The Implementation Gap
Most coaching programs don't have a content problem. They have an implementation problem. And almost nothing in the category is built to solve it.
I've been having a version of the same conversation with coaching founders for months.
Cohort one raves. The content is genuinely good, NPS is fine, the testimonials look like the website promised they would. By cohort three, renewals start drifting. Coaches burn out answering the same five questions every cycle. Someone floats the idea that maybe the curriculum is stale, and six weeks get eaten producing new modules that don't move the metric.
The instinct is to blame the content. It's almost never the content.
The gap
There's a gap between "I learned this" and "I did this." That's the whole story.
Watching the video isn't the bottleneck. Understanding the framework isn't. The bottleneck is the moment a student tries to apply what you taught, at her desk, in her context, with her specific situation, at 11pm on a Tuesday with a half-written cold email and a Wednesday morning deadline. She has no idea what to do next.
The thing that reliably gets her unstuck in that moment is a coach who knows her, knows the material, and is around the second she's stuck.
That's the unscalable thing. The whole reason your unit economics work is that you've already accepted you can't put a 1:1 coach on every student. So you substitute. And the substitutes leak.
Group calls are one-to-many, scheduled, and generic. The student stuck at 11pm Tuesday isn't on one. She's alone with that cold email. Slack is public and slow, and nobody asks the basic question in front of 400 peers. They lurk. Or they DM the coach directly, which is the thing you were specifically trying to prevent when you set up the channel. Community helps too, but peer advice is the second-best version of the answer, delivered by someone with their own gaps in the methodology.
All of it is real value. None of it is the coach at 11pm.
What the data actually says
We don't have to argue from anecdote.
In Katy Jordan's MOOC study (221 courses), completion rates ranged from 0.7% to 52.1%, with a median of 12.6%. MOOCs aren't premium coaching. Different buyer, different stakes, different promise. But the pattern shows up across the category: access to content does not equal progress.
Premium programs do better. Cohort structure helps. Active discussion helps. Live touchpoints help. Platform data from Ruzuku, across 32,000+ courses, shows 65.5% completion in courses with active discussion versus 42.6% without. Cohort-based courses hit 64% versus 48% for self-paced. Meaningful lifts. They still don't close the gap. A student can disappear between calls. She can hit the same wall at midnight. The answer can sit in your 300-hour library with no practical way to retrieve it.
That's a different problem. The substitutes weren't designed for it.
Where the platforms are pointing AI
Open the AI roadmap of any course platform shipping right now. Auto-clipping for social. AI-drafted sales pages. Subject line testers. Email sequence generators. Thumbnail optimizers. AI quiz builders.
These are creator-productivity and acquisition features. Get more students in. Lower CAC. Make the funnel taller.
A few platforms have started gesturing at learner-facing AI. Teaching assistants, course-aware chat, quizzes, summaries, the right idea. The category needs more of it. But most of it still stops short of the real implementation moment.
The student isn't asking, "Can you summarize lesson four?"
She's asking, "I tried to use your cold email framework on this prospect, and here's the draft. Why does it feel wrong?"
That's a different problem.
The center of gravity still sits at the top of the funnel. The implicit bet is that retention is a marketing problem you can outrun with growth. Pour faster than the bucket leaks. The math eventually catches up.
Renewal doesn't happen at renewal
The renewal decision usually happens before the renewal conversation.
It happens in month three, when the student gets stuck and misses a call. It happens in month five, when she stops asking questions because she's embarrassed to be behind. It happens in month eight, when she's still paying but no longer progressing.
By the time renewal comes around, she isn't angry. She just doesn't have enough momentum to buy another year.
The math, conservatively
Pick the smallest version of the number you find believable.
One percentage point of annual gross revenue retention is $10K on a $1M coaching program. $20K on $2M. $50K on $5M. Recurring. Compounding. No extra acquisition spend.
That's one point. And you don't need to believe in a dramatic swing for the math to work. If closing the gap moves retention by three to five points, that's $60K to $100K on a $2M program. On $5M, $150K to $250K.
For monthly subscription programs the picture is sharper. Even at 3 to 4% monthly gross revenue churn, the high-ARPU benchmark from public subscription data, annualized retention lands around 61 to 69%. So a "70% renewal rate" isn't some apocalyptic edge case. For many high-ARPU recurring programs, it can be uncomfortably normal. Even there, you're re-selling 30% of the business every year before you grow.
That 30% usually disappears long before renewal. It disappears when the student gets stuck and stops moving.
The library is supposed to be the answer
I wrote about this last week. Most course platforms can't actually search their own video.
Your stuck student at 11pm is wrestling with a question. The answer is probably somewhere in the 300 hours of content she already has access to. You probably already taught it. You may have taught it three times.
She can't find it. Coaches teach by telling stories, not by saying the keyword, so transcript search misses everything that matters. She gives up looking, posts in Slack, waits, and doesn't get a useful answer in time. She gets unstuck some other way. Or she doesn't get unstuck at all. The moment passes.
The library is technically the answer. Functionally, it's invisible.
What we're building Bold for
The closest scalable approximation of a coach at 11pm. That's it.
Not a replacement for human coaching. The 1:1 work you do still matters, probably more than ever. Bold is for the gap between calls. The student stuck at 11pm who doesn't have a session until Thursday. The cohort member halfway through implementing your framework who needs to know what you would say about her specific situation, right now, before the moment passes.
It's grounded in your library. Your videos, your methodology, your voice. When she asks a question, the answer comes back in the language of your program, citing the moment in module 14 where you covered exactly this.
The category keeps treating the renewal problem as either a content problem or a funnel problem. It's neither. It's an answering problem, in the moment a student is stuck.
If your renewals are softer than they should be, the content probably isn't what's broken. The gap between learning and doing is.
If you're running a coaching program or training academy and your renewals are softer than the content deserves, we'll show you what closing the gap looks like on your library in 30 minutes. See Bold on your library →
— Marcel