Can adopter categories be identified

Firms were classified into five adopter categories consistent with Rogers' adoption curve. Classification was based on the number of years the company had a web site. The longest length of time that any company in the sample had a web site was 10 years, the most recent related to those companies who were in the process of developing a web site. Rogers (1995) uses the mean and standard deviation as key statistics by which to identify adopter categories. Continuous data is required to perform the classification in this manner. Two separate questions were used in the questionnaire to capture the data on web site adoption. One question (categorical) asked respondents whether they had a web site, did not have a web site or were in the process of developing a web site. For those companies which had an "established" web site, they were asked approximately how long they had had the web site for. The question was open-ended and elicited a combination of numerical and alphanumerical responses. While the majority responded giving the number of complete years, those with web sites for less than a year responded in a variety of ways (for example "six months", "half a year" "a couple of weeks") which meant that the data had to be treated as categorical.

Thus, classification of adopter categories could not be performed strictly according to Rogers' (1995) recommendation. Despite this, for those companies with web sites for one year or more, the mean length of time could be calculated and was 3.5 years with a standard deviation of 1.91. According to Rogers (1995), the cut-off point for the innovators should be 7.4 years (mean + 2SD) and the cut-off point for the early adopters should be 5.49 years (mean + SD). While not entirely reliable, it does offer some indication of classification boundaries. Since Rogers' recommendation could not be applied strictly, an alternative means of classification based on categorical groupings was considered.

Beatty et al. (2001) conducted a study into web site adoption and classified companies into adopter categories based on natural groupings in the data. Examination of the frequency of responses for length of time in this study indicated natural groupings as indicated in Table II.

Table II Adopter categories

Category

Timeframe

Frequency

Percentage

Innovators

Between 7 and 10 years

1G

4.8

Early adopters

At least 5 years but less than 7 years

42

2G.1

Early majority

At least 3 years but less than 5 years

6B

BG.1

Late majority

Less than 3 years

49

2B.4

Laggards

Currently developing site

45

21.5

Total

2G9

1GG.G

Beatty et al. (2001) note that commercial web use is estimated to have started around 1993, however the earliest date which any company in this sample developed a web site was in 1994 one year after this date. Due to the differences in time of first adoption, the overall timeframe of the study, and the industrial context of the sample, it is not possible to retain complete parity with the study of Beatty et al. However, some overlaps exist in terms of the timeframe for innovators (adopting around three years before the early adopters), the two-year timeframes of early adopters and the early majority, and the classification of laggards as those currently developing a web site. The main difference is in terms of the late majority, which is less than three years in this study and less than one year in the study by Beatty et al. (2001). Combined with the results from the means and standard deviation analysis, the categories presented in Table II seem to be appropriate.

In terms of how this relates to Rogers' (1995) original model, Figure 1 shows the comparison. While the proportions within each of the categories are different, the general shape and pattern of the curve is similar for the first three groups. The last two groups do not show the same tailing off as in the Rogers' (1995) model. An explanation for this could be that Rogers' model commonly refers to the adoption of an innovation throughout its life-cycle, bearing a similarity to and consistency with the product life-cycle curve. The web could be argued to be at the mature stage of its life cycle; it has certainly not reached the end of its life, and this would

0 0

Post a comment