eLearning development time and cost calculator

Someone has asked you how long the course will take to build, and they want a number today. You know the honest answer is that it depends, and you also know that “it depends” is not going to survive contact with the meeting. What you need is a defensible starting figure and a clear account of what it assumes.

The calculator below uses the ratio the field has cited for more than fifteen years: hours of development work per one finished hour of learning. Enter the seat time, pick the honest complexity level, and it returns development hours, a cost if you supply a rate, and how many calendar weeks that becomes at the pace you can actually work. The number will probably be larger than the one in your head, and that gap is the part worth taking into the meeting.

Development hours = finished hours x the ratio for your complexity level and chosen position in its range.

Cost = development hours x blended hourly rate.

Calendar time = development hours divided by the hours per week you can actually give it.

AI-adjusted hours = development hours x (1 minus your measured time saving).

Why one hour takes a hundred

The first time someone sees the Level 1 average of 79 hours of work per finished hour, they assume it is padded. It is not. The finished hour is the last thing that happens, and the ratio covers everything before it: the analysis that decides what the training is even for, the outline, the script, the storyboard, the media, the build in the authoring tool, the quality pass, the accessibility work, the review cycle with subject-matter experts, and the rebuild after the first review comes back.

People estimate the build and forget every phase around it. That is why internal estimates come in low by a wide margin, and why the schedule slips at the review stage rather than the production stage. If your own number is far below what the tool returns, the question worth asking is not whether the benchmark is wrong. It is which phases your estimate left out.

Pick the honest level, not the flattering one

Level 1 is text, images, narration, and a quiz at the end. Level 2 is genuine interaction: branching, drag-and-drop, scenario questions where the wrong answer leads somewhere, custom media built for this course. Level 3 is simulation, software emulation, or anything with a scoring model behind it.

Most corporate training that gets described as interactive is Level 1 with a few clicks in it. There is nothing wrong with that. Level 1 done well beats Level 2 done badly, and a lot of Level 2 gets specified because it sounds better in a proposal rather than because the learning needs it. What matters for your estimate is that choosing Level 2 when you are building Level 1 inflates the number by more than double, and choosing Level 1 when you are actually building branching scenarios is how a project ends up four months late.

What the tool will not tell you about AI

The AI time saving on this form defaults to zero, and it stays at zero unless you enter a number you have measured yourself. That is deliberate.

There is no credible published benchmark for AI-assisted development ratios. Not a cautious one, not a conservative one, none. Plenty of vendors will quote you a percentage, and those percentages are marketing rather than research. I would rather ship a calculator that admits a gap than one that launders a sales figure into a planning document that someone takes to a budget meeting.

What I can tell you from doing the work is that AI moves effort more than it removes it. Drafting gets faster. Verifying gets slower, and it gets slower in a specific way: a draft that reads well hides its errors better than a rough one does, so the review that used to catch problems by feel now has to be deliberate. Teams that report large savings are often measuring the drafting and not the checking. If you want a real number, run one module both ways, log the hours including review, and put your own figure in the field. That measurement is worth more to you than any benchmark, mine included.

Where the ratios come from

The numbers are the Chapman Alliance benchmark, drawn from roughly 250 organizations and 4,000 learning professionals. Level 1 runs 49 to 125 hours per finished hour, Level 2 runs 127 to 267, Level 3 runs 217 to 716, and a classroom session runs 34 to 49.

That study is from 2010, and I would rather say so than quietly present it as current. It is still the most-cited benchmark in the field, and the reason it has held up is that the changes since then have pushed in both directions. Authoring tools got better and template libraries cut build time. Accessibility conformance, mobile testing, and review governance added work that barely existed in 2010. The two roughly cancel, which is why the ratios still land close for most teams. Treat the result as a planning anchor you adjust with your own project history, the way any experienced estimator treats any benchmark.

What this does not cover

This is development labour only, the cost of making the content. Platform licensing, hosting, learner administration, and the ongoing cost of keeping the material current are separate lines. If the platform side is the part you are budgeting, the WordPress LMS cost calculator works that out against Moodle and Canvas.

Everything here runs in your browser. Nothing you enter is sent anywhere, stored, or logged, which matters when the number you are typing is a salary figure or an unannounced project budget.