Client Onboarding, Converted to Training

The failure mode is not confusion. It is interference.

When a client buys automated lead follow-up, the software starts speaking in their name within days. Every scoping call I have run for this service surfaces the same three failure points: owners keep chasing leads by hand alongside the system, they panic-pause it the first time something feels off, and they miss the moment a real human writes back and the conversation becomes theirs.

None of that is a feature-knowledge gap. A PDF manual answers questions nobody asked. The gap is trust and role clarity: what the system does alone, what stays the client's job, and how the handoff between them works. That is a training problem, so I built the onboarding as a course instead of a document.

The sponsor is me. The metric is the day-30 renewal.

This is training embedded in service delivery. Ghost Catcher is a productized service my studio sells, and I wrote this course as its standard onboarding, so for once the sponsor is not a composite: it is me, and the metric is real. A client who misunderstands the system in week one becomes support load in week two and a cancellation at day 30.

The course is aimed at the four numbers the service is judged on: speed to first touch, lead-to-booked rate, show rate, and founder hours reclaimed. Those are Level 4 outcomes, and the day-30 review where we read them exists in the delivery timeline. The course prepares the client to sit in that review and understand every number in it.

An owner who just gave their voice away.

The learner is the owner of a 5-to-50 person service business. Not technical, permanently short on time, and carrying a specific anxiety this course has to address head on: messages are now going out under their name, to people who know them, written by a system they have never seen.

That anxiety drives the design. The course tells them, in the first minute, exactly what stays theirs, and it repeats the phrase read this as if you wrote it, because in effect you did. These learners are problem-centered in the extreme: they do not want to understand the system, they want to know what to do when a lead writes back at 9pm. Every lesson is built around a moment like that, not around features.

Four objectives, each one checkable in the product.

By the end of the course, the client can:

• Distinguishthe three jobs that stay theirs from the seven the system runs alone, without checking the reference.
• Identifywhich of the eight pipeline stages any lead is in and what, if anything, that stage needs from them.
• Executea takeover of a live conversation, replying from their own inbox and moving the lead when the conversation ends.
• Evaluatetheir four before-and-after numbers at day 30 and decide with me what to change.

The verbs climb Bloom's from recall to evaluation, but the real discipline here is that every objective is observable in the CRM. I do not have to ask the client whether they learned; I can watch the pipeline and see it.

Personalized by merge fields, gated by real approvals.

The biggest decision: this is not a generic course with the client's logo on it. It is built with merge fields, so every client reads their own CRM name, their own call dates, and, most importantly, their own qualification rubric, the three lines I write from their real won and lost deals. The course says those lines are the ones to argue with. Onboarding doubles as design review.

Second decision: the calls to action are real workflow steps, not quiz buttons. Lesson 2 ends in the actual intake form. Lesson 3 ends in the actual approval gate, and nothing sends until the client clicks it. The course sits inside the delivery pipeline instead of beside it, which is why it gets completed. Third: the one knowledge check spends all its weight on the runs-without-you split, because duplicated follow-up is the single most expensive misunderstanding a client can carry.

A learner who never asked for a course.

Nobody buys Ghost Catcher to take training. The course has to be short enough to finish between jobs, work on a phone, and never feel like compliance. It also has to function mid-service: lesson 3 only becomes actionable during the build stretch, when the client's real sequence copy exists to approve, so the course is paced to the 30-day delivery arc rather than consumed in one sitting.

Tool constraints shaped it too. Rise has no native way to branch on a client's answers, so personalization lives in merge fields rather than adaptive paths. And because Articulate's hosted share links disable external links on trial accounts, the course ships as a self-hosted web export, which is also how it survives any future change in tooling.

Rise, exported and self-hosted.

Built in Articulate Rise with the blocks the design actually needed: a numbered process for the four-step system and the eight pipeline stages, accordions for the intake inputs, checklists for the first week, a table for the four numbers, and a single scenario-style knowledge check. Exported as web output and served from this site, so the course link in front of you does not depend on a vendor account staying alive.

The live course is public. The merge fields render as placeholders here, which I left visible on purpose: they show where each client's real rubric, dates, and system names drop in.

The pipeline is the assessment.

This course has an unusual advantage: the behaviors it teaches are all visible in the product, and the business result it serves has a scheduled review. So the evaluation plan leans on observation over surveys, and it is honest below about what is built versus what runs once a client cohort exists.

What exists today, and what day 30 will tell me.

What exists: a complete four-lesson onboarding course, wired to the 30-day delivery timeline, personalized per client through merge fields, with its approval gates and intake form built in, exported and self-hosted. It is the standard onboarding for the service, written once and reused for every client.

What is not yet claimed: results. No client cohort has completed a day-30 review yet, so the four numbers have no after column, and this page will not pretend otherwise. The evaluation machinery is built into the service; the first cohort fills it in. When that happens, the numbers land here.