What it really costs to build training

A $100-billion habit

Training is not a side expense. U.S. organizations spent $102.8 billionon training in 2025 — roughly $874 for every learner on the payroll (Training Magazine, 2025 Industry Report). For most teams the cost shows up as people's time rather than a software invoice, which is exactly why it's easy to underestimate.

Here's the part that surprises people: the slow, expensive stage isn't writing the course. The words are the fast part. What eats the budget is everything that wraps around them — designing the learning, scheduling subject-matter experts, producing media, and rebuilding the whole thing when a source document changes. To see where the money goes, it helps to look at what a single finished hour actually costs — and the hours behind it.

Industry benchmarks, not Scormy One figures — sources cited inline throughout.

What one finished hour costs

Start with today's price tag. Building one finished hour of interactive e-learning runs roughly $15,000–$30,000 (LearningSim, 2024 (consultant estimate)). That number isn't the authoring software — it's the people-hours behind every finished hour.

The most-cited breakdown of those hours comes from the Chapman Alliance development-ratio study (Chapman Alliance, 2010). It's old enough that you should always cite the year — but no survey-grade replacement has displaced it, and recent cost estimates land in the same range, so it still anchors most planning conversations.

For basice-learning — mostly text, images, and simple knowledge checks — the study found an average of 79 hours and roughly $10,054 per finished hour (Chapman Alliance, 2010). Add interactivity and the effort climbs fast.

Read the table as a slope, not a single price. Move from basic to interactive — branching scenarios, drag-and-drop, richer assessment — and you go from 79 hours to 184 hours and about $18,583 per finished hour. Push to advanced — simulations, custom media, heavy interactivity — and a single finished hour can reach 490 hours and roughly $50,371 (Chapman Alliance, 2010). The implied loaded labor rate across those numbers lands somewhere around $100–127 an hour — which is how 184 hours becomes the tens of thousands of dollars you see in current estimates.

Where the time and money go

If a finished hour of basic e-learning costs ~79hours, it helps to ask where those hours actually land. They cluster in three places, and none of them is “writing the script.”

Instructional design.Before anyone writes a sentence, someone has to decide what the course is for: the learning objectives, the structure, how each idea will be assessed, and how it all maps back to the source material. This is skilled work, and it's where a lot of the early hours go.

Subject-matter expert (SME) review cycles. This is the single biggest threat to a timeline. Experts are busy, their feedback arrives in rounds, and every round means rework. Two or three review passes is normal, and each one stalls until the expert has time. When a project ships late, SME availability is usually why.

Media production.Voiceover, video, on-screen graphics, and avatars all take real time to script, produce, and revise — and they have to be redone whenever the underlying content changes. The richer the media, the steeper that table climbs.

There's a fourth cost that doesn't fit neatly in a row: maintenance. When a policy, SOP, or manual changes, the course built from it is now wrong — and bringing it back into line often means walking the whole pipeline again. The expense isn't building a course once; it's keeping a growing library accurate over time.

What changes with an AI-assisted first draft

The benchmark above describes a blank-page workflow: someone reads the source material, designs the course, and writes everything from scratch. That blank page is where AI assistance changes the shape of the work. Instead of starting from nothing, you start from a source-grounded first draft — an outline, scripts, and knowledge checks generated from the documents you already trust. What used to be weeks of drafting can become a first pass measured in hours.

The honest framing matters here, so let's be precise about what moves and what doesn't. The drafting stage compresses. The judgment stage does not. Humans still own the parts that require a human: the learning strategy, the review and approval, and anything with compliance weight (NAVEX, 2026). Recent research on AI-generated instructional content reaches the same conclusion — automated drafts accelerate the early work but still require expert review to be trusted (arXiv, 2024).

So the right way to think about it isn't “X percent faster.” It's a change in whereyour experts spend their hours. Instead of burning the first stretch of every project on a blank page, they spend it reviewing, correcting, and approving a draft that already traces back to its source. The review step doesn't disappear — and it shouldn't. It just starts from a much better place.

A worked example (illustrative)

Numbers per finished hour are abstract, so put them on a realistic year. Say your team produces 20 finished hours of basic e-learning annually — a modest catalog of onboarding, policy, and process courses. At the basic ratio of 79hours per finished hour (Chapman Alliance, 2010), that's:

That's roughly 1,580 hours — close to a full year of one person's working time — and about $197,500 at a ~$125/hour loaded rate. This is a deliberately rough, clearly illustrative figure: your real complexity, rate, and review burden will move it in either direction. The point isn't the exact dollar amount. It's that a “small” training program is a six-figure commitment of expert time, and the first stretch of that time goes into drafting from scratch.

Want to put your own numbers in? The interactive ROI estimator lets you adjust finished hours, complexity, labor rate, and the review hours you'd keep — it never assumes review drops to zero.

The takeaway

Training costs what it costs because it's built by hand, one finished hour at a time, and then rebuilt every time the source changes. The benchmark numbers — $102.8 billion a year across U.S. organizations (Training Magazine, 2025), ~79 hours and ~$10,054for a single finished hour of basic e-learning (Chapman Alliance, 2010) — all point at the same bottleneck: skilled people spending their scarcest hours on the blank page.

That's the part worth changing. Scormy One turns the documents you already trust into a source-grounded first draft — outline, scripts, and quizzes — so your experts start from review instead of a blank page, with every claim tracing back to the page it came from. It doesn't remove the human review your team owns. It just gives them a far better place to start.