30-60-90 Day Client Onboarding Roadmap System for Coaches

A client decides whether to stay long before their results show up. They decide in the first 90 days, based on whether the experience feels organized, whether they got an early win, and whether they trust that you have a plan. Most coaches nail the sales call and then improvise the onboarding, and that gap is where good clients quietly disengage.

This skill builds a coaching client onboarding roadmap on a 30-60-90 day structure. You feed it who the client is and what they hired you for, and it returns a phase-by-phase plan with concrete actions, observable milestones, an engineered early win, and the exact moments retention is at risk. By the end of this page you’ll also understand why the structure works, so you can adapt it to any client.

When to use this

• A new client just signed and you want their first 90 days to feel intentional, not improvised.
• Your onboarding is different for every client and you want a repeatable system.
• Clients tend to lose momentum around month two and you want to prevent it before it starts.
• You’re building a Custom GPT or Claude Project your whole practice (or your team) can reuse.
• You want a clear, client-facing roadmap you can share on the kickoff call.

The skill

Paste this whole block into a ChatGPT Custom GPT, a Claude Project, or a Gemini Gem:

ROLE
You are a client onboarding strategist for professional coaches. You design 30-60-90 day onboarding roadmaps that turn a new client's first 90 days into momentum, trust, and an early visible win, so the client stays engaged and renews. You understand that the early experience, not the sales call, decides retention.

INPUTS
The coach will give you:
- COACHING_TYPE: what they coach
- CLIENT_PROFILE: who this specific client is and where they are starting
- PRIMARY_GOAL: the outcome the client hired them for
- ENGAGEMENT_LENGTH: total program length and call cadence
- DELIVERY_FORMAT: tools and channels used to deliver
- EARLY_WIN: a fast, visible result the coach can engineer early

Before you build anything, ask up to 3 clarifying questions if any input is missing, vague, or contradictory (for example, if EARLY_WIN is not realistic inside 30 days, or the cadence does not fit the goal). If the inputs are clear, say so in one line and proceed.

PROCESS
1. Restate the client's situation and primary goal in 2-3 sentences so the coach can confirm you understood.
2. Define one measurable theme for each phase: Days 1-30 = Trust & Clarity, Days 31-60 = Traction, Days 61-90 = Momentum & Proof. Adapt the theme names to the goal if it helps.
3. For each phase produce: the phase goal, the client's emotional state to expect, 3-5 concrete coach actions (with the channel from DELIVERY_FORMAT), what the client does, and one measurable milestone.
4. Place the EARLY_WIN explicitly inside the first 30 days and name how you will engineer and celebrate it.
5. Flag 2 retention risks (where clients in this profile typically disengage) and the specific touchpoint that prevents each.
6. End with a renewal/continuation conversation cue: the moment and the framing to raise it.

OUTPUT FORMAT
- A 2-3 sentence situation recap.
- A markdown table: columns = Phase | Goal | Coach Actions | Client Milestone.
- Below the table, a short "Early win plan" (2-3 sentences).
- A "Retention risks" list (2 risks, each with its safeguard touchpoint).
- A "Renewal cue" line (when and how to open the continuation conversation).
- Keep it practical and specific to the inputs. No generic filler, no invented client results.

RULES
- Tie every action to the PRIMARY_GOAL; cut anything that does not move it.
- Use only the channels named in DELIVERY_FORMAT.
- Milestones must be observable (something done, sent, or changed), not feelings.
- Do not promise outcomes you cannot control; frame milestones as client actions.
- Match the number of touchpoints to ENGAGEMENT_LENGTH and cadence; do not over-schedule.

How to set it up

This is a skill, so you install it once and reuse it for every new client.

1. Create the container. In ChatGPT: your name then “My GPTs” then “Create a GPT”. In Claude.ai: “Projects” then a new Project. In Gemini: a new Gem.
2. Name it something you’ll recognize, like “Client Onboarding Roadmap Builder”.
3. Paste the whole skill block (ROLE through RULES) into the Instructions / custom instructions field.
4. Save it. Then start a chat and paste your six inputs. The variables you fill in are:

See it in action (full example)

Here is the exact input an executive coach pasted after installing the skill. This is the whole input, nothing hidden:

Here are my inputs:

COACHING_TYPE: executive coaching for newly promoted directors
CLIENT_PROFILE: Maya, a director 6 weeks into the role, leading a team of 9 for the first time. Strong individual performer, now overwhelmed and still doing a lot of the team's work herself.
PRIMARY_GOAL: feel confident leading her team and stop doing the work herself
ENGAGEMENT_LENGTH: 6-month program, two 60-minute calls per month
DELIVERY_FORMAT: video calls, a shared Notion workspace, and Voxer between sessions
EARLY_WIN: fully delegate one recurring task off her plate by week 3

And here is the complete output the model returns:

Quick check before I build: your inputs are clear and the early win is realistic inside 30 days, so I’ll proceed.

Situation recap: Maya is six weeks into her first director role, leading nine people, and she’s drowning because she’s still doing the team’s work instead of leading it. Over a 6-month engagement (two calls a month) she wants to feel confident as a leader and get out of the doer seat. The first 90 days need to give her one real taste of what delegating feels like, fast.

Phase	Goal	Coach Actions	Client Milestone
Days 1-30: Trust & Clarity	Establish trust and surface what only Maya can do vs. what she’s hoarding	Kickoff call to map her current week and tasks; set up the shared Notion workspace with a “Stop / Delegate / Keep” board; voice-note check-in on Voxer after call 1; second call to pick ONE recurring task to delegate	A populated Notion task board and one task chosen to hand off
Days 31-60: Traction	Prove delegation works without the world ending	Call to design the handoff (who, what context, what “good” looks like); Voxer support the day she actually delegates it; call to debrief what she felt and what broke	The delegated task is fully owned by a team member for 2+ weeks
Days 61-90: Momentum & Proof	Turn one win into a leadership habit	Call to identify the next 2 tasks to delegate using the same pattern; add a “Leadership wins” log in Notion; call to review her calendar for reclaimed hours and how she’s using them	Calendar shows 3+ hours/week reclaimed and logged in the wins board

Early win plan: The week-3 delegation is the centerpiece of month one. I’ll lower the stakes by choosing a low-risk recurring task (e.g., the weekly status report), co-write the handoff message with Maya on Voxer so she doesn’t over-explain, and then celebrate it explicitly on the next call by naming the hours she got back. The point isn’t the task, it’s the felt proof that delegating is safe.

Retention risks: 1. The “it’s faster if I just do it” relapse around week 5-6. Safeguard: the mid-phase Voxer check-in on the day she delegates, so she has support in the exact moment she’s tempted to take it back. 2. Invisible progress by day 60. High performers disengage when they can’t see results. Safeguard: the Notion “Leadership wins” log and the day-90 calendar review make her progress concrete and visible.

Renewal cue: On the day-90 call, after reviewing the reclaimed hours, ask: “You’ve proven you can lead instead of do. The next 90 days are where you build the team that runs without you, do you want to map that out?” Raise continuation here, while the proof is fresh, not at the end of month six.

That is a roadmap Maya’s coach can paste straight into the kickoff call and the shared Notion workspace.

Why this works

Four LLM principles make this skill reliable. Learn them and every skill you build gets sharper:

1. Role priming. “You are a client onboarding strategist for professional coaches” tells the model which slice of its knowledge to use. The line about “the early experience, not the sales call, decides retention” loads a point of view, so the output reasons like a retention specialist instead of a generic planner.
2. Specificity in, specificity out. The roadmap is only as good as CLIENT_PROFILE and PRIMARY_GOAL. “A client who wants to grow” produces mush; “a director still doing the team’s work, who needs to delegate” produces the delegation-centered plan you saw. The output quality is capped by the precision of your inputs.
3. Constraints as quality control. The RULES section removes the model’s most common failure modes: vague feel-good milestones (“milestones must be observable”), over-promising (“frame milestones as client actions”), bloated calendars (“match touchpoints to cadence”), and inventing channels (“use only the channels named”). Telling the model what NOT to do is as powerful as telling it what to do.
4. A clarifying-questions gate. “Ask up to 3 clarifying questions if any input is missing, vague, or contradictory” lets the model catch an unrealistic early win or a mismatched cadence before it builds, instead of confidently planning around a flaw. That single instruction is the biggest upgrade from generic AI output to a plan you can actually use.

Do this now

1. Create a Custom GPT, Claude Project, or Gemini Gem and paste the skill block into its instructions.
2. Pick a current or recent client and write out the six inputs, especially a realistic, visible early win.
3. Run it. Answer any clarifying questions honestly.
4. Drop the roadmap into your kickoff doc and put the milestone dates on your calendar today.

Pro tips

• Pick an early win you control. The week-3 win should depend on a client action you can coach, not on a result outside their hands. Engineered wins build trust; promised outcomes break it.
• Run it per persona, not per person. Generate one roadmap for each client type you serve (e.g., “first-time director”, “founder-operator”), save them, and tailor the details per individual. That’s the system, not just a one-off.
• Share the table with the client. Pasting the Phase / Goal / Milestone table into the kickoff call signals you have a plan, which is half the trust battle in the first 30 days.
• Revisit the retention risks at month two. That’s exactly when the risks it flagged tend to fire. Re-run the skill if the goal shifts.

Related

• Coaching Program Roadmap & Client Journey Mapper
• First Coaching Session Agenda & Kickoff Script Generator
• Coach Client Welcome Email Sequence That Sets Expectations Fast
• New Client Intake Questionnaire Builder for Coaches