UPDATE NOTE: From August 4, 2010 – August 31, 2010 we are in the process of updating this information. Click here to participate in the research study. The full results will be shared with participants and later with everyone who wants it.
How long does it take to create different kinds of learning? Here are the answers extracted from several Brandon Hall Research reports. I have reconstructed them into an easy-to-read view; because so many people ask for this information.
Ratio for each Type of learning
34:1 Instructor-Led Training (ILT), including design, lesson plans, handouts, PowerPoint slides, etc.
33:1 PowerPoint to E-Learning Conversion. Not sure why it takes less time then creating ILT, but that’s what we discovered when surveying 200 companies about this practice
220:1 Standard e-learning which includes presentation, audio, some video, test questions, and 20% interactivity
345:1 Time it takes for online learning publishers to design, create, test and package 3rd party courseware
750:1 Simulations from scratch. Creating highly interactive content
==================================================
Here are the bibliographies for each, in case you want to cite these in research:
34:1 Chapman, B. and the staff of Brandon Hall Research (2007). LCMS Knowledgebase 2007: A Comparison of 30+ Enterprise Learning Content Mangement Systems [online database, no page numbers]. Published by Brandon Hall Research, Sunnyvale, CA
33:1 Chapman, B. and the staff of Brandon Hall Research (2006). PowerPoint to E-Learning Development Tools: Comparative Analysis of 20 Leading Systems. Published by Brandon Hall Research, Sunnyvale, CA, p. 20.
220:1 Chapman, B. and the staff of Brandon Hall Research (2006). PowerPoint to E-Learning Development Tools: Comparative Analysis of 20 Leading Systems. Published by Brandon Hall Research, Sunnyvale, CA, p. 20.
345:1 Private study, done for a consulting client, information was not published. No bibliographical reference.
 750:1 Chapman, B. and the staff of Brandon Hall Research (2006). Online Simulations 2006: A Knowledgebase of 100+ Simulation Development Tools and Services [online database, no page numbers]. Published by Brandon Hall Research, Sunnyvale, CA
Workplace Learning Today
Subscribe by Email
Follow Bryan on Twitter
Connect on Facebook
Connect on LinkedIn
{ 12 trackbacks }
{ 29 comments… read them below or add one }
wow nice
Hi Bryan, I’m still alive.
The ratios here are very useful as mean averages, if perplexing in the case of the PowerPoint conversions (unless they mean they start with an expert’s pre-existing PPT, then spruce it up for Flash conversion with questions, animations, audio, and etc.) My question: Are the hours reflective of the “D” and “D” in ADDIE, or the full-monty “A” as well?
When we asked the question as part of the PowerPoint to e-Learning Study (See http://www.brandon-hall.com/publications/ppt/ppt.shtml), we asked end-users to list their average time considering the entire ADDIE process, so it should represent “design” as well as “develop.” Of course, it is sometimes difficult to tell how they interpret the question.
Yes. it was a curiosity that it took less time than classroom prep.
I am assuming that the value for the ratio numbers is:
Development Hours: Delivery Hours
Is that correct?
Yes. The ratio is development; including front end analysis, designa nd development vs. 1 hour of “seat time”
“What’s the basis of your development ratios (instructor-led content)” – I have been asked this question a zillion times, but have never been able to substantiate it with data except for comments available on blogs.
Would it be possible to get some data that I could throw back. I >REQUEST
34:1 Instructor-Led Training (ILT), including design, lesson plans, handouts, PowerPoint slides, etc.
What would this training package include for students? PowerPoint slides, step-by-step demos and workshops, supporting content for both students (and maybe some for instructors)? If not, what material would be developed in this timeline? Trying to compare to our current development model to assess if we take too long or too little time in the development phase compared to others in the industry.
Also, is this average based on a particular industry or many difference ones?
When calculating how long it took to develop one finished hour of classroom instruction, survey respondents were asked how long it took including: front-end analysis, design document, outline, lesson plans, study guides, handouts, tests, PowerPoint and all supporting materials. So the average is based on holistic development. However, not all courses have PowerPoint slides or online tests, etc. The average is across all types.
It would be interesting to redo the survey and itemize length of development in each area (although I suspect that many respondents wouldn’t even know the itemized number for themselves – some do, most don’t).
I agree with PACO. Anyone can argue with these findings and its hard to counter these with believable data. Also, it can vary across industry segments.
In most organizations all aspects of the time involved in the development process is not measured and if it is measured it is very often inaccurate.
Hey Bryan,
If one wants to compare apples to oranges, here are some simulation development prices: http://clarkaldrich.blogspot.com/2007/02/costs-for-simulation.html
There is even a question at the end that I could not answer regarding you!
Anyway, this is great data, very useful for the industry, and thanks for getting it and now updating it.
Wow! This is GREAT. I’ve been in the DE business for about 15 years, and this is the first time I’ve ever seen anybody get this stuff together. Any chance at a peek at the raw data?
Bill McNutt
Chief Technology Officer, AEPro
As for me I create my learning courses using PowerPoint (great tool!) and then convert it to flash with iSpring Ultra (http://www.ispringsolutions.com) which wrapes it into player and makes it easy to publish to any LMS.
I have not seen any numbers for developing using SCORM. Is there a percentage of increase based on this to your published numbers?
Add to my Bookmarks
This is most useful! Would you say this could be applied to language training as well!?
I just want to thank you for putting this together. Once again today a question came up (this time on the eLearning Guild discussion forum), and I appreciate being able to point people to your post. This is still useful to me all these months later. Thanks!
For “345:1″, is this metric for external vendors to create custom eLearning for a client?
Actually, this metric is for vendors who produce off-the-shelf courseware – create once — sell many (i.e. SkillSoft, ElementK, Mindleader, etc.). Please note that this data was captured several years ago and hasn’t been re-validated. I remember that NETg was still around when we did the study. We wondered if they would create faster because they create hundreds of courses; but what we discovered is that more time is spent making sure they get it right the first time, lots of content reviews, testing on multiple platforms, and quality assurance. Hope this helps clarify.
Hi Bryan,
Thanks for clarifying. Do you know where I might be able to find this data? Also interested in knowing how long it takes to manage a learning vendor. I know this is not a straightforward question given the complexity of any given learning program. Would be nice to know if there are any data at all.
Can you provide any insight into the number of programs considered (N) and the Standard Deviations? I find your averages interesting and would like to gain some idea of the variance.
Would the formula work in reverse? Meaning how would you compute for eLearning to ILT? Example: 10 hours of eLearning development including simulations, assessments, etc. converts into ____ hours of developed ILT.
Hi Bryan,
With the wide variety of rapid development tools now available and the continuing maturing of the e-Learning industry, do you believe that these ratios are dropping? (Especially the 220:1)
Best Regards
Brett Christensen
Hi Bryan,
Thanks for sharing its useful for me.
Hi Brian,
As everybody agrres, these numbers have changed in recent times with factors affecting like – technology, expectations, competition, enhanced knowledge and exposure, reusable objects etc. Is it possible to have latest numbers for 2009-10?
Regards
Kapil
What would be the relative effort for each of the stages of the ADDIE model?
Yoiu might consider adding as a reference and comparison TRADOC Pamphlet 350-70-2 which gives estimates for developing IMI for US ARMY training courses.
how many pages of content is required for 1 hr of ILT?
Great idea, however you need better descriptions of what each category is in your survey. For example specifically what tasks do YOU define as Instructional Design versus Authoring. I was making assumptions as I was taking your survey, which of course concerns me… Also by “hour” were you referring to CUE’s? Isn’t around 30-35 pages (or screens) the equivalent of 1 “hour”? Please update the survey with additional clarifications. In addition, you need to better clarify the costs, as one developer may plug in what they get payed per hour, versus what the real cost is takes a company to pay for that “hour” that the developer made.
Hi Susan;
Thanks for your feedback on the survey. Great comments.
On your issue of defining the difference between “instructional design” vs. “authoring,” we’ve done this type of survey many times over the past several years and that’s the first time I’ve ever had anyone ask about the difference. I’ve always seen authoring as part of the overall instructional development process, but haven’t thought of it as “instructional design.”
There is always a tricky balance when creating a survey between brevity (for efficiency) and level of detail.
As for the definition of what constitutes one finished hour of online courseware, I’ve run into many different definitions of what constitutes one finished hour (such as your 30-35 screen example). This definition might work for pure Level 1 courses (pages followed by tests), but how do you factor in an interactive exercise that may take one learner 2 minutes to complete and another 30 minutes to complete. The lines even gets more fuzzy as you move to Level 3 courseware, which includes simulations and serious games. Therefore, we have defined one finished hour very broadly as the average time it takes a learner to go the course offering (for some faster, for others slower; but overall an average of one finished hour). It seems to serve the purpose well. I would appreciate any ideas for locking this down in a way that accommodates the definition for many different groups. Another example, I have heard some people define 1 finished hour as the time it takes to read through the screens at a mid-level pace from front to back. Again, this might work for Level 1 courses, but the model quickly breaks down for Level 2 and Level 3 courses (as defined in the survey).
Finally, I very much appreciate your feedback about internal costs. Our assumption was that the question is worded to capture overall cost (actual) versus the time it takes multiplied by the time spent by the developer. Next go round, we’ll try and make that more specific.
Thanks again for your feedback.