Abstract
D.Com. (Marketing Management)
This study specifically relates to disintermediation and reintermediation in the retail travel
industry in South Africa. Brick and mortar travel agencies in South Africa and across the
world are losing customers due to the advent of the Internet. The Internet has modified
the way traditional distribution channels are structured and how they operate. The
distribution process involves a product or services evolution from raw materials to the
sale of the final product. Intermediaries facilitate the flow in this process by adding value
but also adding costs. Disintermediation refers to removing or bypassing intermediaries
in the distribution process to create shorter supply chains. The core rationalisation for
this process relates to costs of intermediaries versus the value that intermediaries
provide.
If travel agents are to remain relevant in the travel distribution industry, they need to add
value to the travel distribution process again and thus the reintermediation of their
business. Reintermediation can be defined as those previously disintermediated
intermediaries who re-enter the distribution process by offering additional value to
customers or suppliers. The primary objective of this study was thus to develop a model
that assists dissintermediated travel agencies to retain customers via the process of
reintermediation. A range of reintermediation strategies have been suggested for the
travel agency industry such as targeting niche markets, the production of customised
travel packages, the adoption of online strategies, the hiring of more professional and
experienced staff and becoming more customer service orientated. Eight proposed
reintermediation factors were highlighted and explained and evidence was provided to
support their inclusion in the model.
The research methodology employed a 16 step process which encompassed a
qualitative and quantitative component. The qualitative phase of the research
incorporated focus groups with travel agency owners and managers in the Eastern
Cape. Four mini-focus groups were held with a total of 12 owners and managers in this
area. The fundamental aim of the focus groups was of a confirmatory nature in that
participants’ opinions were sought with regard to the composition of the proposed
reintermediation model.
The quantitative phase of the study incorporated an electronically administered survey
with 200 travel agency owners and managers in SA, of which only 149 questionnaires
were eligible for analysis. A questionnaire was developed for the survey using an array
of existing scales that were modified to suit the purposes of this study. The questionnaire
was designed in such a manner, that there were two columns for respondents to rate the
statements which were centred in the questionnaire. The left column required
respondents to indicate their current opinions with regards to the status of the current
reintermediation factors within their businesses, while the right column asked
respondents to rate the importance of the proposed reintermediation factors within their
businesses. Essentially this resulted in the development of two models: Model 1
reflected the current opinions of travel agency owners and managers regarding
reintermediation in their businesses, and Model 2 referred to the opinions of travel
agency owners and managers in terms of the importance of the proposed factors in
reintermediating their businesses and thus retaining customers.
There are three commonly consistent reintermediation factors across Models 1 and 2;
they are: differentiated product (leisure), high quality service and co-exist with
technology. Thus, in terms of reintermediation factors these three can be considered to
be the most important reintermediation factors for travel agents in South Africa. Model 1
contains five of the proposed reintermediation factors while Model 2 only contains four of
the proposed reintermediation factors as presented in Figure 1.2 in Chapter One. Thus
Model 1 was considered the superior model of the two. In addition, Model 1 has better fit
statistics than Model 2.