A FIGURE OF MERIT FOR NON-PROFIT WEB SITES

Web sites of enterprises, I suppose, ought to be managed to achieve the objectives of the enterprises themselves. For a commercial enterprise, one assumes that the motive is profit. For a non-profit, how does one measure success? The figure of merit discussed in the following paragraphs will not really measure success, but still may be useful in managing and improving a web site.

I suggest that a web site should have a target audience, and that indeed it is usually helpful to disaggregate that audience into groups. Thus a web site may have a preferred target audience forming a “community of practice”, a secondary target of people from related communities of practice, and additional audiences of lesser perceived importance such as students, members of a larger “community of interest”, media, etc. I suggest not only that it is frequently useful to disaggregate the intended audience into such groups, but also to define a weight identifying the importance of serving each group. Thus one can define:

w(i) = the weight (say one for least important to 5 for most important) to assign to group i

In general I suspect that one would be interested in two measures of the use of the site by each target group:

n(i) = the number of people in group i using the site, and
f(i) = the average frequency (e.g. times per month) that users from group i visit the site.

One could estimate these figures in some cases from the email addresses of visitors to the site. For example, if one were seeking to attract an audience from certain countries, one could look for domain names from the countries. Alternatively, one could use simple pop-up questionnaires with a portion of the users of the site to estimate to which group they belonged, and how frequently they used the site.

I would suggest that such survey methods also be used to estimate two other parameters:

y(i) = the average importance (say on a scale of 1 to 5) that visitors from group i ascribe to the purpose of his/her visit, and
s(i) = the average satisfaction (say on a scale of 1 to 5) that visitors from group i ascribe to the outcome of the visit.

The figure of merit would then be simply the sum of:

w(i) * n(i) * f(i) * y(i) * s(i)

where the sum is taken over the groups of users.

I suggest such a figure of merit would be useful in tracking the changing value of a web site. It would be useful in allocating resources to improving the site. Thus one might decide that the figure of merit was most sensitive to improvements of one or another kind. If the issue limiting the figure of merit is not the number of eyes attracted, but rather that the users are not satisfied once attracted, there is one solution. If alternatively, the site was getting a lot of visitors, but most from relatively low priority user groups, different steps to improve the situation might be appropriate.

Of course data from survey instruments I am suggesting could be used for more sophisticated purposes. Thus if the users were satisfied with the site for their less important needs, but not for the more important, that would be a significant finding that could be made from cross tabulation of responses to two questions.

I would note that I have specifically left out a commonly used indicator – time on site. Satisfaction of user needs seems more likely to be important to non-profits. (Commercial sites seem to believe that the longer the user is at the site, the more that they will sell; for a non-profit, helping the user get what he/she needs as quickly as possible may be an even more important objective.)

The figure of merit approach is suggested in the context of a sustained effort by those responsible for a web site to improve that site. Variants of the formula given above would probably be more useful for specific non-profit sites than the formula itself. The important thing is to think what the site is to achieve, rather than simply accepting commonly used indices because they are available.