How Long Does It Take to Develop an Hour of eLearning?

The sponsor wants a 30-minute course, and the launch date is already on the calendar. Then someone asks, "So, how many hours is that?" and the conference room goes quiet. "How long does it take to develop eLearning?" sounds like a simple question, but the honest answer can vary by a factor of ten. So how long does it really take to develop one hour of eLearning?

The most widely cited benchmark, a 2010 study by the Chapman Alliance, found that one finished hour of eLearning took about 79 hours to develop for basic content, about 184 hours for moderately interactive courses and about 490 hours for highly interactive ones. The ranges were wide: roughly 49 to 125 hours for basic, 127 to 267 for moderate and 217 to 716 for advanced. Content readiness, review rounds, custom media and team experience determine where a project falls within those ranges. Today, AI tools can shorten the writing, media and coding portions of the job, but reviews, decisions and testing still require human time, so measure your own savings rather than assuming a number.

What Do the Published Benchmarks Say?

Chapman Alliance surveyed 249 organizations about how long their courses took to build and published the resulting ratios in 2010. The figures are old, but they remain the standard starting point for most teams because nobody has replaced them with a study of comparable breadth. The research grouped eLearning into three levels of interactivity and also measured instructor-led training (ILT) for comparison.

The chart below shows the low, average and high development ratios for each category of training.

Here's what each interactivity level typically means in practice:

• Level 1 (basic): text, graphics, maybe narration, simple navigation and a quiz. Think page-turners and policy overviews.
• Level 2 (moderate): more interactions, such as click-to-reveal, drag-and-drop, short scenarios, audio and custom graphics.
• Level 3 (advanced): branching simulations, games, complex scenarios and custom programming.

How Do You Turn the Ratios into a Real Estimate?

To build an estimate, multiply the finished seat time by the appropriate ratio. A 30-minute Level 2 course at the average ratio works out to 0.5 × 184, or about 92 hours. The low end of the range gives about 64 hours, and the high end gives about 134. That spread is your honest planning range before you know much else about the project.

Try it with your own numbers. The estimator below uses the same Chapman ratios and lets you add an assumption for time saved with AI tools.

Then adjust the figure for your situation. If the content is final and the subject matter expert (SME) is responsive, aim toward the low end of the range. If the content is still being written, or you expect six separate reviewers, estimate high and document that assumption in the project plan.

Is eLearning Slower to Build than Classroom Training?

Usually, yes. In the same Chapman Alliance study, instructor-led training averaged about 43 development hours per hour of classroom time, compared with 79 for even basic eLearning. The difference makes sense when you consider the delivery: a trainer in a room fills gaps, answers questions and adjusts on the fly. A self-paced course has to anticipate all of that in advance, on screen, with no facilitator there to explain.

The trade-off appears later, because classroom training costs trainer time, travel and facility time every time it runs. eLearning requires more investment up front and very little per learner after launch. For a course that hundreds or thousands of employees will take, the additional build hours usually pay for themselves.

What Makes eLearning Take Longer to Develop?

The interactivity level sets the starting range, and the following factors determine where in that range your project will land:

1. Content readiness. Final, approved source material is the single biggest time saver. Content that's still changing is the biggest time sink.
2. Review rounds. Each round costs time to gather, reconcile and apply feedback. Unlimited rounds mean an unlimited schedule.
3. Number of reviewers. More voices bring more conflicting comments to sort out.
4. Custom media. Original illustrations, video shoots and professional narration add days.
5. Templates. A tested template with masters and standard interactions cuts build time a lot.
6. Team experience. A developer who has built similar courses works much faster than one learning the tool.
7. Accessibility and translation. Both are worth doing and both add real time when planned late.

Where Do the Hours Actually Go?

The actual build in the authoring tool is often less than half of the total effort. Using a typical ADDIE-style process (Analysis, Design, Development, Implementation, Evaluation), the hours spread across analysis and design, storyboarding and scripting, media, building, reviews and quality assurance, and LMS testing.

That's why the promise to "just build it faster" rarely works, because if reviews take three weeks, even a very efficient developer won't change the launch date. The calendar and the work hours are two separate problems, and you need to plan for both.

Do Rapid Authoring Tools Change the Benchmarks?

They shift the numbers, mostly for Level 1 work. Block-based tools like Rise 360 make it quick to turn approved text into a clean, mobile-friendly course, so a simple policy module can come in well under the old Level 1 average. Templates and asset libraries provide similar efficiencies in slide-based tools.

Rapid authoring tools don't help much with the difficult parts of Level 2 and Level 3 courses. A branching scenario still requires someone to write believable situations, map the consequences and test every path. The software speeds up screen production without speeding up the instructional thinking, which is why the higher levels remain expensive regardless of the tool you choose.

Work Hours vs Calendar Time

An estimate of 92 hours does not mean a course ships in 92 hours. Say one developer can devote about 25 focused hours a week to the project, between meetings and other responsibilities, which works out to roughly four weeks of build time. Now add two review rounds at a week each, a few days for the SME to answer questions, and a week of learning management system (LMS) testing and corrections. Your 92-hour course has quietly become a two-month project.

When someone asks "how long will it take," find out whether they mean effort or launch date. Provide both numbers, and show the review windows as separate line items. That's often the moment the sponsor realizes that faster approvals will move the date more than any developer can.

How to Cut Development Time Without Cutting Quality

Before investing in any new tool, adopt these habits, which reliably pull a project toward the low end of the range:

• Get content signed off before storyboarding starts.
• Name a single decision-maker for review conflicts.
• Limit review rounds to two or three, agreed in writing.
• Build from tested templates with standard interactions.
• Prototype one lesson first and fix the look and navigation there.
• Test in the LMS in week one, not the last week.

None of these practices is glamorous, but all of them work, and they combine well with the AI savings described below.

How Do AI Tools Change eLearning Development Time?

AI tools shorten the parts of the job that involve producing a first draft, and because that represents a large share of the work, the change is genuine. Here's where I consistently see the time come back:

• Outlines and storyboards. ChatGPT, Claude or Microsoft Copilot can turn source documents into a draft outline and screen-by-screen script in minutes. A good AI course outline gives the SME something to react to on day one instead of week two.
• Quiz items. AI drafts questions and plausible wrong answers that you then sharpen against the objectives.
• Narration. Text-to-speech tools such as ElevenLabs produce natural scratch audio early and sometimes final audio for internal courses.
• Images. Generated images can fill gaps that would otherwise mean a stock photo hunt or a custom illustration.
• Code. Assistants like GitHub Copilot help with small scripts and custom interactions. GitHub published research in 2022 in which developers using Copilot finished a coding task 55% faster than those without it. That's about coding, not courses, but the coding slice of a Level 3 build benefits in a similar way.

What AI doesn't shrink much is the human portion of the work: deciding what matters, getting SMEs to agree, verifying accuracy, testing in the LMS and securing sign-off. Those steps have always consumed a large share of the calendar.

Where AI still needs you: every AI draft gets a human check for accuracy, tone and accessibility before a reviewer sees it. Building that check into the plan is what keeps the time savings from turning into rework.

So what number should you use? Start with your own baseline by tracking hours, phase by phase, on two or three projects with and without AI assistance. Then enter your measured savings into the estimator above, because a measured 20% is worth more than an assumed 50%.

How to Estimate Your Next eLearning Project

Here's the method I use when someone needs a number by Friday:

1. Confirm the finished seat time. Not slide count, actual minutes for a typical learner.
2. Pick the interactivity level for each section, not the whole course. Many courses are mostly Level 1 with a Level 2 or 3 core.
3. Multiply each section's minutes by its ratio and add them up.
4. Adjust for content readiness, reviewer count and custom media.
5. Subtract only the AI savings you've actually measured.
6. Add the calendar time for reviews separately from the work hours.

Here's how step 3 plays out in a realistic example. Say a 45-minute compliance course is 30 minutes of straightforward policy content (Level 1) and 15 minutes of scenario practice (Level 2). At the average ratios, that's 0.5 × 79, or about 40 hours, plus 0.25 × 184, or about 46 hours. Call it 86 hours of effort before adjustments. Price the whole thing at the Level 2 ratio and you'd have quoted about 138 hours, which is the kind of gap that makes a sponsor stop trusting the next estimate.

Write your assumptions next to the estimate so that when the scope changes, and it will, you can show exactly which assumption moved.

So, How Long Does It Take to Develop eLearning in Practice?

Plan on roughly 50 to 125 hours per finished hour for basic courses, 125 to 270 for moderately interactive ones, and 220 to 700 or more for advanced simulations, based on the Chapman Alliance ratios. Where you land depends on content readiness, reviews, media and your team's experience.

AI tools now trim the drafting, media and coding work, sometimes considerably, but they don't remove the decisions and reviews that set the calendar. Use the published ranges to set expectations, measure your own projects, and let real data, not hope, shape your next estimate.

Your next step is small: take the course on your desk, split it into sections by interactivity level, and run each one through the estimator above. Then put the review windows on the calendar as separate line items. If the range still feels uncertain, ask an experienced developer to review the scope before you commit to a launch date. A one-hour scoping conversation is far cheaper than a missed deadline.

← All articles