on that too. “We will have it with you Thursday, as long as the analytics access lands by Monday” is a commitment with its condition attached — and if Monday passes, the conversation about Thursday is a conversation you already had.

Where we talk, how fast, and who to ask.

Unstated communication norms are the most reliable source of friction in a new engagement, and the cheapest to remove. Every item here takes fifteen seconds to agree and saves a fortnight of mismatch.

• →One channel, named. Email, a shared channel, a portal — one of them is the record and the others are conversation. Say which. Work that gets agreed in three places gets remembered in none.
• →Response windows, both directions. Yours and theirs, in working hours or working days. Promise the window you can actually staff on a bad week, not the one that sounds impressive on a good one.
• →What counts as urgent, and where urgent goes. If there is a phone number for genuine emergencies, give it and define the emergency. Undefined urgency defaults to everything being urgent, which means nothing is.
• →The standing meeting, or the deliberate absence of one. Day, time, length, and what it is for. Then the rule that makes it survive: no agenda, no meeting. A recurring call with nothing to decide trains people to skip it.
• →Where the current state lives. One link where anyone can see what is in progress without asking. A good share of status requests are not questions at all — they are someone checking that something is still happening.
• →Who to escalate to, by name. Including on your side. A client who knows who to call when something is going wrong is far less likely to need to call them.

Then the last five minutes

Recap out loud before anyone leaves: decisions, owners, dates, and what each side is doing before the next call. Then book the next two dates while everyone still has their calendar open. A call that ends with “I will send some times” ends with a scheduling thread; a call that ends with two accepted invitations ends with a rhythm.

Ask before you press record, in one sentence.

Recording a kickoff is usually the right call. Recording it without asking is the kind of small thing that quietly changes how candid the next fifty minutes are.

Ask at minute four, before anything worth recording has been said, and ask rather than announce. Four facts belong in the sentence: why you record, where it goes, who can watch it, and how long you keep it. Add the exit — that they can ask you to stop or delete at any point — and the whole thing takes twelve seconds.

I record these calls so nobody has to take notes and so anyone
who could not make it can catch up. It goes to [where it lives],
where [who can see it] can watch it, and we keep it for
[retention period]. If you would rather I did not — now or at
any point later — say so and I will stop or delete it.

If anyone hesitates, take the no immediately and without negotiation, then say what you will do instead: written notes circulated after the call. A client who feels they had to justify declining will be more careful about what they say for the rest of the hour, which costs you far more than the recording was worth. Local rules on consent and retention vary by jurisdiction and by industry, and are worth checking once rather than assuming.

Where the recording lives is most of its value

A recording in a folder nobody can name is an archive; a recording attached to the client it belongs to is a resource. The test is simple: when someone joins the team in month three and asks what the client actually wanted, how many clicks is it? If the honest answer involves searching a drive by date, the recording is not doing the job you recorded it for.

Within one working day, the decisions in writing.

The call produced agreement. The recap is what turns agreement into a record — and the speed of it is itself a demonstration of the response window you just promised.

Five sections and nothing else: what was decided and who owns it, the dates that were fixed, what you need from them and by when, what is still open along with when it will be closed, and the recording link. Resist the urge to write a warm narrative of the conversation. The client was there. What they need from you is the part they will forget by Tuesday.

Subject: [Engagement] — what we agreed today

[Name] — good call. Here is the record, so we are both working
from the same one.

DECIDED
· [Decision] — owned by [name]
· [Decision] — owned by [name]

DATES
· [First deliverable] — [date]
· Next call — [date], [time] [timezone]

WITH YOU
· [What we need] — by [date]
· [What we need] — by [date]

STILL OPEN
· [The question we did not settle, and when we will]

Recording: [link]

If anything above is wrong or missing, reply and say so. I would
rather fix it today than discover it in three weeks.

[Your name]

The last line is the mechanism

Inviting correction is what converts your summary into a shared one. Without it, a recap is one party’s account of a meeting. With it, silence becomes something closer to assent — and in the month-three conversation where two people remember the same sentence differently, a dated document that asked to be corrected and was not is a considerably better position than a good memory.

Send it inside one working day. Same day is better, because it arrives while the conversation is still live in their head and corrections are cheap. A recap that arrives on Friday for a Tuesday call gets skimmed, and skimmed recaps do not get corrected.

Seven days later, before drift becomes habit.

The kickoff created a set of expectations. Week one is when you find out whether they survived contact with everyone’s actual calendar — and it is the cheapest week in the engagement to fix what did not.

1. 01Send it on day seven, not day fourteen. A week is long enough for the first deliverable to have landed or slipped, and short enough that a stalled client has not yet built a habit of not replying. Fourteen days is where a small drift becomes an awkward conversation.
2. 02Ask three questions, not a status form. Has anything landed. Is anything blocked on you. Is anything blocked on them. Three lines are answerable from a phone in a taxi, which is the point — a check-in that requires effort gets postponed and then forgotten.
3. 03Read the silence as data. The client who was enthusiastic in the kickoff and has not replied to two messages in week one is not busy; something has changed. Pick up the phone rather than sending a third message, and ask directly whether the priorities have moved.
4. 04Fix the norm, not the incident. If feedback came back late, the answer is not to chase harder. It is to say out loud that the agreed window did not hold, ask what would make it hold, and change either the window or the reviewer. The first week is the cheapest moment to renegotiate something that will otherwise break every phase.

[Name] — one week in. Three questions, and one-line answers are
perfect:

1. Has anything landed that you expected to land?
2. Is anything blocked on us?
3. Is anything blocked on your side that we could help unblock?

If it is easier said out loud, here is fifteen minutes on [day]:
[booking link].

Three lines, answerable from a phone. The check-in that asks for a status update gets postponed until the client has time to write one properly, which is a polite way of saying it gets postponed. Offering a short call as the alternative rather than the default respects the fact that most weeks, three sentences are genuinely enough.

What the answers tell you

A client who answers all three in a sentence each is fine. A client who answers only the first, or answers warmly but without content, has usually hit something internal they have not decided how to tell you about. And a client who does not answer at all in week one has given you the most useful signal of the lot, well before the first invoice becomes an awkward conversation. In all three cases the next move is the same: a short call, asked for directly, without an apology attached to it.