Fall 2000



“How much,” the LiNE Zine editor asked me one day, “does elearning cost?” I told her that was a difficult question to answer. Apparently oblivious to the value of my time, she responded, “I didn’t say it was going to be easy for you to answer.” To which I retorted, “Let me be more clear. That is an impossible question to answer. It all depends on the circumstances.” 

That conversation was the genesis of this article, which outlines the major components of the cost of both traditional and eLearning. I still maintain that the question is an impossible one to answer in the abstract. One of two things is true if someone claims to have “the” answer: (1) they don’t know what they are talking about, or (2) they are trying to sell you something. 

Nonetheless, there are clearly identifiable categories of costs. Some are unique to elearning, some are unique to traditional training, and some are shared by both. By working through these various categories of cost for the particular circumstances you face, you should be able to derive responsible estimates of the total cost of elearning, as well as how it compares to traditional training.

First—A Few Rules

If you’re really serious about creating credible estimates of costs, there are some iron-clad rules to keep in mind:

Rule #1   The direct cost of developing and delivering learning—whether it is via elearning or traditional training—is just the tip of the iceberg. 

Indirect costs—which include overhead and employees’ compensation while they are learning—are typically far greater than the direct costs of learning. Opportunity cost (the value of lost productivity) can, under some circumstances be greater still. So equating the cost of learning with the cost of developing and delivering can be a big mistake.

Rule #2   Costs and benefits are co-determined. 

It can be inexpensive to develop learning interventionstraditional or elearningthat bore people silly, and/or produce no lasting learning or behavioral changes. (This, by the way, can and is frequently done at greater expense than planned.) Creating highly beneficial learning, however, is rarely inexpensive.

Rule #3   Just because costs can be shifted does not mean that they are zero.

One of the great attractions of elearning is that it enables employees to engage in learning “on their own time.” While this does have the benefit (from the employers’ perspective) of reducing the indirect cost of learning, it does not mean that this cost goes away. Rather, it shifts to employees.

Rule #4   When costs are shifted without a corresponding benefit, they will ultimately be shifted back.

It’s probably best not to say too much about this one, and just let you think about it for a while.

The Components of Cost

The cost of workplace learning has three components: direct, indirect, and opportunity cost.

Direct costs for “traditional” learning interventions, consist of:  wages and salaries of trainers; payments to outside vendors; facilities expenses; development, production and distribution of materials; travel expenses; and administrative and support costs.

For elearning, the direct costs consist of development, purchase, and/or licensing of materials and hardware for delivery. Almost all elements of cost have both marginal costs (those that vary with the number of learners) and fixed costs (those that do not vary with the number of learners).

Indirect costs consist of compensationthe wages and benefits paid to learners while they are learningas well as the overhead costs associated with both the direct and indirect costs. The available evidence suggests that the indirect, compensation costs of traditional learning are typically at least as great as the direct costs. When overhead (which also applies against the direct cost) is added in, the indirect costs of traditional learning are likely to be twice the direct costs.

Opportunity cost consists of the lost productivity that occurs when employees take time away from their work to learn. Although we know little on a systematic basis about the opportunity cost of traditional learning, a reasonable estimate is that it is twice the indirect compensation cost.

Consequently, a conservative estimate of the total cost of typical, classroom learning is that it is approximately five times the direct costs. That explains one of the major attractions of elearning; it holds the promise of reducing all three categories of costs, but perhaps, most significantly, the indirect and opportunity costs of learning. Under some circumstances, firms have estimated the total cost of elearning to be less than half the cost of traditional learning.

However, before you leap on the elearning bandwagon, keep at least two things in mind:

1.   While elearning can eliminate many of the components of direct cost and reduce indirect and opportunity cost, the fixed costs can be very high. Developing high quality elearning is an expensive proposition. The economics of elearning are highly dependent on the number of learners involved. The greater the number of learners, the greater the probability that economies of scale will make elearning an attractive proposition from a cost perspective. Conversely, the smaller the number of learners for any given content, the smaller the probability that elearning will be economically viable.

2.   The relative cost effectiveness of traditional versus elearning is a function of the costs and benefits of both. Putting hype aside, there is very little that is systematically known about this. 

Below is a prototype of a worksheet that could help make a choice between traditional learning and elearning from a cost perspective.


Traditional Learning



Fixed Cost

Marginal Costs

Fixed Cost

Marginal Costs

Direct Cost


Trainers’ compensation




Outside vendors


Materials, development



Materials, production




Materials, distribution












Travel expenses




Administrative support


Indirect Cost


Learners’ compensation






Opportunity Cost




* Indicates that these costs are likely to be (or can be made to be) negligible

The total cost of a learning intervention is calculated as follows:

Total cost = sum of all fixed costs + (sum of all marginal costs X the number of learners)

A quick glance at the table makes the allure of elearning apparentthe number of asterisks on the right side of the table far exceeds the number of asterisks on the left side. Some of the elearning asterisks are (or can be made to appear) in the especially important columns of learners’ compensation and opportunity costs. 

Filling in the elements of this matrix is by no means a trivial assignment. In the past, few companies had tracking systems in place that could identify, with precision, the various cost components of traditional learning. Now elearning is new enough that few have the experience necessary to determine the components of its cost—even if they do have a sophisticated tracking system in place.

Here are some suggestions for tackling these cost estimates:

       Do not try to do it in the abstract (since that is impossible). Rather, consider a specific learning intervention with a well-defined audience and learning objective.

       If you do not have a learning management system in place that enables you to track the historic costs of traditional training, sit down with your finance folks to develop an educated guess of the various elements of cost.

      Work with a reputable vendor who is willing to assist you in estimating the major cost components of elearning that would be incurred in producing an outcome comparable to that produced by traditional learning. (If they cannot address the outcomes that their products produce, you should immediately become suspicious).

The cost estimates that you devise will be highly dependent on a variety of factors associated with the specific learning intervention under consideration. Some of the factors likely to favor one choice over the other, from a cost perspective, are summarized below.          

Factors favoring traditional learning

Factors favoring elearning

The nature of the content requires human feedback and practice

The content can be readily modularized and/or focuses on facts and information

The most effective instructional design is highly interactive and/or a great deal of customization is required

Passive learning is acceptable and/or little customization is necessary

The content has a short-shelf life (in cases where elearning development costs are high)

The content has a long-shelf life (in cases where development costs are high) or short shelf-life (in cases where development costs are low)

Relatively few learners need access to the content

Many learners need access to the content

Learners are highly geographically concentrated, so travel costs are relatively low

Learners are highly dispersed geographically, so travel costs are high

The speed with which the learning is rolled out is not particularly critical

Learning must be rolled out quickly to a large number of learners

Words of Caution

Costs are irrelevant. Cost-effectiveness is what matters. A cheap solution that does not produce the intended result is a waste of money.

As has already been mentioned, when all hyperbole is put aside, very little is known on a systematic, rigorous basis about the cost-effectiveness of one type of learning intervention relative to another. Although some good research has been done on “level 2” assessmentswhat it is that learners actually learnedthat compare traditional and elearning, this research tells us nothing about the comparative success of the two in transfer of learning to behavior and productivity. This represents a huge hole in our understanding. If you would like to see a longer discussion of this point, contact me and I can send you a white paper on it.

Rememberwhen costs are shifted without a corresponding benefit, they will ultimately shift back. Think hard and long if the (only?) attraction of elearning is that it reduces indirect and opportunity costs. It better be really good and beneficial learning—from the perspective of the learner—or else it is only a matter of time (which will probably be short) before rule #4 comes into play.

Be suspicious of silver bullet solutions to complex problems.

Laurie Bassi is a former Georgetown University economics professor and vice president of research for the American Society for Training and Development (ASTD). She now spends her time helping organizations understand how the numbers stand on their own. Last year she wrote a whitepaper on elearning for Saba. If you'd like to receive the whitepaper or ask her to remove her tongue from her cheek, write to her at lbassi@hcdynamics.com




Copyright (c) 2000-2004 LiNE Zine (www.linezine.com)

LiNE Zine retains the copyright in all of the material on these web pages as a collective work under copyright laws. You may not republish, redistribute or exploit in any manner any material from these pages without the express consent of LiNE Zine and the author. Contact permissions@linezine.com for reprints and permissions. You may, however, download or print copyrighted material for your individual and non-commercial use.