Matching is a form of search in which the object is to create a relationship, such as marriage or employment (marriage could be regarded as a form of employment). Matching is more complex and often more protracted than most searching for goods or services because the stakes tend to be greater. The costs of exit may be high and often there are high opportunity costs as well. The higher these costs of mistaken matching, the more it pays to invest in the search for a good match. But a protracted search is very costly; for example, if one spends 10 years searching for the perfect mate, one has lost 10 years of benefits of marriage.
If one divides the process of searching for a match into two parts, which I'll call "screening" and "meeting," one can see more easily the precise benefits of Internet matching services. The universes of potential mates, workers, and jobs are immense, and while very few members of these universes are suitable candidates for a particular match, it is very difficult to determine in advance who those few might be. So the first task in the search for a match is to screen out the vast number of unsuitable candidates, and this is done much more quickly, completely, and accurately (and therefore costs less and confers greater benefits) by an Internet search than by such traditional alternatives as mixer dances, marriage brokers, the personal columns in magazines, wanted ads, and job fairs. Once the candidates for a match have been reduced to a manageable number, the more time-intensive, face-to-face meeting phase of the matching process takes over; an efficient preliminary screening greatly reduces the aggregate costs of the expensive "meeting" phase of matching. In job search, the meeting phase is the interviewing of the most promising candidates.
However, two forms of non-Internet marriage-matching screening should be mentioned, as they are also relatively new (and increasingly common) and highly efficient. They are coeducational higher education and gender integration of the workplace. In effect, both the college, which wants a homogeneous student body, and the employer, who wants a homogeneous work force within each job category, do the prescreening at no cost to the searcher, and the fact of taking classes with, engaging in extracurricular activities with, or working side by side with someone of the other sex reduces the incremental costs of the "meeting" phase of the search: you don't have to go out of your way to meet someone you work with or go to class with.
My guess would be, therefore, that the demand for Internet marriage screening would be less among students and among young workers in workplaces that have a fairly even balance of men and women, and greater among persons who are not in such advantageous situations and among persons who have high opportunity costs of time, like successful businessmen and professionals. It should also be greater among persons who have idiosyncratic or minority tastes. For example, since homosexuals are a relatively small fraction of the population, I would expect their demand for Internet match-screening services to be proportionately greater than that of heterosexuals. Personal ads in magazines that have a specialized readership are another form of non-Internet preliminary screening.
The advantages of Internet job matching over newspaper want ads are particularly great, and this for four reasons. First, in any community in which there is more than one newspaper, the job searcher (whether looking to be hired or to hire) has to buy and read both newspapers (or however many there are), even though he may derive no value whatsoever from anything in the second through nth papers except the want ads. Second, a job search is often regional or national rather than local, and it is infeasible to do a regional or national job search by means of local newspapers. Third, the costs of paper greatly limit the number of jobs that can be advertised in a newspaper and the amount of information that can be conveyed about each job or job hunter. And fourth (though related to the first point), a newspaper is a bundled commodity, and people searching for jobs or workers may derive very little value from the other sticks in the bundle.